Update shared drm code, inteldrm(4) and radeondrm(4) from linux 4.4 to

linux 4.19.34.

Adds support for more Intel hardware:
Broxton/Apollo Lake (was is_preliminary in 4.4)
Amber Lake (another Kaby Lake refresh)
Gemini Lake
Coffee Lake
Whiskey Lake
Cannon Lake (though no hardware with Intel graphics ever shipped)
Ice Lake (alpha support, hardware not released)

This does not add support for new radeon hardware on the AMD side as
newer radeons have a different kernel driver (amdgpu).

Thanks to the OpenBSD Foundation for sponsoring this work, kettenis@ for
helping and a bunch of other developers for testing.

130 files changed, 5021 insertions, 8425 deletions

diff --git a/sys/dev/pci/drm/radeon/atom.c b/sys/dev/pci/drm/radeon/atom.c
index 1ce19696ee0..18658cd05aa 100644
--- a/sys/dev/pci/drm/radeon/atom.c
+++ b/sys/dev/pci/drm/radeon/atom.c
@@ -22,6 +22,11 @@
  * Author: Stanislaw Skowronek
  */
 
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <asm/unaligned.h>
+
 #define ATOM_DEBUG
 
 #include "atom.h"
@@ -61,9 +66,10 @@ int atom_debug = 0;
 static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32_t * params);
 int atom_execute_table(struct atom_context *ctx, int index, uint32_t * params);
 
-static uint32_t atom_arg_mask[8] =
-    { 0xFFFFFFFF, 0xFFFF, 0xFFFF00, 0xFFFF0000, 0xFF, 0xFF00, 0xFF0000,
-0xFF000000 };
+static uint32_t atom_arg_mask[8] = {
+	0xFFFFFFFF, 0x0000FFFF, 0x00FFFF00, 0xFFFF0000,
+	0x000000FF, 0x0000FF00, 0x00FF0000, 0xFF000000
+};
 static int atom_arg_shift[8] = { 0, 0, 8, 16, 0, 8, 16, 24 };
 
 static int atom_dst_to_src[8][4] = {
@@ -168,7 +174,7 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base,
 		case ATOM_IIO_END:
 			return temp;
 		default:
-			printk(KERN_INFO "Unknown IIO opcode.\n");
+			pr_info("Unknown IIO opcode\n");
 			return 0;
 		}
 }
@@ -192,22 +198,19 @@ static uint32_t atom_get_src_int(atom_exec_context *ctx, uint8_t attr,
 			val = gctx->card->reg_read(gctx->card, idx);
 			break;
 		case ATOM_IO_PCI:
-			printk(KERN_INFO
-			       "PCI registers are not implemented.\n");
+			pr_info("PCI registers are not implemented\n");
 			return 0;
 		case ATOM_IO_SYSIO:
-			printk(KERN_INFO
-			       "SYSIO registers are not implemented.\n");
+			pr_info("SYSIO registers are not implemented\n");
 			return 0;
 		default:
 			if (!(gctx->io_mode & 0x80)) {
-				printk(KERN_INFO "Bad IO mode.\n");
+				pr_info("Bad IO mode\n");
 				return 0;
 			}
 			if (!gctx->iio[gctx->io_mode & 0x7F]) {
-				printk(KERN_INFO
-				       "Undefined indirect IO read method %d.\n",
-				       gctx->io_mode & 0x7F);
+				pr_info("Undefined indirect IO read method %d\n",
+					gctx->io_mode & 0x7F);
 				return 0;
 			}
 			val =
@@ -471,22 +474,19 @@ static void atom_put_dst(atom_exec_context *ctx, int arg, uint8_t attr,
 				gctx->card->reg_write(gctx->card, idx, val);
 			break;
 		case ATOM_IO_PCI:
-			printk(KERN_INFO
-			       "PCI registers are not implemented.\n");
+			pr_info("PCI registers are not implemented\n");
 			return;
 		case ATOM_IO_SYSIO:
-			printk(KERN_INFO
-			       "SYSIO registers are not implemented.\n");
+			pr_info("SYSIO registers are not implemented\n");
 			return;
 		default:
 			if (!(gctx->io_mode & 0x80)) {
-				printk(KERN_INFO "Bad IO mode.\n");
+				pr_info("Bad IO mode\n");
 				return;
 			}
 			if (!gctx->iio[gctx->io_mode & 0xFF]) {
-				printk(KERN_INFO
-				       "Undefined indirect IO write method %d.\n",
-				       gctx->io_mode & 0x7F);
+				pr_info("Undefined indirect IO write method %d\n",
+					gctx->io_mode & 0x7F);
 				return;
 			}
 			atom_iio_execute(gctx, gctx->iio[gctx->io_mode & 0xFF],
@@ -817,17 +817,17 @@ static void atom_op_postcard(atom_exec_context *ctx, int *ptr, int arg)
 
 static void atom_op_repeat(atom_exec_context *ctx, int *ptr, int arg)
 {
-	printk(KERN_INFO "unimplemented!\n");
+	pr_info("unimplemented!\n");
 }
 
 static void atom_op_restorereg(atom_exec_context *ctx, int *ptr, int arg)
 {
-	printk(KERN_INFO "unimplemented!\n");
+	pr_info("unimplemented!\n");
 }
 
 static void atom_op_savereg(atom_exec_context *ctx, int *ptr, int arg)
 {
-	printk(KERN_INFO "unimplemented!\n");
+	pr_info("unimplemented!\n");
 }
 
 static void atom_op_setdatablock(atom_exec_context *ctx, int *ptr, int arg)
@@ -990,7 +990,7 @@ static void atom_op_switch(atom_exec_context *ctx, int *ptr, int arg)
 			}
 			(*ptr) += 2;
 		} else {
-			printk(KERN_INFO "Bad case.\n");
+			pr_info("Bad case\n");
 			return;
 		}
 	(*ptr) += 2;
@@ -1024,7 +1024,7 @@ static void atom_op_xor(atom_exec_context *ctx, int *ptr, int arg)
 
 static void atom_op_debug(atom_exec_context *ctx, int *ptr, int arg)
 {
-	printk(KERN_INFO "unimplemented!\n");
+	pr_info("unimplemented!\n");
 }
 
 static struct {
@@ -1180,7 +1180,7 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32
 	ectx.abort = false;
 	ectx.last_jump = 0;
 	if (ws)
-		ectx.ws = kzalloc(4 * ws, GFP_KERNEL);
+		ectx.ws = kcalloc(4, ws, GFP_KERNEL);
 	else
 		ectx.ws = NULL;
 
@@ -1278,14 +1278,14 @@ struct atom_context *atom_parse(struct card_info *card, void *bios)
 	ctx->bios = bios;
 
 	if (CU16(0) != ATOM_BIOS_MAGIC) {
-		printk(KERN_INFO "Invalid BIOS magic.\n");
+		pr_info("Invalid BIOS magic\n");
 		kfree(ctx);
 		return NULL;
 	}
 	if (strncmp
 	    (CSTR(ATOM_ATI_MAGIC_PTR), ATOM_ATI_MAGIC,
 	     strlen(ATOM_ATI_MAGIC))) {
-		printk(KERN_INFO "Invalid ATI magic.\n");
+		pr_info("Invalid ATI magic\n");
 		kfree(ctx);
 		return NULL;
 	}
@@ -1294,7 +1294,7 @@ struct atom_context *atom_parse(struct card_info *card, void *bios)
 	if (strncmp
 	    (CSTR(base + ATOM_ROM_MAGIC_PTR), ATOM_ROM_MAGIC,
 	     strlen(ATOM_ROM_MAGIC))) {
-		printk(KERN_INFO "Invalid ATOM magic.\n");
+		pr_info("Invalid ATOM magic\n");
 		kfree(ctx);
 		return NULL;
 	}
@@ -1319,7 +1319,7 @@ struct atom_context *atom_parse(struct card_info *card, void *bios)
 		}
 	}
 #ifdef DRMDEBUG
-	printk(KERN_INFO "ATOM BIOS: %s\n", name);
+	pr_info("ATOM BIOS: %s\n", name);
 #endif
 
 	return ctx;
diff --git a/sys/dev/pci/drm/radeon/atom.h b/sys/dev/pci/drm/radeon/atom.h
index 0d710ef330a..e34d1650558 100644
--- a/sys/dev/pci/drm/radeon/atom.h
+++ b/sys/dev/pci/drm/radeon/atom.h
@@ -25,8 +25,8 @@
 #ifndef ATOM_H
 #define ATOM_H
 
-#include <sys/types.h>
-#include <dev/pci/drm/drmP.h>
+#include <linux/types.h>
+#include <drm/drmP.h>
 
 #define ATOM_BIOS_MAGIC		0xAA55
 #define ATOM_ATI_MAGIC_PTR	0x30
diff --git a/sys/dev/pci/drm/radeon/atombios.h b/sys/dev/pci/drm/radeon/atombios.h
index ab89eed9ddd..4b86e8b4500 100644
--- a/sys/dev/pci/drm/radeon/atombios.h
+++ b/sys/dev/pci/drm/radeon/atombios.h
@@ -181,7 +181,7 @@
 #define HW_ASSISTED_I2C_STATUS_FAILURE          2
 #define HW_ASSISTED_I2C_STATUS_SUCCESS          1
 
-#pragma pack(1)                                       /* BIOS data must use byte aligment */
+#pragma pack(1)                                       /* BIOS data must use byte alignment */
 
 /*  Define offset to location of ROM header. */
 
@@ -3883,7 +3883,7 @@ typedef struct _ATOM_GPIO_PIN_ASSIGNMENT
 }ATOM_GPIO_PIN_ASSIGNMENT;
 
 //ucGPIO_ID pre-define id for multiple usage
-//from SMU7.x, if ucGPIO_ID=PP_AC_DC_SWITCH_GPIO_PINID in GPIO_LUTTable, AC/DC swithing feature is enable
+//from SMU7.x, if ucGPIO_ID=PP_AC_DC_SWITCH_GPIO_PINID in GPIO_LUTTable, AC/DC switching feature is enable
 #define PP_AC_DC_SWITCH_GPIO_PINID          60
 //from SMU7.x, if ucGPIO_ID=VDDC_REGULATOR_VRHOT_GPIO_PINID in GPIO_LUTable, VRHot feature is enable
 #define VDDC_VRHOT_GPIO_PINID               61
@@ -7909,7 +7909,7 @@ typedef struct  _ATOM_POWERPLAY_INFO_V3
 
 /*********************************************************************************/
 
-#pragma pack() // BIOS data must use byte aligment
+#pragma pack() // BIOS data must use byte alignment
 
 //
 // AMD ACPI Table
diff --git a/sys/dev/pci/drm/radeon/atombios_crtc.c b/sys/dev/pci/drm/radeon/atombios_crtc.c
index 8f3826560ba..efbd5816082 100644
--- a/sys/dev/pci/drm/radeon/atombios_crtc.c
+++ b/sys/dev/pci/drm/radeon/atombios_crtc.c
@@ -23,10 +23,11 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
-#include <dev/pci/drm/drm_fixed.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/radeon_drm.h>
+#include <drm/drm_fixed.h>
 #include "radeon.h"
 #include "atom.h"
 #include "atom-bits.h"
@@ -274,15 +275,15 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode)
 		if (ASIC_IS_DCE3(rdev) && !ASIC_IS_DCE6(rdev))
 			atombios_enable_crtc_memreq(crtc, ATOM_ENABLE);
 		atombios_blank_crtc(crtc, ATOM_DISABLE);
-		drm_vblank_post_modeset(dev, radeon_crtc->crtc_id);
-		/* Make sure vblank interrupt is still enabled if needed */
-		radeon_irq_set(rdev);
+		if (dev->num_crtcs > radeon_crtc->crtc_id)
+			drm_crtc_vblank_on(crtc);
 		radeon_crtc_load_lut(crtc);
 		break;
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 	case DRM_MODE_DPMS_OFF:
-		drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id);
+		if (dev->num_crtcs > radeon_crtc->crtc_id)
+			drm_crtc_vblank_off(crtc);
 		if (radeon_crtc->enabled)
 			atombios_blank_crtc(crtc, ATOM_ENABLE);
 		if (ASIC_IS_DCE3(rdev) && !ASIC_IS_DCE6(rdev))
@@ -588,7 +589,8 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 		if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE61(rdev) || ASIC_IS_DCE8(rdev))
 			radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV;
 		/* use frac fb div on RS780/RS880 */
-		if ((rdev->family == CHIP_RS780) || (rdev->family == CHIP_RS880))
+		if (((rdev->family == CHIP_RS780) || (rdev->family == CHIP_RS880))
+		    && !radeon_crtc->ss_enabled)
 			radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV;
 		if (ASIC_IS_DCE32(rdev) && mode->clock > 165000)
 			radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV;
@@ -625,7 +627,9 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 			if (radeon_crtc->ss.refdiv) {
 				radeon_crtc->pll_flags |= RADEON_PLL_USE_REF_DIV;
 				radeon_crtc->pll_reference_div = radeon_crtc->ss.refdiv;
-				if (ASIC_IS_AVIVO(rdev))
+				if (ASIC_IS_AVIVO(rdev) &&
+				    rdev->family != CHIP_RS780 &&
+				    rdev->family != CHIP_RS880)
 					radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV;
 			}
 		}
@@ -1141,7 +1145,6 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -1152,6 +1155,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	u32 tmp, viewport_w, viewport_h;
 	int r;
 	bool bypass_lut = false;
+	struct drm_format_name_buf format_name;
 
 	/* no fb bound */
 	if (!atomic && !crtc->primary->fb) {
@@ -1159,19 +1163,15 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1190,7 +1190,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(rbo);
 
-	switch (target_fb->pixel_format) {
+	switch (target_fb->format->format) {
 	case DRM_FORMAT_C8:
 		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_8BPP) |
 			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_INDEXED));
@@ -1256,7 +1256,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		break;
 	default:
 		DRM_ERROR("Unsupported screen format %s\n",
-			  drm_get_format_name(target_fb->pixel_format));
+		          drm_get_format_name(target_fb->format->format, &format_name));
 		return -EINVAL;
 	}
 
@@ -1272,7 +1272,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 
 				/* Calculate the macrotile mode index. */
 				tile_split_bytes = 64 << tile_split;
-				tileb = 8 * 8 * target_fb->bits_per_pixel / 8;
+				tileb = 8 * 8 * target_fb->format->cpp[0];
 				tileb = min(tile_split_bytes, tileb);
 
 				for (index = 0; tileb > 64; index++)
@@ -1280,13 +1280,14 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 
 				if (index >= 16) {
 					DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n",
-						  target_fb->bits_per_pixel, tile_split);
+						  target_fb->format->cpp[0] * 8,
+						  tile_split);
 					return -EINVAL;
 				}
 
 				num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3;
 			} else {
-				switch (target_fb->bits_per_pixel) {
+				switch (target_fb->format->cpp[0] * 8) {
 				case 8:
 					index = 10;
 					break;
@@ -1374,6 +1375,11 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		break;
 	}
 
+	/* Make sure surface address is updated at vertical blank rather than
+	 * horizontal blank
+	 */
+	WREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, 0);
+
 	WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
 	       upper_32_bits(fb_location));
 	WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
@@ -1404,7 +1410,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(EVERGREEN_GRPH_X_END + radeon_crtc->crtc_offset, target_fb->width);
 	WREG32(EVERGREEN_GRPH_Y_END + radeon_crtc->crtc_offset, target_fb->height);
 
-	fb_pitch_pixels = target_fb->pitches[0] / (target_fb->bits_per_pixel / 8);
+	fb_pitch_pixels = target_fb->pitches[0] / target_fb->format->cpp[0];
 	WREG32(EVERGREEN_GRPH_PITCH + radeon_crtc->crtc_offset, fb_pitch_pixels);
 	WREG32(EVERGREEN_GRPH_ENABLE + radeon_crtc->crtc_offset, 1);
 
@@ -1426,18 +1432,11 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(EVERGREEN_VIEWPORT_SIZE + radeon_crtc->crtc_offset,
 	       (viewport_w << 16) | viewport_h);
 
-	/* pageflip setup */
-	/* make sure flip is at vb rather than hb */
-	tmp = RREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset);
-	tmp &= ~EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN;
-	WREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp);
-
-	/* set pageflip to happen only at start of vblank interval (front porch) */
-	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
+	/* set pageflip to happen anywhere in vblank interval */
+	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1458,16 +1457,16 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
 	struct drm_framebuffer *target_fb;
 	uint64_t fb_location;
 	uint32_t fb_format, fb_pitch_pixels, tiling_flags;
 	u32 fb_swap = R600_D1GRPH_SWAP_ENDIAN_NONE;
-	u32 tmp, viewport_w, viewport_h;
+	u32 viewport_w, viewport_h;
 	int r;
 	bool bypass_lut = false;
+	struct drm_format_name_buf format_name;
 
 	/* no fb bound */
 	if (!atomic && !crtc->primary->fb) {
@@ -1475,16 +1474,12 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1505,7 +1500,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(rbo);
 
-	switch (target_fb->pixel_format) {
+	switch (target_fb->format->format) {
 	case DRM_FORMAT_C8:
 		fb_format =
 		    AVIVO_D1GRPH_CONTROL_DEPTH_8BPP |
@@ -1558,7 +1553,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		break;
 	default:
 		DRM_ERROR("Unsupported screen format %s\n",
-			  drm_get_format_name(target_fb->pixel_format));
+		          drm_get_format_name(target_fb->format->format, &format_name));
 		return -EINVAL;
 	}
 
@@ -1580,6 +1575,11 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	else
 		WREG32(AVIVO_D2VGA_CONTROL, 0);
 
+	/* Make sure surface address is update at vertical blank rather than
+	 * horizontal blank
+	 */
+	WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, 0);
+
 	if (rdev->family >= CHIP_RV770) {
 		if (radeon_crtc->crtc_id) {
 			WREG32(R700_D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH, upper_32_bits(fb_location));
@@ -1611,7 +1611,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(AVIVO_D1GRPH_X_END + radeon_crtc->crtc_offset, target_fb->width);
 	WREG32(AVIVO_D1GRPH_Y_END + radeon_crtc->crtc_offset, target_fb->height);
 
-	fb_pitch_pixels = target_fb->pitches[0] / (target_fb->bits_per_pixel / 8);
+	fb_pitch_pixels = target_fb->pitches[0] / target_fb->format->cpp[0];
 	WREG32(AVIVO_D1GRPH_PITCH + radeon_crtc->crtc_offset, fb_pitch_pixels);
 	WREG32(AVIVO_D1GRPH_ENABLE + radeon_crtc->crtc_offset, 1);
 
@@ -1626,18 +1626,11 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(AVIVO_D1MODE_VIEWPORT_SIZE + radeon_crtc->crtc_offset,
 	       (viewport_w << 16) | viewport_h);
 
-	/* pageflip setup */
-	/* make sure flip is at vb rather than hb */
-	tmp = RREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset);
-	tmp &= ~AVIVO_D1GRPH_SURFACE_UPDATE_H_RETRACE_EN;
-	WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp);
-
 	/* set pageflip to happen only at start of vblank interval (front porch) */
 	WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1666,11 +1659,11 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 }
 
 int atombios_crtc_set_base_atomic(struct drm_crtc *crtc,
-                                  struct drm_framebuffer *fb,
+				  struct drm_framebuffer *fb,
 				  int x, int y, enum mode_set_atomic state)
 {
-       struct drm_device *dev = crtc->dev;
-       struct radeon_device *rdev = dev->dev_private;
+	struct drm_device *dev = crtc->dev;
+	struct radeon_device *rdev = dev->dev_private;
 
 	if (ASIC_IS_DCE4(rdev))
 		return dce4_crtc_do_set_base(crtc, fb, x, y, 1);
@@ -2144,11 +2137,9 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 	atombios_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
@@ -2212,7 +2203,6 @@ static const struct drm_crtc_helper_funcs atombios_helper_funcs = {
 	.mode_set_base_atomic = atombios_crtc_set_base_atomic,
 	.prepare = atombios_crtc_prepare,
 	.commit = atombios_crtc_commit,
-	.load_lut = radeon_crtc_load_lut,
 	.disable = atombios_crtc_disable,
 };
 
diff --git a/sys/dev/pci/drm/radeon/atombios_dp.c b/sys/dev/pci/drm/radeon/atombios_dp.c
index b590be0b5a0..b874b2b1127 100644
--- a/sys/dev/pci/drm/radeon/atombios_dp.c
+++ b/sys/dev/pci/drm/radeon/atombios_dp.c
@@ -24,13 +24,13 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 #include "atom.h"
 #include "atom-bits.h"
-#include <dev/pci/drm/drm_dp_helper.h>
+#include <drm/drm_dp_helper.h>
 
 /* move these to drm_dp_helper.c/h */
 #define DP_LINK_CONFIGURATION_SIZE 9
@@ -38,10 +38,10 @@
 
 #ifdef DRMDEBUG
 static char *voltage_names[] = {
-        "0.4V", "0.6V", "0.8V", "1.2V"
+	"0.4V", "0.6V", "0.8V", "1.2V"
 };
 static char *pre_emph_names[] = {
-        "0dB", "3.5dB", "6dB", "9.5dB"
+	"0dB", "3.5dB", "6dB", "9.5dB"
 };
 #endif
 
@@ -304,10 +304,10 @@ static int convert_bpc_to_bpp(int bpc)
 
 /***** radeon specific DP functions *****/
 
-int radeon_dp_get_dp_link_config(struct drm_connector *connector,
-				 const u8 dpcd[DP_DPCD_SIZE],
-				 unsigned pix_clock,
-				 unsigned *dp_lanes, unsigned *dp_rate)
+static int radeon_dp_get_dp_link_config(struct drm_connector *connector,
+					const u8 dpcd[DP_DPCD_SIZE],
+					unsigned pix_clock,
+					unsigned *dp_lanes, unsigned *dp_rate)
 {
 	int bpp = convert_bpc_to_bpp(radeon_get_monitor_bpc(connector));
 	static const unsigned link_rates[3] = { 162000, 270000, 540000 };
@@ -389,22 +389,21 @@ bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector)
 {
 	struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
 	u8 msg[DP_DPCD_SIZE];
-	int ret, i;
+	int ret;
 
-	for (i = 0; i < 7; i++) {
-		ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg,
-				       DP_DPCD_SIZE);
-		if (ret == DP_DPCD_SIZE) {
-			memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
+	ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg,
+			       DP_DPCD_SIZE);
+	if (ret == DP_DPCD_SIZE) {
+		memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
 
-			DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
-				      dig_connector->dpcd);
+		DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
+			      dig_connector->dpcd);
 
-			radeon_dp_probe_oui(radeon_connector);
+		radeon_dp_probe_oui(radeon_connector);
 
-			return true;
-		}
+		return true;
 	}
+
 	dig_connector->dpcd[0] = 0;
 	return false;
 }
diff --git a/sys/dev/pci/drm/radeon/atombios_encoders.c b/sys/dev/pci/drm/radeon/atombios_encoders.c
index 65ca0364b3b..78a2b3b8be1 100644
--- a/sys/dev/pci/drm/radeon/atombios_encoders.c
+++ b/sys/dev/pci/drm/radeon/atombios_encoders.c
@@ -23,12 +23,15 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_audio.h"
+#include "radeon_asic.h"
 #include "atom.h"
+#include <linux/backlight.h>
+#include <linux/dmi.h>
 
 extern int atom_debug;
 
@@ -2646,16 +2649,8 @@ radeon_atom_ext_dpms(struct drm_encoder *encoder, int mode)
 
 }
 
-static bool radeon_atom_ext_mode_fixup(struct drm_encoder *encoder,
-				       const struct drm_display_mode *mode,
-				       struct drm_display_mode *adjusted_mode)
-{
-	return true;
-}
-
 static const struct drm_encoder_helper_funcs radeon_atom_ext_helper_funcs = {
 	.dpms = radeon_atom_ext_dpms,
-	.mode_fixup = radeon_atom_ext_mode_fixup,
 	.prepare = radeon_atom_ext_prepare,
 	.mode_set = radeon_atom_ext_mode_set,
 	.commit = radeon_atom_ext_commit,
@@ -2790,23 +2785,27 @@ radeon_add_atom_encoder(struct drm_device *dev,
 	case ENCODER_OBJECT_ID_INTERNAL_LVTM1:
 		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
 			radeon_encoder->rmx_type = RMX_FULL;
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_LVDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_LVDS, NULL);
 			radeon_encoder->enc_priv = radeon_atombios_get_lvds_info(radeon_encoder);
 		} else {
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TMDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_TMDS, NULL);
 			radeon_encoder->enc_priv = radeon_atombios_set_dig_info(radeon_encoder);
 		}
 		drm_encoder_helper_add(encoder, &radeon_atom_dig_helper_funcs);
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_DAC1:
-		drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC);
+		drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+				 DRM_MODE_ENCODER_DAC, NULL);
 		radeon_encoder->enc_priv = radeon_atombios_set_dac_info(radeon_encoder);
 		drm_encoder_helper_add(encoder, &radeon_atom_dac_helper_funcs);
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_DAC2:
 	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1:
 	case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2:
-		drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TVDAC);
+		drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+				 DRM_MODE_ENCODER_TVDAC, NULL);
 		radeon_encoder->enc_priv = radeon_atombios_set_dac_info(radeon_encoder);
 		drm_encoder_helper_add(encoder, &radeon_atom_dac_helper_funcs);
 		break;
@@ -2820,13 +2819,16 @@ radeon_add_atom_encoder(struct drm_device *dev,
 	case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
 		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
 			radeon_encoder->rmx_type = RMX_FULL;
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_LVDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_LVDS, NULL);
 			radeon_encoder->enc_priv = radeon_atombios_get_lvds_info(radeon_encoder);
 		} else if (radeon_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT)) {
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_DAC, NULL);
 			radeon_encoder->enc_priv = radeon_atombios_set_dig_info(radeon_encoder);
 		} else {
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TMDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_TMDS, NULL);
 			radeon_encoder->enc_priv = radeon_atombios_set_dig_info(radeon_encoder);
 		}
 		drm_encoder_helper_add(encoder, &radeon_atom_dig_helper_funcs);
@@ -2843,11 +2845,14 @@ radeon_add_atom_encoder(struct drm_device *dev,
 		/* these are handled by the primary encoders */
 		radeon_encoder->is_ext_encoder = true;
 		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_LVDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_LVDS, NULL);
 		else if (radeon_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT))
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_DAC, NULL);
 		else
-			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TMDS);
+			drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs,
+					 DRM_MODE_ENCODER_TMDS, NULL);
 		drm_encoder_helper_add(encoder, &radeon_atom_ext_helper_funcs);
 		break;
 	}
diff --git a/sys/dev/pci/drm/radeon/atombios_i2c.c b/sys/dev/pci/drm/radeon/atombios_i2c.c
index b817c037af6..ab8a3d24233 100644
--- a/sys/dev/pci/drm/radeon/atombios_i2c.c
+++ b/sys/dev/pci/drm/radeon/atombios_i2c.c
@@ -22,8 +22,8 @@
  * Authors: Alex Deucher
  *
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "atom.h"
 
diff --git a/sys/dev/pci/drm/radeon/btc_dpm.c b/sys/dev/pci/drm/radeon/btc_dpm.c
index 2ba7bce168e..0aef4937c90 100644
--- a/sys/dev/pci/drm/radeon/btc_dpm.c
+++ b/sys/dev/pci/drm/radeon/btc_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "btcd.h"
@@ -30,6 +30,7 @@
 #include "cypress_dpm.h"
 #include "btc_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -1162,12 +1163,11 @@ u32 btc_valid_sclk[40] =
 	155000, 160000, 165000, 170000, 175000, 180000, 185000, 190000, 195000, 200000
 };
 
-static const struct radeon_blacklist_clocks btc_blacklist_clocks[] =
-{
-        { 10000, 30000, RADEON_SCLK_UP },
-        { 15000, 30000, RADEON_SCLK_UP },
-        { 20000, 30000, RADEON_SCLK_UP },
-        { 25000, 30000, RADEON_SCLK_UP }
+static const struct radeon_blacklist_clocks btc_blacklist_clocks[] = {
+	{ 10000, 30000, RADEON_SCLK_UP },
+	{ 15000, 30000, RADEON_SCLK_UP },
+	{ 20000, 30000, RADEON_SCLK_UP },
+	{ 25000, 30000, RADEON_SCLK_UP }
 };
 
 void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table,
@@ -1636,14 +1636,14 @@ static int btc_init_smc_table(struct radeon_device *rdev,
 	cypress_populate_smc_voltage_tables(rdev, table);
 
 	switch (rdev->pm.int_thermal_type) {
-        case THERMAL_TYPE_EVERGREEN:
-        case THERMAL_TYPE_EMC2103_WITH_INTERNAL:
+	case THERMAL_TYPE_EVERGREEN:
+	case THERMAL_TYPE_EMC2103_WITH_INTERNAL:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_INTERNAL;
 		break;
-        case THERMAL_TYPE_NONE:
+	case THERMAL_TYPE_NONE:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_NONE;
 		break;
-        default:
+	default:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_EXTERNAL;
 		break;
 	}
@@ -1859,37 +1859,37 @@ static bool btc_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg)
 	case MC_SEQ_RAS_TIMING >> 2:
 		*out_reg = MC_SEQ_RAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_CAS_TIMING >> 2:
+	case MC_SEQ_CAS_TIMING >> 2:
 		*out_reg = MC_SEQ_CAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING >> 2:
+	case MC_SEQ_MISC_TIMING >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING2 >> 2:
+	case MC_SEQ_MISC_TIMING2 >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING2_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D0 >> 2:
+	case MC_SEQ_RD_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D1 >> 2:
+	case MC_SEQ_RD_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D1_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D0 >> 2:
+	case MC_SEQ_WR_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D1 >> 2:
+	case MC_SEQ_WR_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D1_LP >> 2;
 		break;
-        case MC_PMG_CMD_EMRS >> 2:
+	case MC_PMG_CMD_EMRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_EMRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS >> 2:
+	case MC_PMG_CMD_MRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS1 >> 2:
+	case MC_PMG_CMD_MRS1 >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS1_LP >> 2;
 		break;
-        default:
+	default:
 		result = false;
 		break;
 	}
@@ -2581,7 +2581,9 @@ int btc_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/sys/dev/pci/drm/radeon/cayman_blit_shaders.c b/sys/dev/pci/drm/radeon/cayman_blit_shaders.c
index b9ca2001bc5..9fec4d09f38 100644
--- a/sys/dev/pci/drm/radeon/cayman_blit_shaders.c
+++ b/sys/dev/pci/drm/radeon/cayman_blit_shaders.c
@@ -24,7 +24,9 @@
  *     Alex Deucher <alexander.deucher@amd.com>
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/bug.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
 
 /*
  * evergreen cards need to use the 3D engine to blit data which requires
diff --git a/sys/dev/pci/drm/radeon/ci_dpm.c b/sys/dev/pci/drm/radeon/ci_dpm.c
index 6702407d7f9..a97294ac96d 100644
--- a/sys/dev/pci/drm/radeon/ci_dpm.c
+++ b/sys/dev/pci/drm/radeon/ci_dpm.c
@@ -21,7 +21,8 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_ucode.h"
@@ -29,6 +30,7 @@
 #include "r600_dpm.h"
 #include "ci_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -182,6 +184,7 @@ static int ci_set_overdrive_target_tdp(struct radeon_device *rdev,
 				       u32 target_tdp);
 static int ci_update_uvd_dpm(struct radeon_device *rdev, bool gate);
 
+static PPSMC_Result ci_send_msg_to_smc(struct radeon_device *rdev, PPSMC_Msg msg);
 static PPSMC_Result ci_send_msg_to_smc_with_parameter(struct radeon_device *rdev,
 						      PPSMC_Msg msg, u32 parameter);
 
@@ -190,9 +193,9 @@ static void ci_fan_ctrl_set_default_mode(struct radeon_device *rdev);
 
 static struct ci_power_info *ci_get_pi(struct radeon_device *rdev)
 {
-        struct ci_power_info *pi = rdev->pm.dpm.priv;
+	struct ci_power_info *pi = rdev->pm.dpm.priv;
 
-        return pi;
+	return pi;
 }
 
 static struct ci_ps *ci_get_ps(struct radeon_ps *rps)
@@ -780,12 +783,6 @@ bool ci_dpm_vblank_too_short(struct radeon_device *rdev)
 	if (r600_dpm_get_vrefresh(rdev) > 120)
 		return true;
 
-	/* disable mclk switching if the refresh is >120Hz, even if the
-        * blanking period would allow it
-        */
-	if (r600_dpm_get_vrefresh(rdev) > 120)
-		return true;
-
 	if (vblank_time < switch_limit)
 		return true;
 	else
@@ -1642,7 +1639,7 @@ static int ci_notify_hw_of_power_source(struct radeon_device *rdev,
 	else
 		power_limit = (u32)(cac_tdp_table->battery_power_limit * 256);
 
-        ci_set_power_limit(rdev, power_limit);
+	ci_set_power_limit(rdev, power_limit);
 
 	if (pi->caps_automatic_dc_transition) {
 		if (ac_power)
@@ -1655,6 +1652,27 @@ static int ci_notify_hw_of_power_source(struct radeon_device *rdev,
 }
 #endif
 
+static PPSMC_Result ci_send_msg_to_smc(struct radeon_device *rdev, PPSMC_Msg msg)
+{
+	u32 tmp;
+	int i;
+
+	if (!ci_is_smc_running(rdev))
+		return PPSMC_Result_Failed;
+
+	WREG32(SMC_MESSAGE_0, msg);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		tmp = RREG32(SMC_RESP_0);
+		if (tmp != 0)
+			break;
+		udelay(1);
+	}
+	tmp = RREG32(SMC_RESP_0);
+
+	return (PPSMC_Result)tmp;
+}
+
 static PPSMC_Result ci_send_msg_to_smc_with_parameter(struct radeon_device *rdev,
 						      PPSMC_Msg msg, u32 parameter)
 {
@@ -2027,9 +2045,9 @@ static void ci_enable_display_gap(struct radeon_device *rdev)
 {
 	u32 tmp = RREG32_SMC(CG_DISPLAY_GAP_CNTL);
 
-        tmp &= ~(DISP_GAP_MASK | DISP_GAP_MCHG_MASK);
-        tmp |= (DISP_GAP(R600_PM_DISPLAY_GAP_IGNORE) |
-                DISP_GAP_MCHG(R600_PM_DISPLAY_GAP_VBLANK));
+	tmp &= ~(DISP_GAP_MASK | DISP_GAP_MCHG_MASK);
+	tmp |= (DISP_GAP(R600_PM_DISPLAY_GAP_IGNORE) |
+		DISP_GAP_MCHG(R600_PM_DISPLAY_GAP_VBLANK));
 
 	WREG32_SMC(CG_DISPLAY_GAP_CNTL, tmp);
 }
@@ -2948,8 +2966,8 @@ static int ci_populate_single_memory_level(struct radeon_device *rdev,
 
 	memory_level->MinVddc = cpu_to_be32(memory_level->MinVddc * VOLTAGE_SCALE);
 	memory_level->MinVddcPhases = cpu_to_be32(memory_level->MinVddcPhases);
-        memory_level->MinVddci = cpu_to_be32(memory_level->MinVddci * VOLTAGE_SCALE);
-        memory_level->MinMvdd = cpu_to_be32(memory_level->MinMvdd * VOLTAGE_SCALE);
+	memory_level->MinVddci = cpu_to_be32(memory_level->MinVddci * VOLTAGE_SCALE);
+	memory_level->MinMvdd = cpu_to_be32(memory_level->MinMvdd * VOLTAGE_SCALE);
 
 	memory_level->MclkFrequency = cpu_to_be32(memory_level->MclkFrequency);
 	memory_level->ActivityLevel = cpu_to_be16(memory_level->ActivityLevel);
@@ -3162,7 +3180,7 @@ static int ci_calculate_sclk_params(struct radeon_device *rdev,
 
 	spll_func_cntl_3 &= ~SPLL_FB_DIV_MASK;
 	spll_func_cntl_3 |= SPLL_FB_DIV(fbdiv);
-        spll_func_cntl_3 |= SPLL_DITHEN;
+	spll_func_cntl_3 |= SPLL_DITHEN;
 
 	if (pi->caps_sclk_ss_support) {
 		struct radeon_atom_ss ss;
@@ -3239,7 +3257,7 @@ static int ci_populate_single_graphic_level(struct radeon_device *rdev,
 	graphic_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
 	graphic_level->Flags = cpu_to_be32(graphic_level->Flags);
-        graphic_level->MinVddc = cpu_to_be32(graphic_level->MinVddc * VOLTAGE_SCALE);
+	graphic_level->MinVddc = cpu_to_be32(graphic_level->MinVddc * VOLTAGE_SCALE);
 	graphic_level->MinVddcPhases = cpu_to_be32(graphic_level->MinVddcPhases);
 	graphic_level->SclkFrequency = cpu_to_be32(graphic_level->SclkFrequency);
 	graphic_level->ActivityLevel = cpu_to_be16(graphic_level->ActivityLevel);
@@ -3853,7 +3871,10 @@ static void ci_find_dpm_states_clocks_in_dpm_table(struct radeon_device *rdev,
 	if (i >= sclk_table->count) {
 		pi->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
 	} else {
-		/* XXX check display min clock requirements */
+		/* XXX The current code always reprogrammed the sclk levels,
+		 * but we don't currently handle disp sclk requirements
+		 * so just skip it.
+		 */
 		if (CISLAND_MINIMUM_ENGINE_CLOCK != CISLAND_MINIMUM_ENGINE_CLOCK)
 			pi->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_SCLK;
 	}
@@ -4403,7 +4424,7 @@ static bool ci_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg)
 		break;
 	case MC_SEQ_CAS_TIMING >> 2:
 		*out_reg = MC_SEQ_CAS_TIMING_LP >> 2;
-            break;
+		break;
 	case MC_SEQ_MISC_TIMING >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING_LP >> 2;
 		break;
@@ -4635,7 +4656,7 @@ static int ci_initialize_mc_reg_table(struct radeon_device *rdev)
 	if (ret)
 		goto init_mc_done;
 
-        ret = ci_copy_vbios_mc_reg_table(table, ci_table);
+	ret = ci_copy_vbios_mc_reg_table(table, ci_table);
 	if (ret)
 		goto init_mc_done;
 
@@ -4926,7 +4947,7 @@ static int ci_set_private_data_variables_based_on_pptable(struct radeon_device *
 		allowed_mclk_vddc_table->entries[allowed_sclk_vddc_table->count - 1].clk;
 	rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc =
 		allowed_sclk_vddc_table->entries[allowed_sclk_vddc_table->count - 1].v;
-        rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci =
+	rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci =
 		allowed_mclk_vddci_table->entries[allowed_mclk_vddci_table->count - 1].v;
 
 	return 0;
@@ -5527,7 +5548,7 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 	struct _NonClockInfoArray *non_clock_info_array;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 	struct ci_ps *ps;
@@ -5547,8 +5568,9 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -5654,19 +5676,30 @@ int ci_dpm_init(struct radeon_device *rdev)
 	u16 data_offset, size;
 	u8 frev, crev;
 	struct ci_power_info *pi;
+	enum pci_bus_speed speed_cap = PCI_SPEED_UNKNOWN;
+	struct pci_dev *root = rdev->pdev->bus->self;
 	int ret;
-	u32 mask;
 
 	pi = kzalloc(sizeof(struct ci_power_info), GFP_KERNEL);
 	if (pi == NULL)
 		return -ENOMEM;
 	rdev->pm.dpm.priv = pi;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret)
+	if (!pci_is_root_bus(rdev->pdev->bus))
+		speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN) {
 		pi->sys_pcie_mask = 0;
-	else
-		pi->sys_pcie_mask = mask;
+	} else {
+		if (speed_cap == PCIE_SPEED_8_0GT)
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50 |
+				RADEON_PCIE_SPEED_80;
+		else if (speed_cap == PCIE_SPEED_5_0GT)
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50;
+		else
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25;
+	}
 	pi->force_pcie_gen = RADEON_PCIE_GEN_INVALID;
 
 	pi->pcie_gen_performance.max = RADEON_PCIE_GEN1;
@@ -5703,8 +5736,8 @@ int ci_dpm_init(struct radeon_device *rdev)
 		return ret;
 	}
 
-        pi->dll_default_on = false;
-        pi->sram_end = SMC_RAM_END;
+	pi->dll_default_on = false;
+	pi->sram_end = SMC_RAM_END;
 
 	pi->activity_target[0] = CISLAND_TARGETACTIVITY_DFLT;
 	pi->activity_target[1] = CISLAND_TARGETACTIVITY_DFLT;
@@ -5744,12 +5777,14 @@ int ci_dpm_init(struct radeon_device *rdev)
 	pi->caps_uvd_dpm = true;
 	pi->caps_vce_dpm = true;
 
-        ci_get_leakage_voltages(rdev);
-        ci_patch_dependency_tables_with_leakage(rdev);
-        ci_set_private_data_variables_based_on_pptable(rdev);
+	ci_get_leakage_voltages(rdev);
+	ci_patch_dependency_tables_with_leakage(rdev);
+	ci_set_private_data_variables_based_on_pptable(rdev);
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		ci_dpm_fini(rdev);
 		return -ENOMEM;
@@ -5849,7 +5884,7 @@ int ci_dpm_init(struct radeon_device *rdev)
 			pi->vddci_control = CISLANDS_VOLTAGE_CONTROL_BY_SVID2;
 		else
 			rdev->pm.dpm.platform_caps &= ~ATOM_PP_PLATFORM_CAP_VDDCI_CONTROL;
-        }
+	}
 
 	if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_MVDDCONTROL) {
 		if (radeon_atom_is_voltage_gpio(rdev, VOLTAGE_TYPE_MVDDC, VOLTAGE_OBJ_GPIO_LUT))
@@ -5870,7 +5905,7 @@ int ci_dpm_init(struct radeon_device *rdev)
 #endif
 
 	if (atom_parse_data_header(rdev->mode_info.atom_context, index, &size,
-                                   &frev, &crev, &data_offset)) {
+				   &frev, &crev, &data_offset)) {
 		pi->caps_sclk_ss_support = true;
 		pi->caps_mclk_ss_support = true;
 		pi->dynamic_ss = true;
diff --git a/sys/dev/pci/drm/radeon/ci_dpm.h b/sys/dev/pci/drm/radeon/ci_dpm.h
index 723220ffbea..dff2a63df38 100644
--- a/sys/dev/pci/drm/radeon/ci_dpm.h
+++ b/sys/dev/pci/drm/radeon/ci_dpm.h
@@ -330,7 +330,6 @@ int ci_program_jump_on_start(struct radeon_device *rdev);
 void ci_stop_smc_clock(struct radeon_device *rdev);
 void ci_start_smc_clock(struct radeon_device *rdev);
 bool ci_is_smc_running(struct radeon_device *rdev);
-PPSMC_Result ci_send_msg_to_smc(struct radeon_device *rdev, PPSMC_Msg msg);
 PPSMC_Result ci_wait_for_smc_inactive(struct radeon_device *rdev);
 int ci_load_smc_ucode(struct radeon_device *rdev, u32 limit);
 int ci_read_smc_sram_dword(struct radeon_device *rdev,
diff --git a/sys/dev/pci/drm/radeon/ci_smc.c b/sys/dev/pci/drm/radeon/ci_smc.c
index 32fe3d07e34..37112191375 100644
--- a/sys/dev/pci/drm/radeon/ci_smc.c
+++ b/sys/dev/pci/drm/radeon/ci_smc.c
@@ -22,7 +22,8 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "ppsmc.h"
@@ -162,27 +163,6 @@ bool ci_is_smc_running(struct radeon_device *rdev)
 	return false;
 }
 
-PPSMC_Result ci_send_msg_to_smc(struct radeon_device *rdev, PPSMC_Msg msg)
-{
-	u32 tmp;
-	int i;
-
-	if (!ci_is_smc_running(rdev))
-		return PPSMC_Result_Failed;
-
-	WREG32(SMC_MESSAGE_0, msg);
-
-	for (i = 0; i < rdev->usec_timeout; i++) {
-		tmp = RREG32(SMC_RESP_0);
-		if (tmp != 0)
-			break;
-		udelay(1);
-	}
-	tmp = RREG32(SMC_RESP_0);
-
-	return (PPSMC_Result)tmp;
-}
-
 #if 0
 PPSMC_Result ci_wait_for_smc_inactive(struct radeon_device *rdev)
 {
@@ -193,11 +173,11 @@ PPSMC_Result ci_wait_for_smc_inactive(struct radeon_device *rdev)
 		return PPSMC_Result_OK;
 
 	for (i = 0; i < rdev->usec_timeout; i++) {
-                tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0);
-                if ((tmp & CKEN) == 0)
+		tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0);
+		if ((tmp & CKEN) == 0)
 			break;
-                udelay(1);
-        }
+		udelay(1);
+	}
 
 	return PPSMC_Result_OK;
 }
diff --git a/sys/dev/pci/drm/radeon/cik.c b/sys/dev/pci/drm/radeon/cik.c
index bd5983471dc..b6c91ff4906 100644
--- a/sys/dev/pci/drm/radeon/cik.c
+++ b/sys/dev/pci/drm/radeon/cik.c
@@ -21,7 +21,10 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -30,7 +33,9 @@
 #include "cik_blit_shaders.h"
 #include "radeon_ucode.h"
 #include "clearstate_ci.h"
-#include "radeon_kfd.h"
+
+#define SH_MEM_CONFIG_GFX_DEFAULT \
+	ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED)
 
 MODULE_FIRMWARE("radeon/BONAIRE_pfp.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_me.bin");
@@ -50,6 +55,7 @@ MODULE_FIRMWARE("radeon/bonaire_mc.bin");
 MODULE_FIRMWARE("radeon/bonaire_rlc.bin");
 MODULE_FIRMWARE("radeon/bonaire_sdma.bin");
 MODULE_FIRMWARE("radeon/bonaire_smc.bin");
+MODULE_FIRMWARE("radeon/bonaire_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/HAWAII_pfp.bin");
 MODULE_FIRMWARE("radeon/HAWAII_me.bin");
@@ -69,6 +75,7 @@ MODULE_FIRMWARE("radeon/hawaii_mc.bin");
 MODULE_FIRMWARE("radeon/hawaii_rlc.bin");
 MODULE_FIRMWARE("radeon/hawaii_sdma.bin");
 MODULE_FIRMWARE("radeon/hawaii_smc.bin");
+MODULE_FIRMWARE("radeon/hawaii_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/KAVERI_pfp.bin");
 MODULE_FIRMWARE("radeon/KAVERI_me.bin");
@@ -1620,8 +1627,6 @@ static const u32 godavari_golden_registers[] =
 
 static void cik_init_golden_registers(struct radeon_device *rdev)
 {
-	/* Some of the registers might be dependent on GRBM_GFX_INDEX */
-	mutex_lock(&rdev->grbm_idx_mutex);
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
 		radeon_program_register_sequence(rdev,
@@ -1696,7 +1701,6 @@ static void cik_init_golden_registers(struct radeon_device *rdev)
 	default:
 		break;
 	}
-	mutex_unlock(&rdev->grbm_idx_mutex);
 }
 
 /**
@@ -1709,7 +1713,7 @@ static void cik_init_golden_registers(struct radeon_device *rdev)
  */
 u32 cik_get_xclk(struct radeon_device *rdev)
 {
-        u32 reference_clock = rdev->clock.spll.reference_freq;
+	u32 reference_clock = rdev->clock.spll.reference_freq;
 
 	if (rdev->flags & RADEON_IS_IGP) {
 		if (RREG32_SMC(GENERAL_PWRMGT) & GPU_COUNTER_CLK)
@@ -1866,7 +1870,7 @@ int ci_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data = NULL;
 	const __le32 *new_fw_data = NULL;
-	u32 running, blackout = 0, tmp;
+	u32 running, tmp;
 	u32 *io_mc_regs = NULL;
 	const __le32 *new_io_mc_regs = NULL;
 	int i, regs_size, ucode_size;
@@ -1907,11 +1911,6 @@ int ci_mc_load_microcode(struct radeon_device *rdev)
 	running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK;
 
 	if (running == 0) {
-		if (running) {
-			blackout = RREG32(MC_SHARED_BLACKOUT_CNTL);
-			WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1);
-		}
-
 		/* reset the engine and set to writable */
 		WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
 		WREG32(MC_SEQ_SUP_CNTL, 0x00000010);
@@ -1959,9 +1958,6 @@ int ci_mc_load_microcode(struct radeon_device *rdev)
 				break;
 			udelay(1);
 		}
-
-		if (running)
-			WREG32(MC_SHARED_BLACKOUT_CNTL, blackout);
 	}
 
 	return 0;
@@ -1987,12 +1983,17 @@ static int cik_init_microcode(struct radeon_device *rdev)
 	int new_fw = 0;
 	int err;
 	int num_fw;
+	bool new_smc = false;
 
 	DRM_DEBUG("\n");
 
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
 		chip_name = "BONAIRE";
+		if ((rdev->pdev->revision == 0x80) ||
+		    (rdev->pdev->revision == 0x81) ||
+		    (rdev->pdev->device == 0x665f))
+			new_smc = true;
 		new_chip_name = "bonaire";
 		pfp_req_size = CIK_PFP_UCODE_SIZE * 4;
 		me_req_size = CIK_ME_UCODE_SIZE * 4;
@@ -2007,6 +2008,8 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_HAWAII:
 		chip_name = "HAWAII";
+		if (rdev->pdev->revision == 0x80)
+			new_smc = true;
 		new_chip_name = "hawaii";
 		pfp_req_size = CIK_PFP_UCODE_SIZE * 4;
 		me_req_size = CIK_ME_UCODE_SIZE * 4;
@@ -2065,8 +2068,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->pfp_fw->size != pfp_req_size) {
-			printk(KERN_ERR
-			       "cik_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->pfp_fw->size, fw_name);
 			err = -EINVAL;
 			goto out;
@@ -2074,8 +2076,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 	} else {
 		err = radeon_ucode_validate(rdev->pfp_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2091,16 +2092,14 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->me_fw->size != me_req_size) {
-			printk(KERN_ERR
-			       "cik_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->me_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->me_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2116,16 +2115,14 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->ce_fw->size != ce_req_size) {
-			printk(KERN_ERR
-			       "cik_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->ce_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->ce_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2141,16 +2138,14 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->mec_fw->size != mec_req_size) {
-			printk(KERN_ERR
-			       "cik_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->mec_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->mec_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2181,16 +2176,14 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->rlc_fw->size != rlc_req_size) {
-			printk(KERN_ERR
-			       "cik_rlc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_rlc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->rlc_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->rlc_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2206,16 +2199,14 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->sdma_fw->size != sdma_req_size) {
-			printk(KERN_ERR
-			       "cik_sdma: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("cik_sdma: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->sdma_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->sdma_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "cik_fw: validation failed for firmware \"%s\"\n",
+			pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -2238,8 +2229,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 			}
 			if ((rdev->mc_fw->size != mc_req_size) &&
 			    (rdev->mc_fw->size != mc2_req_size)){
-				printk(KERN_ERR
-				       "cik_mc: Bogus length %zu in firmware \"%s\"\n",
+				pr_err("cik_mc: Bogus length %zu in firmware \"%s\"\n",
 				       rdev->mc_fw->size, fw_name);
 				err = -EINVAL;
 			}
@@ -2247,8 +2237,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		} else {
 			err = radeon_ucode_validate(rdev->mc_fw);
 			if (err) {
-				printk(KERN_ERR
-				       "cik_fw: validation failed for firmware \"%s\"\n",
+				pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 				       fw_name);
 				goto out;
 			} else {
@@ -2256,29 +2245,29 @@ static int cik_init_microcode(struct radeon_device *rdev)
 			}
 		}
 
-		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name);
+		if (new_smc)
+			snprintf(fw_name, sizeof(fw_name), "radeon/%s_k_smc.bin", new_chip_name);
+		else
+			snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name);
 		err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 		if (err) {
 			snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
 			err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 			if (err) {
-				printk(KERN_ERR
-				       "smc: error loading firmware \"%s\"\n",
+				pr_err("smc: error loading firmware \"%s\"\n",
 				       fw_name);
 				release_firmware(rdev->smc_fw);
 				rdev->smc_fw = NULL;
 				err = 0;
 			} else if (rdev->smc_fw->size != smc_req_size) {
-				printk(KERN_ERR
-				       "cik_smc: Bogus length %zu in firmware \"%s\"\n",
+				pr_err("cik_smc: Bogus length %zu in firmware \"%s\"\n",
 				       rdev->smc_fw->size, fw_name);
 				err = -EINVAL;
 			}
 		} else {
 			err = radeon_ucode_validate(rdev->smc_fw);
 			if (err) {
-				printk(KERN_ERR
-				       "cik_fw: validation failed for firmware \"%s\"\n",
+				pr_err("cik_fw: validation failed for firmware \"%s\"\n",
 				       fw_name);
 				goto out;
 			} else {
@@ -2290,7 +2279,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 	if (new_fw == 0) {
 		rdev->new_fw = false;
 	} else if (new_fw < num_fw) {
-		printk(KERN_ERR "ci_fw: mixing new and old firmware!\n");
+		pr_err("ci_fw: mixing new and old firmware!\n");
 		err = -EINVAL;
 	} else {
 		rdev->new_fw = true;
@@ -2299,8 +2288,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 out:
 	if (err) {
 		if (err != -EINVAL)
-			printk(KERN_ERR
-			       "cik_cp: Failed to load firmware \"%s\"\n",
+			pr_err("cik_cp: Failed to load firmware \"%s\"\n",
 			       fw_name);
 		release_firmware(rdev->pfp_fw);
 		rdev->pfp_fw = NULL;
@@ -2340,9 +2328,13 @@ out:
  */
 static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 {
-	const u32 num_tile_mode_states = 32;
-	const u32 num_secondary_tile_mode_states = 16;
-	u32 reg_offset, gb_tile_moden, split_equal_to_row_size;
+	u32 *tile = rdev->config.cik.tile_mode_array;
+	u32 *macrotile = rdev->config.cik.macrotile_mode_array;
+	const u32 num_tile_mode_states =
+			ARRAY_SIZE(rdev->config.cik.tile_mode_array);
+	const u32 num_secondary_tile_mode_states =
+			ARRAY_SIZE(rdev->config.cik.macrotile_mode_array);
+	u32 reg_offset, split_equal_to_row_size;
 	u32 num_pipe_configs;
 	u32 num_rbs = rdev->config.cik.max_backends_per_se *
 		rdev->config.cik.max_shader_engines;
@@ -2364,1032 +2356,669 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 	if (num_pipe_configs > 8)
 		num_pipe_configs = 16;
 
-	if (num_pipe_configs == 16) {
-		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
-				break;
-			case 1:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
-				break;
-			case 2:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 3:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
-				break;
-			case 4:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 5:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
-				break;
-			case 6:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 7:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 8:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16));
-				break;
-			case 9:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
-				break;
-			case 10:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 11:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 12:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 13:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
-				break;
-			case 14:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 16:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 17:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 27:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
-				break;
-			case 28:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 29:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 30:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.tile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 1:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 2:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 3:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 4:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 5:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			case 6:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_2_BANK));
-				break;
-			case 8:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 9:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 10:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 11:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 12:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			case 13:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_2_BANK));
-				break;
-			case 14:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_2_BANK));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-	} else if (num_pipe_configs == 8) {
-		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
-				break;
-			case 1:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
-				break;
-			case 2:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 3:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
-				break;
-			case 4:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 5:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
-				break;
-			case 6:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 7:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 8:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16));
-				break;
-			case 9:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
-				break;
-			case 10:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 11:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 12:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 13:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
-				break;
-			case 14:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 16:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 17:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 27:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
-				break;
-			case 28:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 29:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 30:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.tile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 1:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 2:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 3:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 4:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 5:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			case 6:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_2_BANK));
-				break;
-			case 8:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 9:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 10:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 11:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 12:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 13:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			case 14:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_2_BANK));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-	} else if (num_pipe_configs == 4) {
+	for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+		tile[reg_offset] = 0;
+	for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++)
+		macrotile[reg_offset] = 0;
+
+	switch(num_pipe_configs) {
+	case 16:
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
+		tile[4] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[5] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+		tile[6] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[7] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16));
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[11] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[12] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[16] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[17] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[27] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
+		tile[28] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[29] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[30] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+
+		macrotile[0] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[1] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[2] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[3] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[4] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			   NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[5] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			   NUM_BANKS(ADDR_SURF_4_BANK));
+		macrotile[6] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			   NUM_BANKS(ADDR_SURF_2_BANK));
+		macrotile[8] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[9] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+			   NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[10] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			    BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			    MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			    NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[11] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			    BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			    MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			    NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[12] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			    BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			    MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			    NUM_BANKS(ADDR_SURF_4_BANK));
+		macrotile[13] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			    BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			    MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			    NUM_BANKS(ADDR_SURF_2_BANK));
+		macrotile[14] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			    BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			    MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+			    NUM_BANKS(ADDR_SURF_2_BANK));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++)
+			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), macrotile[reg_offset]);
+		break;
+
+	case 8:
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
+		tile[4] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[5] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+		tile[6] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[7] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16));
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[11] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[12] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[16] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[17] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[27] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
+		tile[28] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[29] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[30] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+
+		macrotile[0] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[1] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[2] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[3] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[4] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[5] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_4_BANK));
+		macrotile[6] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_2_BANK));
+		macrotile[8] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[9] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[10] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[11] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[12] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[13] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_4_BANK));
+		macrotile[14] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_2_BANK));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++)
+			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), macrotile[reg_offset]);
+		break;
+
+	case 4:
 		if (num_rbs == 4) {
-			for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-				switch (reg_offset) {
-				case 0:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
-					break;
-				case 1:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
-					break;
-				case 2:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-					break;
-				case 3:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
-					break;
-				case 4:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(split_equal_to_row_size));
-					break;
-				case 5:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
-					break;
-				case 6:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-					break;
-				case 7:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 TILE_SPLIT(split_equal_to_row_size));
-					break;
-				case 8:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16));
-					break;
-				case 9:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
-					break;
-				case 10:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 11:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 12:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 13:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
-					break;
-				case 14:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 16:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 17:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 27:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
-					break;
-				case 28:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 29:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 30:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				default:
-					gb_tile_moden = 0;
-					break;
-				}
-				rdev->config.cik.tile_mode_array[reg_offset] = gb_tile_moden;
-				WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-			}
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
+		tile[4] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[5] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+		tile[6] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[7] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16));
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[11] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[12] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[16] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[17] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[27] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
+		tile[28] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[29] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[30] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_16x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+
 		} else if (num_rbs < 4) {
-			for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-				switch (reg_offset) {
-				case 0:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
-					break;
-				case 1:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
-					break;
-				case 2:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-					break;
-				case 3:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
-					break;
-				case 4:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(split_equal_to_row_size));
-					break;
-				case 5:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
-					break;
-				case 6:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-					break;
-				case 7:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 TILE_SPLIT(split_equal_to_row_size));
-					break;
-				case 8:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16));
-					break;
-				case 9:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
-					break;
-				case 10:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 11:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 12:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 13:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
-					break;
-				case 14:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 16:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 17:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 27:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
-					break;
-				case 28:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 29:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				case 30:
-					gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-							 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-					break;
-				default:
-					gb_tile_moden = 0;
-					break;
-				}
-				rdev->config.cik.tile_mode_array[reg_offset] = gb_tile_moden;
-				WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-			}
-		}
-		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 1:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 2:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 3:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 4:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 5:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 6:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			case 8:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 9:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 10:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 11:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 12:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 13:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 14:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
-						 NUM_BANKS(ADDR_SURF_4_BANK));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-	} else if (num_pipe_configs == 2) {
-		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
-				break;
-			case 1:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
-				break;
-			case 2:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 3:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
-				break;
-			case 4:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 5:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
-				break;
-			case 6:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
-				break;
-			case 7:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 TILE_SPLIT(split_equal_to_row_size));
-				break;
-			case 8:
-				gb_tile_moden = ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						PIPE_CONFIG(ADDR_SURF_P2);
-				break;
-			case 9:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2));
-				break;
-			case 10:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 11:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 12:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 13:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
-				break;
-			case 14:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 16:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 17:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 27:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2));
-				break;
-			case 28:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 29:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			case 30:
-				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P2) |
-						 SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.tile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 1:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 2:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 3:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 4:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 5:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 6:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			case 8:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 9:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 10:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 11:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 12:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 13:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
-						 NUM_BANKS(ADDR_SURF_16_BANK));
-				break;
-			case 14:
-				gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
-						 NUM_BANKS(ADDR_SURF_8_BANK));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden);
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
+		tile[4] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[5] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+		tile[6] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[7] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16));
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[11] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[12] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[16] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[17] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[27] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
+		tile[28] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[29] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[30] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
 		}
-	} else
+
+		macrotile[0] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[1] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[2] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[3] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[4] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[5] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[6] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_4_BANK));
+		macrotile[8] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[9] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[10] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[11] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[12] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[13] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[14] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
+				NUM_BANKS(ADDR_SURF_4_BANK));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++)
+			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), macrotile[reg_offset]);
+		break;
+
+	case 2:
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B));
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B));
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B));
+		tile[4] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[5] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
+		tile[6] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B));
+		tile[7] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2) |
+			   TILE_SPLIT(split_equal_to_row_size));
+		tile[8] = ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   PIPE_CONFIG(ADDR_SURF_P2);
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P2));
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[11] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[12] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[16] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[17] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[27] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2));
+		tile[28] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[29] = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+		tile[30] = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
+			    MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+			    PIPE_CONFIG(ADDR_SURF_P2) |
+			    SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
+
+		macrotile[0] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[1] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[2] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[3] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[4] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[5] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[6] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+		macrotile[8] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[9] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_4) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[10] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[11] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[12] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[13] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4) |
+				NUM_BANKS(ADDR_SURF_16_BANK));
+		macrotile[14] = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+				BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+				MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
+				NUM_BANKS(ADDR_SURF_8_BANK));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++)
+			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), macrotile[reg_offset]);
+		break;
+
+	default:
 		DRM_ERROR("unknown num pipe config: 0x%x\n", num_pipe_configs);
+	}
 }
 
 /**
@@ -3488,7 +3117,6 @@ static void cik_setup_rb(struct radeon_device *rdev,
 	u32 disabled_rbs = 0;
 	u32 enabled_rbs = 0;
 
-	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < se_num; i++) {
 		for (j = 0; j < sh_per_se; j++) {
 			cik_select_se_sh(rdev, i, j);
@@ -3500,7 +3128,6 @@ static void cik_setup_rb(struct radeon_device *rdev,
 		}
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	mask = 1;
 	for (i = 0; i < max_rb_num_per_se * se_num; i++) {
@@ -3511,7 +3138,6 @@ static void cik_setup_rb(struct radeon_device *rdev,
 
 	rdev->config.cik.backend_enable_mask = enabled_rbs;
 
-	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < se_num; i++) {
 		cik_select_se_sh(rdev, i, 0xffffffff);
 		data = 0;
@@ -3539,7 +3165,6 @@ static void cik_setup_rb(struct radeon_device *rdev,
 		WREG32(PA_SC_RASTER_CONFIG, data);
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 }
 
 /**
@@ -3732,12 +3357,6 @@ static void cik_gpu_init(struct radeon_device *rdev)
 	/* set HW defaults for 3D engine */
 	WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60));
 
-	mutex_lock(&rdev->grbm_idx_mutex);
-	/*
-	 * making sure that the following register writes will be broadcasted
-	 * to all the shaders
-	 */
-	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 	WREG32(SX_DEBUG_1, 0x20);
 
 	WREG32(TA_CNTL_AUX, 0x00010000);
@@ -3793,7 +3412,6 @@ static void cik_gpu_init(struct radeon_device *rdev)
 
 	WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3));
 	WREG32(PA_SC_ENHANCE, ENABLE_PA_SC_OUT_OF_ORDER);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	udelay(50);
 }
@@ -4102,10 +3720,10 @@ struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev,
  * @rdev: radeon_device pointer
  * @ib: radeon indirect buffer object
  *
- * Emits an DE (drawing engine) or CE (constant engine) IB
+ * Emits a DE (drawing engine) or CE (constant engine) IB
  * on the gfx ring.  IBs are usually generated by userspace
  * acceleration drivers and submitted to the kernel for
- * sheduling on the ring.  This function schedules the IB
+ * scheduling on the ring.  This function schedules the IB
  * on the gfx ring for execution by the GPU.
  */
 void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
@@ -4189,13 +3807,20 @@ int cik_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 		return r;
 	}
-	r = radeon_fence_wait(ib.fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(ib.fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		radeon_scratch_free(rdev, scratch);
 		radeon_ib_free(rdev, &ib);
 		return r;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		radeon_scratch_free(rdev, scratch);
+		radeon_ib_free(rdev, &ib);
+		return -ETIMEDOUT;
 	}
+	r = 0;
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = RREG32(scratch);
 		if (tmp == 0xDEADBEEF)
@@ -4511,11 +4136,7 @@ u32 cik_gfx_get_rptr(struct radeon_device *rdev,
 u32 cik_gfx_get_wptr(struct radeon_device *rdev,
 		     struct radeon_ring *ring)
 {
-	u32 wptr;
-
-	wptr = RREG32(CP_RB0_WPTR);
-
-	return wptr;
+	return RREG32(CP_RB0_WPTR);
 }
 
 void cik_gfx_set_wptr(struct radeon_device *rdev,
@@ -4770,11 +4391,12 @@ static int cik_mec_init(struct radeon_device *rdev)
 	/*
 	 * KV:    2 MEC, 4 Pipes/MEC, 8 Queues/Pipe - 64 Queues total
 	 * CI/KB: 1 MEC, 4 Pipes/MEC, 8 Queues/Pipe - 32 Queues total
-	 * Nonetheless, we assign only 1 pipe because all other pipes will
-	 * be handled by KFD
 	 */
-	rdev->mec.num_mec = 1;
-	rdev->mec.num_pipe = 1;
+	if (rdev->family == CHIP_KAVERI)
+		rdev->mec.num_mec = 2;
+	else
+		rdev->mec.num_mec = 1;
+	rdev->mec.num_pipe = 4;
 	rdev->mec.num_queue = rdev->mec.num_mec * rdev->mec.num_pipe * 8;
 
 	if (rdev->mec.hpd_eop_obj == NULL) {
@@ -4917,23 +4539,28 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 	/* init the pipes */
 	mutex_lock(&rdev->srbm_mutex);
 
-	eop_gpu_addr = rdev->mec.hpd_eop_gpu_addr;
+	for (i = 0; i < (rdev->mec.num_pipe * rdev->mec.num_mec); ++i) {
+		int me = (i < 4) ? 1 : 2;
+		int pipe = (i < 4) ? i : (i - 4);
 
-	cik_srbm_select(rdev, 0, 0, 0, 0);
+		cik_srbm_select(rdev, me, pipe, 0, 0);
 
-	/* write the EOP addr */
-	WREG32(CP_HPD_EOP_BASE_ADDR, eop_gpu_addr >> 8);
-	WREG32(CP_HPD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr) >> 8);
+		eop_gpu_addr = rdev->mec.hpd_eop_gpu_addr + (i * MEC_HPD_SIZE * 2) ;
+		/* write the EOP addr */
+		WREG32(CP_HPD_EOP_BASE_ADDR, eop_gpu_addr >> 8);
+		WREG32(CP_HPD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr) >> 8);
 
-	/* set the VMID assigned */
-	WREG32(CP_HPD_EOP_VMID, 0);
+		/* set the VMID assigned */
+		WREG32(CP_HPD_EOP_VMID, 0);
 
-	/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
-	tmp = RREG32(CP_HPD_EOP_CONTROL);
-	tmp &= ~EOP_SIZE_MASK;
-	tmp |= order_base_2(MEC_HPD_SIZE / 8);
-	WREG32(CP_HPD_EOP_CONTROL, tmp);
+		/* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */
+		tmp = RREG32(CP_HPD_EOP_CONTROL);
+		tmp &= ~EOP_SIZE_MASK;
+		tmp |= order_base_2(MEC_HPD_SIZE / 8);
+		WREG32(CP_HPD_EOP_CONTROL, tmp);
 
+	}
+	cik_srbm_select(rdev, 0, 0, 0, 0);
 	mutex_unlock(&rdev->srbm_mutex);
 
 	/* init the queues.  Just two for now. */
@@ -5583,15 +5210,21 @@ static void cik_gpu_pci_config_reset(struct radeon_device *rdev)
  * cik_asic_reset - soft reset GPU
  *
  * @rdev: radeon_device pointer
+ * @hard: force hard reset
  *
  * Look up which blocks are hung and attempt
  * to reset them.
  * Returns 0 for success.
  */
-int cik_asic_reset(struct radeon_device *rdev)
+int cik_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	u32 reset_mask;
 
+	if (hard) {
+		cik_gpu_pci_config_reset(rdev);
+		return 0;
+	}
+
 	reset_mask = cik_gpu_check_soft_reset(rdev);
 
 	if (reset_mask)
@@ -5782,28 +5415,6 @@ void cik_pcie_gart_tlb_flush(struct radeon_device *rdev)
 	WREG32(VM_INVALIDATE_REQUEST, 0x1);
 }
 
-static void cik_pcie_init_compute_vmid(struct radeon_device *rdev)
-{
-	int i;
-	uint32_t sh_mem_bases, sh_mem_config;
-
-	sh_mem_bases = 0x6000 | 0x6000 << 16;
-	sh_mem_config = ALIGNMENT_MODE(SH_MEM_ALIGNMENT_MODE_UNALIGNED);
-	sh_mem_config |= DEFAULT_MTYPE(MTYPE_NONCACHED);
-
-	mutex_lock(&rdev->srbm_mutex);
-	for (i = 8; i < 16; i++) {
-		cik_srbm_select(rdev, 0, 0, 0, i);
-		/* CP and shaders */
-		WREG32(SH_MEM_CONFIG, sh_mem_config);
-		WREG32(SH_MEM_APE1_BASE, 1);
-		WREG32(SH_MEM_APE1_LIMIT, 0);
-		WREG32(SH_MEM_BASES, sh_mem_bases);
-	}
-	cik_srbm_select(rdev, 0, 0, 0, 0);
-	mutex_unlock(&rdev->srbm_mutex);
-}
-
 /**
  * cik_pcie_gart_enable - gart enable
  *
@@ -5903,7 +5514,7 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 	for (i = 0; i < 16; i++) {
 		cik_srbm_select(rdev, 0, 0, 0, i);
 		/* CP and shaders */
-		WREG32(SH_MEM_CONFIG, 0);
+		WREG32(SH_MEM_CONFIG, SH_MEM_CONFIG_GFX_DEFAULT);
 		WREG32(SH_MEM_APE1_BASE, 1);
 		WREG32(SH_MEM_APE1_LIMIT, 0);
 		WREG32(SH_MEM_BASES, 0);
@@ -5917,8 +5528,6 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 	cik_srbm_select(rdev, 0, 0, 0, 0);
 	mutex_unlock(&rdev->srbm_mutex);
 
-	cik_pcie_init_compute_vmid(rdev);
-
 	cik_pcie_gart_tlb_flush(rdev);
 	DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n",
 		 (unsigned)(rdev->mc.gtt_size >> 20),
@@ -6014,10 +5623,9 @@ int cik_vm_init(struct radeon_device *rdev)
 	/*
 	 * number of VMs
 	 * VMID 0 is reserved for System
-	 * radeon graphics/compute will use VMIDs 1-7
-	 * amdkfd will use VMIDs 8-15
+	 * radeon graphics/compute will use VMIDs 1-15
 	 */
-	rdev->vm_manager.nvm = RADEON_NUM_OF_VMIDS;
+	rdev->vm_manager.nvm = 16;
 	/* base offset of vram pages */
 	if (rdev->flags & RADEON_IS_IGP) {
 		u64 tmp = RREG32(MC_VM_FB_OFFSET);
@@ -6110,7 +5718,7 @@ void cik_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
 	radeon_ring_write(ring, 0);
 
 	radeon_ring_write(ring, 0); /* SH_MEM_BASES */
-	radeon_ring_write(ring, 0); /* SH_MEM_CONFIG */
+	radeon_ring_write(ring, SH_MEM_CONFIG_GFX_DEFAULT); /* SH_MEM_CONFIG */
 	radeon_ring_write(ring, 1); /* SH_MEM_APE1_BASE */
 	radeon_ring_write(ring, 0); /* SH_MEM_APE1_LIMIT */
 
@@ -6186,7 +5794,6 @@ static void cik_wait_for_rlc_serdes(struct radeon_device *rdev)
 	u32 i, j, k;
 	u32 mask;
 
-	mutex_lock(&rdev->grbm_idx_mutex);
 	for (i = 0; i < rdev->config.cik.max_shader_engines; i++) {
 		for (j = 0; j < rdev->config.cik.max_sh_per_se; j++) {
 			cik_select_se_sh(rdev, i, j);
@@ -6198,7 +5805,6 @@ static void cik_wait_for_rlc_serdes(struct radeon_device *rdev)
 		}
 	}
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	mask = SE_MASTER_BUSY_MASK | GC_MASTER_BUSY | TC0_MASTER_BUSY | TC1_MASTER_BUSY;
 	for (k = 0; k < rdev->usec_timeout; k++) {
@@ -6333,12 +5939,10 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 	WREG32(RLC_LB_CNTR_INIT, 0);
 	WREG32(RLC_LB_CNTR_MAX, 0x00008000);
 
-	mutex_lock(&rdev->grbm_idx_mutex);
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 	WREG32(RLC_LB_INIT_CU_MASK, 0xffffffff);
 	WREG32(RLC_LB_PARAMS, 0x00600408);
 	WREG32(RLC_LB_CNTL, 0x80000004);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	WREG32(RLC_MC_CNTL, 0);
 	WREG32(RLC_UCODE_CNTL, 0);
@@ -6405,13 +6009,11 @@ static void cik_enable_cgcg(struct radeon_device *rdev, bool enable)
 
 		tmp = cik_halt_rlc(rdev);
 
-		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		tmp2 = BPM_ADDR_MASK | CGCG_OVERRIDE_0 | CGLS_ENABLE;
 		WREG32(RLC_SERDES_WR_CTRL, tmp2);
-		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 
@@ -6454,13 +6056,11 @@ static void cik_enable_mgcg(struct radeon_device *rdev, bool enable)
 
 		tmp = cik_halt_rlc(rdev);
 
-		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		data = BPM_ADDR_MASK | MGCG_OVERRIDE_0;
 		WREG32(RLC_SERDES_WR_CTRL, data);
-		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 
@@ -6504,13 +6104,11 @@ static void cik_enable_mgcg(struct radeon_device *rdev, bool enable)
 
 		tmp = cik_halt_rlc(rdev);
 
-		mutex_lock(&rdev->grbm_idx_mutex);
 		cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
 		WREG32(RLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff);
 		WREG32(RLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff);
 		data = BPM_ADDR_MASK | MGCG_OVERRIDE_1;
 		WREG32(RLC_SERDES_WR_CTRL, data);
-		mutex_unlock(&rdev->grbm_idx_mutex);
 
 		cik_update_rlc(rdev, tmp);
 	}
@@ -6939,12 +6537,10 @@ static u32 cik_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh)
 	u32 mask = 0, tmp, tmp1;
 	int i;
 
-	mutex_lock(&rdev->grbm_idx_mutex);
 	cik_select_se_sh(rdev, se, sh);
 	tmp = RREG32(CC_GC_SHADER_ARRAY_CONFIG);
 	tmp1 = RREG32(GC_USER_SHADER_ARRAY_CONFIG);
 	cik_select_se_sh(rdev, 0xffffffff, 0xffffffff);
-	mutex_unlock(&rdev->grbm_idx_mutex);
 
 	tmp &= 0xffff0000;
 
@@ -7430,7 +7026,8 @@ static int cik_irq_init(struct radeon_device *rdev)
 int cik_irq_set(struct radeon_device *rdev)
 {
 	u32 cp_int_cntl;
-	u32 cp_m1p0;
+	u32 cp_m1p0, cp_m1p1, cp_m1p2, cp_m1p3;
+	u32 cp_m2p0, cp_m2p1, cp_m2p2, cp_m2p3;
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6;
 	u32 grbm_int_cntl = 0;
@@ -7463,6 +7060,13 @@ int cik_irq_set(struct radeon_device *rdev)
 	dma_cntl1 = RREG32(SDMA0_CNTL + SDMA1_REGISTER_OFFSET) & ~TRAP_ENABLE;
 
 	cp_m1p0 = RREG32(CP_ME1_PIPE0_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m1p1 = RREG32(CP_ME1_PIPE1_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m1p2 = RREG32(CP_ME1_PIPE2_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m1p3 = RREG32(CP_ME1_PIPE3_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m2p0 = RREG32(CP_ME2_PIPE0_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m2p1 = RREG32(CP_ME2_PIPE1_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m2p2 = RREG32(CP_ME2_PIPE2_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
+	cp_m2p3 = RREG32(CP_ME2_PIPE3_INT_CNTL) & ~TIME_STAMP_INT_ENABLE;
 
 	/* enable CP interrupts on all rings */
 	if (atomic_read(&rdev->irq.ring_int[RADEON_RING_TYPE_GFX_INDEX])) {
@@ -7477,6 +7081,33 @@ int cik_irq_set(struct radeon_device *rdev)
 			case 0:
 				cp_m1p0 |= TIME_STAMP_INT_ENABLE;
 				break;
+			case 1:
+				cp_m1p1 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 2:
+				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 3:
+				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			default:
+				DRM_DEBUG("si_irq_set: sw int cp1 invalid pipe %d\n", ring->pipe);
+				break;
+			}
+		} else if (ring->me == 2) {
+			switch (ring->pipe) {
+			case 0:
+				cp_m2p0 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 1:
+				cp_m2p1 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 2:
+				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 3:
+				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
+				break;
 			default:
 				DRM_DEBUG("si_irq_set: sw int cp1 invalid pipe %d\n", ring->pipe);
 				break;
@@ -7493,6 +7124,33 @@ int cik_irq_set(struct radeon_device *rdev)
 			case 0:
 				cp_m1p0 |= TIME_STAMP_INT_ENABLE;
 				break;
+			case 1:
+				cp_m1p1 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 2:
+				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 3:
+				cp_m1p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			default:
+				DRM_DEBUG("si_irq_set: sw int cp2 invalid pipe %d\n", ring->pipe);
+				break;
+			}
+		} else if (ring->me == 2) {
+			switch (ring->pipe) {
+			case 0:
+				cp_m2p0 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 1:
+				cp_m2p1 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 2:
+				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
+				break;
+			case 3:
+				cp_m2p2 |= TIME_STAMP_INT_ENABLE;
+				break;
 			default:
 				DRM_DEBUG("si_irq_set: sw int cp2 invalid pipe %d\n", ring->pipe);
 				break;
@@ -7573,6 +7231,13 @@ int cik_irq_set(struct radeon_device *rdev)
 	WREG32(SDMA0_CNTL + SDMA1_REGISTER_OFFSET, dma_cntl1);
 
 	WREG32(CP_ME1_PIPE0_INT_CNTL, cp_m1p0);
+	WREG32(CP_ME1_PIPE1_INT_CNTL, cp_m1p1);
+	WREG32(CP_ME1_PIPE2_INT_CNTL, cp_m1p2);
+	WREG32(CP_ME1_PIPE3_INT_CNTL, cp_m1p3);
+	WREG32(CP_ME2_PIPE0_INT_CNTL, cp_m2p0);
+	WREG32(CP_ME2_PIPE1_INT_CNTL, cp_m2p1);
+	WREG32(CP_ME2_PIPE2_INT_CNTL, cp_m2p2);
+	WREG32(CP_ME2_PIPE3_INT_CNTL, cp_m2p3);
 
 	WREG32(GRBM_INT_CNTL, grbm_int_cntl);
 
@@ -7921,11 +7586,6 @@ restart_ih:
 		/* wptr/rptr are in bytes! */
 		ring_index = rptr / 4;
 
-#ifdef notyet
-		radeon_kfd_interrupt(rdev,
-				(const void *) &rdev->ih.ring[ring_index]);
-#endif
-
 		src_id =  le32_to_cpu(rdev->ih.ring[ring_index]) & 0xff;
 		src_data = le32_to_cpu(rdev->ih.ring[ring_index + 1]) & 0xfffffff;
 		ring_id = le32_to_cpu(rdev->ih.ring[ring_index + 2]) & 0xff;
@@ -8463,6 +8123,164 @@ restart_ih:
 /*
  * startup/shutdown callbacks
  */
+static void cik_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails cik_uvd_start() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void cik_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = uvd_v4_2_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD 4.2 resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void cik_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
+static void cik_vce_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE (%d) init.\n", r);
+		/*
+		 * At this point rdev->vce.vcpu_bo is NULL which trickles down
+		 * to early fails cik_vce_start() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable vce here.
+		 */
+		rdev->has_vce = 0;
+		return;
+	}
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE1_INDEX], 4096);
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE2_INDEX], 4096);
+}
+
+static void cik_vce_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = vce_v2_0_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE1_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 fences (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE2_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE2 fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
+}
+
+static void cik_vce_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_vce || !rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, VCE_CMD_NO_OP);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, VCE_CMD_NO_OP);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	r = vce_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE (%d).\n", r);
+		return;
+	}
+}
+
 /**
  * cik_startup - program the asic to a functional state
  *
@@ -8516,7 +8334,8 @@ static int cik_startup(struct radeon_device *rdev)
 		}
 	}
 	rdev->rlc.cs_data = ci_cs_data;
-	rdev->rlc.cp_table_size = CP_ME_TABLE_SIZE * 5 * 4;
+	rdev->rlc.cp_table_size = roundup2(CP_ME_TABLE_SIZE * 5 * 4, 2048); /* CP JT */
+	rdev->rlc.cp_table_size += 64 * 1024; /* GDS */
 	r = sumo_rlc_init(rdev);
 	if (r) {
 		DRM_ERROR("Failed to init rlc BOs!\n");
@@ -8565,34 +8384,8 @@ static int cik_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = radeon_uvd_resume(rdev);
-	if (!r) {
-		r = uvd_v4_2_resume(rdev);
-		if (!r) {
-			r = radeon_fence_driver_start_ring(rdev,
-							   R600_RING_TYPE_UVD_INDEX);
-			if (r)
-				dev_err(rdev->dev, "UVD fences init error (%d).\n", r);
-		}
-	}
-	if (r)
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
-
-	r = radeon_vce_resume(rdev);
-	if (!r) {
-		r = vce_v2_0_resume(rdev);
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE1_INDEX);
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE2_INDEX);
-	}
-	if (r) {
-		dev_err(rdev->dev, "VCE init error (%d).\n", r);
-		rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
-		rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
-	}
+	cik_uvd_start(rdev);
+	cik_vce_start(rdev);
 
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
@@ -8668,32 +8461,8 @@ static int cik_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-	if (ring->ring_size) {
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     RADEON_CP_PACKET2);
-		if (!r)
-			r = uvd_v1_0_init(rdev);
-		if (r)
-			DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
-	}
-
-	r = -ENOENT;
-
-	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-	if (ring->ring_size)
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     VCE_CMD_NO_OP);
-
-	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-	if (ring->ring_size)
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     VCE_CMD_NO_OP);
-
-	if (!r)
-		r = vce_v1_0_init(rdev);
-	else if (r != -ENOENT)
-		DRM_ERROR("radeon: failed initializing VCE (%d).\n", r);
+	cik_uvd_resume(rdev);
+	cik_vce_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -8711,12 +8480,6 @@ static int cik_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-#ifdef notyet
-	r = radeon_kfd_resume(rdev);
-	if (r)
-		return r;
-#endif
-
 	return 0;
 }
 
@@ -8765,17 +8528,17 @@ int cik_resume(struct radeon_device *rdev)
  */
 int cik_suspend(struct radeon_device *rdev)
 {
-#ifdef notyet
-	radeon_kfd_suspend(rdev);
-#endif
 	radeon_pm_suspend(rdev);
 	radeon_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
 	cik_cp_enable(rdev, false);
 	cik_sdma_enable(rdev, false);
-	uvd_v1_0_fini(rdev);
-	radeon_uvd_suspend(rdev);
-	radeon_vce_suspend(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
+	if (rdev->has_vce)
+		radeon_vce_suspend(rdev);
 	cik_fini_pg(rdev);
 	cik_fini_cg(rdev);
 	cik_irq_suspend(rdev);
@@ -8901,23 +8664,8 @@ int cik_init(struct radeon_device *rdev)
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 256 * 1024);
 
-	r = radeon_uvd_init(rdev);
-	if (!r) {
-		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-	}
-
-	r = radeon_vce_init(rdev);
-	if (!r) {
-		ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-
-		ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-	}
+	cik_uvd_init(rdev);
+	cik_vce_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -9392,23 +9140,10 @@ static u32 dce8_latency_watermark(struct dce8_wm_params *wm)
 	a.full = dfixed_const(available_bandwidth);
 	b.full = dfixed_const(wm->num_heads);
 	a.full = dfixed_div(a, b);
+	tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512);
+	tmp = min(dfixed_trunc(a), tmp);
 
-	b.full = dfixed_const(mc_latency + 512);
-	c.full = dfixed_const(wm->disp_clk);
-	b.full = dfixed_div(b, c);
-
-	c.full = dfixed_const(dmif_size);
-	b.full = dfixed_div(c, b);
-
-	tmp = min(dfixed_trunc(a), dfixed_trunc(b));
-
-	b.full = dfixed_const(1000);
-	c.full = dfixed_const(wm->disp_clk);
-	b.full = dfixed_div(c, b);
-	c.full = dfixed_const(wm->bytes_per_pixel);
-	b.full = dfixed_mul(b, c);
-
-	lb_fill_bw = min(tmp, dfixed_trunc(b));
+	lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000);
 
 	a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel);
 	b.full = dfixed_const(1000);
@@ -9516,14 +9251,17 @@ static void dce8_program_watermarks(struct radeon_device *rdev,
 {
 	struct drm_display_mode *mode = &radeon_crtc->base.mode;
 	struct dce8_wm_params wm_low, wm_high;
-	u32 pixel_period;
+	u32 active_time;
 	u32 line_time = 0;
 	u32 latency_watermark_a = 0, latency_watermark_b = 0;
 	u32 tmp, wm_mask;
 
 	if (radeon_crtc->base.enabled && num_heads && mode) {
-		pixel_period = 1000000 / (u32)mode->clock;
-		line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535);
+		active_time = (u32) div_u64((u64)mode->crtc_hdisplay * 1000000,
+					    (u32)mode->clock);
+		line_time = (u32) div_u64((u64)mode->crtc_htotal * 1000000,
+					  (u32)mode->clock);
+		line_time = min(line_time, (u32)65535);
 
 		/* watermark for high clocks */
 		if ((rdev->pm.pm_method == PM_METHOD_DPM) &&
@@ -9539,7 +9277,7 @@ static void dce8_program_watermarks(struct radeon_device *rdev,
 
 		wm_high.disp_clk = mode->clock;
 		wm_high.src_width = mode->crtc_hdisplay;
-		wm_high.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_high.active_time = active_time;
 		wm_high.blank_time = line_time - wm_high.active_time;
 		wm_high.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -9579,7 +9317,7 @@ static void dce8_program_watermarks(struct radeon_device *rdev,
 
 		wm_low.disp_clk = mode->clock;
 		wm_low.src_width = mode->crtc_hdisplay;
-		wm_low.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_low.active_time = active_time;
 		wm_low.blank_time = line_time - wm_low.active_time;
 		wm_low.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -9680,13 +9418,13 @@ uint64_t cik_get_gpu_clock_counter(struct radeon_device *rdev)
 	mutex_lock(&rdev->gpu_clock_mutex);
 	WREG32(RLC_CAPTURE_GPU_CLOCK_COUNT, 1);
 	clock = (uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_LSB) |
-	        ((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+		((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
 	mutex_unlock(&rdev->gpu_clock_mutex);
 	return clock;
 }
 
 static int cik_set_uvd_clock(struct radeon_device *rdev, u32 clock,
-                              u32 cntl_reg, u32 status_reg)
+			     u32 cntl_reg, u32 status_reg)
 {
 	int r, i;
 	struct atom_clock_dividers dividers;
@@ -9762,17 +9500,13 @@ int cik_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk)
 
 static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 {
-	struct pci_dev _root;
-	struct pci_dev *root;
+	struct pci_dev *root = rdev->pdev->bus->self;
+	enum pci_bus_speed speed_cap;
 	int bridge_pos, gpu_pos;
-	u32 speed_cntl, mask, current_data_rate;
-	int ret, i;
+	u32 speed_cntl, current_data_rate;
+	int i;
 	u16 tmp16;
 
-	root = &_root;
-	root->pc = rdev->pdev->pc;
-	root->tag = *rdev->ddev->bridgetag;
-
 	if (pci_is_root_bus(rdev->pdev->bus))
 		return;
 
@@ -9785,23 +9519,24 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret != 0)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN)
 		return;
 
-	if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+	if ((speed_cap != PCIE_SPEED_8_0GT) &&
+	    (speed_cap != PCIE_SPEED_5_0GT))
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
 		LC_CURRENT_DATA_RATE_SHIFT;
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		if (current_data_rate == 2) {
 			DRM_INFO("PCIE gen 3 link speeds already enabled\n");
 			return;
 		}
 		DRM_INFO("enabling PCIE gen 3 link speeds, disable with radeon.pcie_gen2=0\n");
-	} else if (mask & DRM_PCIE_SPEED_50) {
+	} else if (speed_cap == PCIE_SPEED_5_0GT) {
 		if (current_data_rate == 1) {
 			DRM_INFO("PCIE gen 2 link speeds already enabled\n");
 			return;
@@ -9817,7 +9552,7 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!gpu_pos)
 		return;
 
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		/* re-try equalization if gen3 is not already enabled */
 		if (current_data_rate != 2) {
 			u16 bridge_cfg, gpu_cfg;
@@ -9905,9 +9640,9 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 
 	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
 	tmp16 &= ~0xf;
-	if (mask & DRM_PCIE_SPEED_80)
+	if (speed_cap == PCIE_SPEED_8_0GT)
 		tmp16 |= 3; /* gen3 */
-	else if (mask & DRM_PCIE_SPEED_50)
+	else if (speed_cap == PCIE_SPEED_5_0GT)
 		tmp16 |= 2; /* gen2 */
 	else
 		tmp16 |= 1; /* gen1 */
@@ -10004,13 +9739,8 @@ static void cik_program_aspm(struct radeon_device *rdev)
 
 			if (!disable_clkreq &&
 			    !pci_is_root_bus(rdev->pdev->bus)) {
+				struct pci_dev *root = rdev->pdev->bus->self;
 				u32 lnkcap;
-				struct pci_dev _root;
-				struct pci_dev *root;
-
-				root = &_root;
-				root->pc = rdev->pdev->pc;
-				root->tag = *rdev->ddev->bridgetag;
 
 				clk_req_support = false;
 				pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap);
diff --git a/sys/dev/pci/drm/radeon/cik_blit_shaders.c b/sys/dev/pci/drm/radeon/cik_blit_shaders.c
index 2edcce04bdc..ff1311806e9 100644
--- a/sys/dev/pci/drm/radeon/cik_blit_shaders.c
+++ b/sys/dev/pci/drm/radeon/cik_blit_shaders.c
@@ -24,7 +24,9 @@
  *     Alex Deucher <alexander.deucher@amd.com>
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/types.h>
+#include <linux/bug.h>
+#include <linux/kernel.h>
 
 const u32 cik_default_state[] =
 {
diff --git a/sys/dev/pci/drm/radeon/cik_reg.h b/sys/dev/pci/drm/radeon/cik_reg.h
index 4e883fdc59d..318377df09e 100644
--- a/sys/dev/pci/drm/radeon/cik_reg.h
+++ b/sys/dev/pci/drm/radeon/cik_reg.h
@@ -147,8 +147,6 @@
 
 #define CIK_LB_DESKTOP_HEIGHT                     0x6b0c
 
-#define KFD_CIK_SDMA_QUEUE_OFFSET		0x200
-
 #define SQ_IND_INDEX					0x8DE0
 #define SQ_CMD						0x8DEC
 #define SQ_IND_DATA					0x8DE4
diff --git a/sys/dev/pci/drm/radeon/cik_sdma.c b/sys/dev/pci/drm/radeon/cik_sdma.c
index 3f36fa1e3d5..9c351dc8a9e 100644
--- a/sys/dev/pci/drm/radeon/cik_sdma.c
+++ b/sys/dev/pci/drm/radeon/cik_sdma.c
@@ -21,7 +21,8 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_ucode.h"
 #include "radeon_asic.h"
@@ -736,11 +737,16 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 		return r;
 	}
-	r = radeon_fence_wait(ib.fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(ib.fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		return r;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		return -ETIMEDOUT;
 	}
+	r = 0;
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
diff --git a/sys/dev/pci/drm/radeon/cikd.h b/sys/dev/pci/drm/radeon/cikd.h
index 391ff9d5d70..cda16fcd43b 100644
--- a/sys/dev/pci/drm/radeon/cikd.h
+++ b/sys/dev/pci/drm/radeon/cikd.h
@@ -30,8 +30,6 @@
 #define CIK_RB_BITMAP_WIDTH_PER_SH     2
 #define HAWAII_RB_BITMAP_WIDTH_PER_SH  4
 
-#define RADEON_NUM_OF_VMIDS	8
-
 /* DIDT IND registers */
 #define DIDT_SQ_CTRL0                                     0x0
 #       define DIDT_CTRL_EN                               (1 << 0)
@@ -2016,7 +2014,7 @@
 #       define SDMA_COPY_SUB_OPCODE_T2T_SUB_WINDOW        6
 #define	SDMA_OPCODE_WRITE				  2
 #       define SDMA_WRITE_SUB_OPCODE_LINEAR               0
-#       define SDMA_WRTIE_SUB_OPCODE_TILED                1
+#       define SDMA_WRITE_SUB_OPCODE_TILED                1
 #define	SDMA_OPCODE_INDIRECT_BUFFER			  4
 #define	SDMA_OPCODE_FENCE				  5
 #define	SDMA_OPCODE_TRAP				  6
@@ -2069,8 +2067,10 @@
 #define UVD_UDEC_ADDR_CONFIG		0xef4c
 #define UVD_UDEC_DB_ADDR_CONFIG		0xef50
 #define UVD_UDEC_DBW_ADDR_CONFIG	0xef54
+#define UVD_NO_OP			0xeffc
 
 #define UVD_LMI_EXT40_ADDR		0xf498
+#define UVD_GP_SCRATCH4			0xf4e0
 #define UVD_LMI_ADDR_EXT		0xf594
 #define UVD_VCPU_CACHE_OFFSET0		0xf608
 #define UVD_VCPU_CACHE_SIZE0		0xf60c
diff --git a/sys/dev/pci/drm/radeon/cypress_dpm.c b/sys/dev/pci/drm/radeon/cypress_dpm.c
index 6a18da7c757..3eb7899a403 100644
--- a/sys/dev/pci/drm/radeon/cypress_dpm.c
+++ b/sys/dev/pci/drm/radeon/cypress_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "evergreend.h"
@@ -1620,14 +1620,14 @@ static int cypress_init_smc_table(struct radeon_device *rdev,
 	cypress_populate_smc_voltage_tables(rdev, table);
 
 	switch (rdev->pm.int_thermal_type) {
-        case THERMAL_TYPE_EVERGREEN:
-        case THERMAL_TYPE_EMC2103_WITH_INTERNAL:
+	case THERMAL_TYPE_EVERGREEN:
+	case THERMAL_TYPE_EMC2103_WITH_INTERNAL:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_INTERNAL;
 		break;
-        case THERMAL_TYPE_NONE:
+	case THERMAL_TYPE_NONE:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_NONE;
 		break;
-        default:
+	default:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_EXTERNAL;
 		break;
 	}
diff --git a/sys/dev/pci/drm/radeon/dce3_1_afmt.c b/sys/dev/pci/drm/radeon/dce3_1_afmt.c
index 228e8ac00f2..cfa3a84a2af 100644
--- a/sys/dev/pci/drm/radeon/dce3_1_afmt.c
+++ b/sys/dev/pci/drm/radeon/dce3_1_afmt.c
@@ -20,8 +20,8 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  */
-#include <dev/pci/drm/linux_hdmi.h>
-#include <dev/pci/drm/drmP.h>
+#include <linux/hdmi.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
diff --git a/sys/dev/pci/drm/radeon/dce6_afmt.c b/sys/dev/pci/drm/radeon/dce6_afmt.c
index c342965908b..367a916f364 100644
--- a/sys/dev/pci/drm/radeon/dce6_afmt.c
+++ b/sys/dev/pci/drm/radeon/dce6_afmt.c
@@ -20,8 +20,8 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include <dev/pci/drm/linux_hdmi.h>
-#include <dev/pci/drm/drmP.h>
+#include <linux/hdmi.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_audio.h"
 #include "sid.h"
diff --git a/sys/dev/pci/drm/radeon/drm_buffer.h b/sys/dev/pci/drm/radeon/drm_buffer.h
deleted file mode 100644
index 4eaa26fbad1..00000000000
--- a/sys/dev/pci/drm/radeon/drm_buffer.h
+++ /dev/null
@@ -1,148 +0,0 @@
-/**************************************************************************
- *
- * Copyright 2010 Pauli Nieminen.
- * All Rights Reserved.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the
- * "Software"), to deal in the Software without restriction, including
- * without limitation the rights to use, copy, modify, merge, publish,
- * distribute, sub license, and/or sell copies of the Software, and to
- * permit persons to whom the Software is furnished to do so, subject to
- * the following conditions:
- *
- * The above copyright notice and this permission notice (including the
- * next paragraph) shall be included in all copies or substantial portions
- * of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
- * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
- * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
- * USE OR OTHER DEALINGS IN THE SOFTWARE.
- *
- *
- **************************************************************************/
-/*
- * Multipart buffer for coping data which is larger than the page size.
- *
- * Authors:
- * Pauli Nieminen <suokkos-at-gmail-dot-com>
- */
-
-#ifndef _DRM_BUFFER_H_
-#define _DRM_BUFFER_H_
-
-#include <dev/pci/drm/drmP.h>
-
-struct drm_buffer {
-	int iterator;
-	int size;
-	char *data[];
-};
-
-
-/**
- * Return the index of page that buffer is currently pointing at.
- */
-static inline int drm_buffer_page(struct drm_buffer *buf)
-{
-	return buf->iterator / PAGE_SIZE;
-}
-/**
- * Return the index of the current byte in the page
- */
-static inline int drm_buffer_index(struct drm_buffer *buf)
-{
-	return buf->iterator & (PAGE_SIZE - 1);
-}
-/**
- * Return number of bytes that is left to process
- */
-static inline int drm_buffer_unprocessed(struct drm_buffer *buf)
-{
-	return buf->size - buf->iterator;
-}
-
-/**
- * Advance the buffer iterator number of bytes that is given.
- */
-static inline void drm_buffer_advance(struct drm_buffer *buf, int bytes)
-{
-	buf->iterator += bytes;
-}
-
-/**
- * Allocate the drm buffer object.
- *
- *   buf: A pointer to a pointer where the object is stored.
- *   size: The number of bytes to allocate.
- */
-extern int drm_buffer_alloc(struct drm_buffer **buf, int size);
-
-/**
- * Copy the user data to the begin of the buffer and reset the processing
- * iterator.
- *
- *   user_data: A pointer the data that is copied to the buffer.
- *   size: The Number of bytes to copy.
- */
-extern int drm_buffer_copy_from_user(struct drm_buffer *buf,
-		void __user *user_data, int size);
-
-/**
- * Free the drm buffer object
- */
-extern void drm_buffer_free(struct drm_buffer *buf);
-
-/**
- * Read an object from buffer that may be split to multiple parts. If object
- * is not split function just returns the pointer to object in buffer. But in
- * case of split object data is copied to given stack object that is suplied
- * by caller.
- *
- * The processing location of the buffer is also advanced to the next byte
- * after the object.
- *
- *   objsize: The size of the objet in bytes.
- *   stack_obj: A pointer to a memory location where object can be copied.
- */
-extern void *drm_buffer_read_object(struct drm_buffer *buf,
-		int objsize, void *stack_obj);
-
-/**
- * Returns the pointer to the dword which is offset number of elements from the
- * current processing location.
- *
- * Caller must make sure that dword is not split in the buffer. This
- * requirement is easily met if all the sizes of objects in buffer are
- * multiples of dword and PAGE_SIZE is multiple dword.
- *
- * Call to this function doesn't change the processing location.
- *
- *   offset: The index of the dword relative to the internat iterator.
- */
-static inline void *drm_buffer_pointer_to_dword(struct drm_buffer *buffer,
-		int offset)
-{
-	int iter = buffer->iterator + offset * 4;
-	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
-}
-/**
- * Returns the pointer to the dword which is offset number of elements from
- * the current processing location.
- *
- * Call to this function doesn't change the processing location.
- *
- *   offset: The index of the byte relative to the internat iterator.
- */
-static inline void *drm_buffer_pointer_to_byte(struct drm_buffer *buffer,
-		int offset)
-{
-	int iter = buffer->iterator + offset;
-	return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)];
-}
-
-#endif
diff --git a/sys/dev/pci/drm/radeon/evergreen.c b/sys/dev/pci/drm/radeon/evergreen.c
index e10aca71c96..df1f7cd7a6b 100644
--- a/sys/dev/pci/drm/radeon/evergreen.c
+++ b/sys/dev/pci/drm/radeon/evergreen.c
@@ -21,11 +21,13 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "evergreend.h"
 #include "atom.h"
 #include "avivod.h"
@@ -33,6 +35,10 @@
 #include "evergreen_blit_shaders.h"
 #include "radeon_ucode.h"
 
+#define DC_HPDx_CONTROL(x)        (DC_HPD1_CONTROL     + (x * 0xc))
+#define DC_HPDx_INT_CONTROL(x)    (DC_HPD1_INT_CONTROL + (x * 0xc))
+#define DC_HPDx_INT_STATUS_REG(x) (DC_HPD1_INT_STATUS  + (x * 0xc))
+
 /*
  * Indirect registers accessor
  */
@@ -1138,7 +1144,7 @@ static int sumo_set_uvd_clock(struct radeon_device *rdev, u32 clock,
 	int r, i;
 	struct atom_clock_dividers dividers;
 
-        r = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+	r = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
 					   clock, false, &dividers);
 	if (r)
 		return r;
@@ -1430,11 +1436,14 @@ void dce4_wait_for_vblank(struct radeon_device *rdev, int crtc)
  * Triggers the actual pageflip by updating the primary
  * surface base address (evergreen+).
  */
-void evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base,
+			 bool async)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 
 	/* update the scanout addresses */
+	WREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset,
+	       async ? EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN : 0);
 	WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + radeon_crtc->crtc_offset,
 	       upper_32_bits(crtc_base));
 	WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
@@ -1734,38 +1743,10 @@ void evergreen_pm_finish(struct radeon_device *rdev)
  */
 bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd)
 {
-	bool connected = false;
-
-	switch (hpd) {
-	case RADEON_HPD_1:
-		if (RREG32(DC_HPD1_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	case RADEON_HPD_2:
-		if (RREG32(DC_HPD2_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	case RADEON_HPD_3:
-		if (RREG32(DC_HPD3_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	case RADEON_HPD_4:
-		if (RREG32(DC_HPD4_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	case RADEON_HPD_5:
-		if (RREG32(DC_HPD5_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	case RADEON_HPD_6:
-		if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE)
-			connected = true;
-		break;
-	default:
-		break;
-	}
+	if (hpd == RADEON_HPD_NONE)
+		return false;
 
-	return connected;
+	return !!(RREG32(DC_HPDx_INT_STATUS_REG(hpd)) & DC_HPDx_SENSE);
 }
 
 /**
@@ -1779,61 +1760,15 @@ bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd)
 void evergreen_hpd_set_polarity(struct radeon_device *rdev,
 				enum radeon_hpd_id hpd)
 {
-	u32 tmp;
 	bool connected = evergreen_hpd_sense(rdev, hpd);
 
-	switch (hpd) {
-	case RADEON_HPD_1:
-		tmp = RREG32(DC_HPD1_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-		break;
-	case RADEON_HPD_2:
-		tmp = RREG32(DC_HPD2_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-		break;
-	case RADEON_HPD_3:
-		tmp = RREG32(DC_HPD3_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-		break;
-	case RADEON_HPD_4:
-		tmp = RREG32(DC_HPD4_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-		break;
-	case RADEON_HPD_5:
-		tmp = RREG32(DC_HPD5_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-			break;
-	case RADEON_HPD_6:
-		tmp = RREG32(DC_HPD6_INT_CONTROL);
-		if (connected)
-			tmp &= ~DC_HPDx_INT_POLARITY;
-		else
-			tmp |= DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
-		break;
-	default:
-		break;
-	}
+	if (hpd == RADEON_HPD_NONE)
+		return;
+
+	if (connected)
+		WREG32_AND(DC_HPDx_INT_CONTROL(hpd), ~DC_HPDx_INT_POLARITY);
+	else
+		WREG32_OR(DC_HPDx_INT_CONTROL(hpd), DC_HPDx_INT_POLARITY);
 }
 
 /**
@@ -1853,7 +1788,8 @@ void evergreen_hpd_init(struct radeon_device *rdev)
 		DC_HPDx_RX_INT_TIMER(0xfa) | DC_HPDx_EN;
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+		enum radeon_hpd_id hpd =
+			to_radeon_connector(connector)->hpd.hpd;
 
 		if (connector->connector_type == DRM_MODE_CONNECTOR_eDP ||
 		    connector->connector_type == DRM_MODE_CONNECTOR_LVDS) {
@@ -1864,30 +1800,14 @@ void evergreen_hpd_init(struct radeon_device *rdev)
 			 */
 			continue;
 		}
-		switch (radeon_connector->hpd.hpd) {
-		case RADEON_HPD_1:
-			WREG32(DC_HPD1_CONTROL, tmp);
-			break;
-		case RADEON_HPD_2:
-			WREG32(DC_HPD2_CONTROL, tmp);
-			break;
-		case RADEON_HPD_3:
-			WREG32(DC_HPD3_CONTROL, tmp);
-			break;
-		case RADEON_HPD_4:
-			WREG32(DC_HPD4_CONTROL, tmp);
-			break;
-		case RADEON_HPD_5:
-			WREG32(DC_HPD5_CONTROL, tmp);
-			break;
-		case RADEON_HPD_6:
-			WREG32(DC_HPD6_CONTROL, tmp);
-			break;
-		default:
-			break;
-		}
-		radeon_hpd_set_polarity(rdev, radeon_connector->hpd.hpd);
-		enabled |= 1 << radeon_connector->hpd.hpd;
+
+		if (hpd == RADEON_HPD_NONE)
+			continue;
+
+		WREG32(DC_HPDx_CONTROL(hpd), tmp);
+		enabled |= 1 << hpd;
+
+		radeon_hpd_set_polarity(rdev, hpd);
 	}
 	radeon_irq_kms_enable_hpd(rdev, enabled);
 }
@@ -1907,30 +1827,14 @@ void evergreen_hpd_fini(struct radeon_device *rdev)
 	unsigned disabled = 0;
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		struct radeon_connector *radeon_connector = to_radeon_connector(connector);
-		switch (radeon_connector->hpd.hpd) {
-		case RADEON_HPD_1:
-			WREG32(DC_HPD1_CONTROL, 0);
-			break;
-		case RADEON_HPD_2:
-			WREG32(DC_HPD2_CONTROL, 0);
-			break;
-		case RADEON_HPD_3:
-			WREG32(DC_HPD3_CONTROL, 0);
-			break;
-		case RADEON_HPD_4:
-			WREG32(DC_HPD4_CONTROL, 0);
-			break;
-		case RADEON_HPD_5:
-			WREG32(DC_HPD5_CONTROL, 0);
-			break;
-		case RADEON_HPD_6:
-			WREG32(DC_HPD6_CONTROL, 0);
-			break;
-		default:
-			break;
-		}
-		disabled |= 1 << radeon_connector->hpd.hpd;
+		enum radeon_hpd_id hpd =
+			to_radeon_connector(connector)->hpd.hpd;
+
+		if (hpd == RADEON_HPD_NONE)
+			continue;
+
+		WREG32(DC_HPDx_CONTROL(hpd), 0);
+		disabled |= 1 << hpd;
 	}
 	radeon_irq_kms_disable_hpd(rdev, disabled);
 }
@@ -2206,13 +2110,7 @@ static u32 evergreen_latency_watermark(struct evergreen_wm_params *wm)
 	b.full = dfixed_const(wm->num_heads);
 	a.full = dfixed_div(a, b);
 
-	b.full = dfixed_const(1000);
-	c.full = dfixed_const(wm->disp_clk);
-	b.full = dfixed_div(c, b);
-	c.full = dfixed_const(wm->bytes_per_pixel);
-	b.full = dfixed_mul(b, c);
-
-	lb_fill_bw = min(dfixed_trunc(a), dfixed_trunc(b));
+	lb_fill_bw = min(dfixed_trunc(a), wm->disp_clk * wm->bytes_per_pixel / 1000);
 
 	a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel);
 	b.full = dfixed_const(1000);
@@ -2279,7 +2177,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev,
 	struct drm_display_mode *mode = &radeon_crtc->base.mode;
 	struct evergreen_wm_params wm_low, wm_high;
 	u32 dram_channels;
-	u32 pixel_period;
+	u32 active_time;
 	u32 line_time = 0;
 	u32 latency_watermark_a = 0, latency_watermark_b = 0;
 	u32 priority_a_mark = 0, priority_b_mark = 0;
@@ -2290,8 +2188,11 @@ static void evergreen_program_watermarks(struct radeon_device *rdev,
 	fixed20_12 a, b, c;
 
 	if (radeon_crtc->base.enabled && num_heads && mode) {
-		pixel_period = 1000000 / (u32)mode->clock;
-		line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535);
+		active_time = (u32) div_u64((u64)mode->crtc_hdisplay * 1000000,
+					    (u32)mode->clock);
+		line_time = (u32) div_u64((u64)mode->crtc_htotal * 1000000,
+					  (u32)mode->clock);
+		line_time = min(line_time, (u32)65535);
 		priority_a_cnt = 0;
 		priority_b_cnt = 0;
 		dram_channels = evergreen_get_number_of_dram_channels(rdev);
@@ -2309,7 +2210,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev,
 
 		wm_high.disp_clk = mode->clock;
 		wm_high.src_width = mode->crtc_hdisplay;
-		wm_high.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_high.active_time = active_time;
 		wm_high.blank_time = line_time - wm_high.active_time;
 		wm_high.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -2336,7 +2237,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev,
 
 		wm_low.disp_clk = mode->clock;
 		wm_low.src_width = mode->crtc_hdisplay;
-		wm_low.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_low.active_time = active_time;
 		wm_low.blank_time = line_time - wm_low.active_time;
 		wm_low.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -2504,7 +2405,7 @@ void evergreen_pcie_gart_tlb_flush(struct radeon_device *rdev)
 		tmp = RREG32(VM_CONTEXT0_REQUEST_RESPONSE);
 		tmp = (tmp & RESPONSE_TYPE_MASK) >> RESPONSE_TYPE_SHIFT;
 		if (tmp == 2) {
-			printk(KERN_WARNING "[drm] r600 flush TLB failed\n");
+			pr_warn("[drm] r600 flush TLB failed\n");
 			return;
 		}
 		if (tmp) {
@@ -2661,6 +2562,15 @@ static const unsigned evergreen_dp_offsets[] =
 	EVERGREEN_DP5_REGISTER_OFFSET
 };
 
+static const unsigned evergreen_disp_int_status[] =
+{
+	DISP_INTERRUPT_STATUS,
+	DISP_INTERRUPT_STATUS_CONTINUE,
+	DISP_INTERRUPT_STATUS_CONTINUE2,
+	DISP_INTERRUPT_STATUS_CONTINUE3,
+	DISP_INTERRUPT_STATUS_CONTINUE4,
+	DISP_INTERRUPT_STATUS_CONTINUE5
+};
 
 /*
  * Assumption is that EVERGREEN_CRTC_MASTER_EN enable for requested crtc
@@ -2896,9 +2806,8 @@ void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *s
 	for (i = 0; i < rdev->num_crtc; i++) {
 		if (save->crtc_enabled[i]) {
 			tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]);
-			if ((tmp & 0x7) != 3) {
+			if ((tmp & 0x7) != 0) {
 				tmp &= ~0x7;
-				tmp |= 0x3;
 				WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
 			}
 			tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]);
@@ -3629,6 +3538,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
 		tmp = r6xx_remap_render_backend(rdev, tmp, rdev->config.evergreen.max_backends,
 						EVERGREEN_MAX_BACKENDS, disabled_rb_mask);
 	}
+	rdev->config.evergreen.backend_map = tmp;
 	WREG32(GB_BACKEND_MAP, tmp);
 
 	WREG32(CGTS_SYS_TCC_DISABLE, 0);
@@ -4159,10 +4069,15 @@ void evergreen_gpu_pci_config_reset(struct radeon_device *rdev)
 	}
 }
 
-int evergreen_asic_reset(struct radeon_device *rdev)
+int evergreen_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	u32 reset_mask;
 
+	if (hard) {
+		evergreen_gpu_pci_config_reset(rdev);
+		return 0;
+	}
+
 	reset_mask = evergreen_gpu_check_soft_reset(rdev);
 
 	if (reset_mask)
@@ -4563,6 +4478,7 @@ u32 evergreen_get_vblank_counter(struct radeon_device *rdev, int crtc)
 
 void evergreen_disable_interrupt_state(struct radeon_device *rdev)
 {
+	int i;
 	u32 tmp;
 
 	if (rdev->family >= CHIP_CAYMAN) {
@@ -4578,56 +4494,27 @@ void evergreen_disable_interrupt_state(struct radeon_device *rdev)
 	WREG32(DMA_CNTL, tmp);
 	WREG32(GRBM_INT_CNTL, 0);
 	WREG32(SRBM_INT_CNTL, 0);
-	WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
-	WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
-	if (rdev->num_crtc >= 4) {
-		WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
-		WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
-		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
-	}
-
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
-	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
-	}
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(INT_MASK + crtc_offsets[i], 0);
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(GRPH_INT_CONTROL + crtc_offsets[i], 0);
 
 	/* only one DAC on DCE5 */
 	if (!ASIC_IS_DCE5(rdev))
 		WREG32(DACA_AUTODETECT_INT_CONTROL, 0);
 	WREG32(DACB_AUTODETECT_INT_CONTROL, 0);
 
-	tmp = RREG32(DC_HPD1_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD1_INT_CONTROL, tmp);
-	tmp = RREG32(DC_HPD2_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD2_INT_CONTROL, tmp);
-	tmp = RREG32(DC_HPD3_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD3_INT_CONTROL, tmp);
-	tmp = RREG32(DC_HPD4_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD4_INT_CONTROL, tmp);
-	tmp = RREG32(DC_HPD5_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD5_INT_CONTROL, tmp);
-	tmp = RREG32(DC_HPD6_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-	WREG32(DC_HPD6_INT_CONTROL, tmp);
-
+	for (i = 0; i < 6; i++)
+		WREG32_AND(DC_HPDx_INT_CONTROL(i), DC_HPDx_INT_POLARITY);
 }
 
+/* Note that the order we write back regs here is important */
 int evergreen_irq_set(struct radeon_device *rdev)
 {
+	int i;
 	u32 cp_int_cntl = CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE;
 	u32 cp_int_cntl1 = 0, cp_int_cntl2 = 0;
-	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
-	u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6;
 	u32 grbm_int_cntl = 0;
-	u32 afmt1 = 0, afmt2 = 0, afmt3 = 0, afmt4 = 0, afmt5 = 0, afmt6 = 0;
 	u32 dma_cntl, dma_cntl1 = 0;
 	u32 thermal_int = 0;
 
@@ -4643,12 +4530,6 @@ int evergreen_irq_set(struct radeon_device *rdev)
 		return 0;
 	}
 
-	hpd1 = RREG32(DC_HPD1_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	hpd2 = RREG32(DC_HPD2_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	hpd3 = RREG32(DC_HPD3_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	hpd4 = RREG32(DC_HPD4_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	hpd5 = RREG32(DC_HPD5_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	hpd6 = RREG32(DC_HPD6_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
 	if (rdev->family == CHIP_ARUBA)
 		thermal_int = RREG32(TN_CG_THERMAL_INT_CTRL) &
 			~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW);
@@ -4656,13 +4537,6 @@ int evergreen_irq_set(struct radeon_device *rdev)
 		thermal_int = RREG32(CG_THERMAL_INT) &
 			~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW);
 
-	afmt1 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-	afmt2 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-	afmt3 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-	afmt4 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-	afmt5 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-	afmt6 = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET) & ~AFMT_AZ_FORMAT_WTRIG_MASK;
-
 	dma_cntl = RREG32(DMA_CNTL) & ~TRAP_ENABLE;
 
 	if (rdev->family >= CHIP_CAYMAN) {
@@ -4705,85 +4579,6 @@ int evergreen_irq_set(struct radeon_device *rdev)
 		thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW;
 	}
 
-	if (rdev->irq.crtc_vblank_int[0] ||
-	    atomic_read(&rdev->irq.pflip[0])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 0\n");
-		crtc1 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[1] ||
-	    atomic_read(&rdev->irq.pflip[1])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 1\n");
-		crtc2 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[2] ||
-	    atomic_read(&rdev->irq.pflip[2])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 2\n");
-		crtc3 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[3] ||
-	    atomic_read(&rdev->irq.pflip[3])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 3\n");
-		crtc4 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[4] ||
-	    atomic_read(&rdev->irq.pflip[4])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 4\n");
-		crtc5 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[5] ||
-	    atomic_read(&rdev->irq.pflip[5])) {
-		DRM_DEBUG("evergreen_irq_set: vblank 5\n");
-		crtc6 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.hpd[0]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 1\n");
-		hpd1 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[1]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 2\n");
-		hpd2 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[2]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 3\n");
-		hpd3 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[3]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 4\n");
-		hpd4 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[4]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 5\n");
-		hpd5 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[5]) {
-		DRM_DEBUG("evergreen_irq_set: hpd 6\n");
-		hpd6 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.afmt[0]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 0\n");
-		afmt1 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-	if (rdev->irq.afmt[1]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 1\n");
-		afmt2 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-	if (rdev->irq.afmt[2]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 2\n");
-		afmt3 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-	if (rdev->irq.afmt[3]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 3\n");
-		afmt4 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-	if (rdev->irq.afmt[4]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 4\n");
-		afmt5 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-	if (rdev->irq.afmt[5]) {
-		DRM_DEBUG("evergreen_irq_set: hdmi 5\n");
-		afmt6 |= AFMT_AZ_FORMAT_WTRIG_MASK;
-	}
-
 	if (rdev->family >= CHIP_CAYMAN) {
 		cayman_cp_int_cntl_setup(rdev, 0, cp_int_cntl);
 		cayman_cp_int_cntl_setup(rdev, 1, cp_int_cntl1);
@@ -4798,51 +4593,35 @@ int evergreen_irq_set(struct radeon_device *rdev)
 
 	WREG32(GRBM_INT_CNTL, grbm_int_cntl);
 
-	WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, crtc1);
-	WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, crtc2);
-	if (rdev->num_crtc >= 4) {
-		WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, crtc3);
-		WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, crtc4);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, crtc5);
-		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
-	}
-
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
-	       GRPH_PFLIP_INT_MASK);
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
-	       GRPH_PFLIP_INT_MASK);
-	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-	}
-
-	WREG32(DC_HPD1_INT_CONTROL, hpd1);
-	WREG32(DC_HPD2_INT_CONTROL, hpd2);
-	WREG32(DC_HPD3_INT_CONTROL, hpd3);
-	WREG32(DC_HPD4_INT_CONTROL, hpd4);
-	WREG32(DC_HPD5_INT_CONTROL, hpd5);
-	WREG32(DC_HPD6_INT_CONTROL, hpd6);
+	for (i = 0; i < rdev->num_crtc; i++) {
+		radeon_irq_kms_set_irq_n_enabled(
+		    rdev, INT_MASK + crtc_offsets[i],
+		    VBLANK_INT_MASK,
+		    rdev->irq.crtc_vblank_int[i] ||
+		    atomic_read(&rdev->irq.pflip[i]), "vblank", i);
+	}
+
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(GRPH_INT_CONTROL + crtc_offsets[i], GRPH_PFLIP_INT_MASK);
+
+	for (i = 0; i < 6; i++) {
+		radeon_irq_kms_set_irq_n_enabled(
+		    rdev, DC_HPDx_INT_CONTROL(i),
+		    DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN,
+		    rdev->irq.hpd[i], "HPD", i);
+	}
+
 	if (rdev->family == CHIP_ARUBA)
 		WREG32(TN_CG_THERMAL_INT_CTRL, thermal_int);
 	else
 		WREG32(CG_THERMAL_INT, thermal_int);
 
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, afmt1);
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, afmt2);
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, afmt3);
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, afmt4);
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, afmt5);
-	WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, afmt6);
+	for (i = 0; i < 6; i++) {
+		radeon_irq_kms_set_irq_n_enabled(
+		    rdev, AFMT_AUDIO_PACKET_CONTROL + crtc_offsets[i],
+		    AFMT_AZ_FORMAT_WTRIG_MASK,
+		    rdev->irq.afmt[i], "HDMI", i);
+	}
 
 	/* posting read */
 	RREG32(SRBM_STATUS);
@@ -4850,168 +4629,53 @@ int evergreen_irq_set(struct radeon_device *rdev)
 	return 0;
 }
 
+/* Note that the order we write back regs here is important */
 static void evergreen_irq_ack(struct radeon_device *rdev)
 {
-	u32 tmp;
+	int i, j;
+	u32 *grph_int = rdev->irq.stat_regs.evergreen.grph_int;
+	u32 *disp_int = rdev->irq.stat_regs.evergreen.disp_int;
+	u32 *afmt_status = rdev->irq.stat_regs.evergreen.afmt_status;
 
-	rdev->irq.stat_regs.evergreen.disp_int = RREG32(DISP_INTERRUPT_STATUS);
-	rdev->irq.stat_regs.evergreen.disp_int_cont = RREG32(DISP_INTERRUPT_STATUS_CONTINUE);
-	rdev->irq.stat_regs.evergreen.disp_int_cont2 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE2);
-	rdev->irq.stat_regs.evergreen.disp_int_cont3 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE3);
-	rdev->irq.stat_regs.evergreen.disp_int_cont4 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE4);
-	rdev->irq.stat_regs.evergreen.disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5);
-	rdev->irq.stat_regs.evergreen.d1grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.d2grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET);
-	if (rdev->num_crtc >= 4) {
-		rdev->irq.stat_regs.evergreen.d3grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET);
-		rdev->irq.stat_regs.evergreen.d4grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET);
-	}
-	if (rdev->num_crtc >= 6) {
-		rdev->irq.stat_regs.evergreen.d5grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET);
-		rdev->irq.stat_regs.evergreen.d6grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET);
-	}
-
-	rdev->irq.stat_regs.evergreen.afmt_status1 = RREG32(AFMT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.afmt_status2 = RREG32(AFMT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.afmt_status3 = RREG32(AFMT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.afmt_status4 = RREG32(AFMT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.afmt_status5 = RREG32(AFMT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.afmt_status6 = RREG32(AFMT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET);
-
-	if (rdev->irq.stat_regs.evergreen.d1grph_int & GRPH_PFLIP_INT_OCCURRED)
-		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-	if (rdev->irq.stat_regs.evergreen.d2grph_int & GRPH_PFLIP_INT_OCCURRED)
-		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-	if (rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VBLANK_INTERRUPT)
-		WREG32(VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VLINE_INTERRUPT)
-		WREG32(VLINE_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VLINE_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VBLANK_INTERRUPT)
-		WREG32(VBLANK_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VBLANK_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VLINE_INTERRUPT)
-		WREG32(VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK);
-
-	if (rdev->num_crtc >= 4) {
-		if (rdev->irq.stat_regs.evergreen.d3grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.d4grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VLINE_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VLINE_ACK);
-	}
-
-	if (rdev->num_crtc >= 6) {
-		if (rdev->irq.stat_regs.evergreen.d5grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.d6grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VLINE_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VLINE_ACK);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_INTERRUPT) {
-		tmp = RREG32(DC_HPD1_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_INTERRUPT) {
-		tmp = RREG32(DC_HPD2_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_INTERRUPT) {
-		tmp = RREG32(DC_HPD3_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_INTERRUPT) {
-		tmp = RREG32(DC_HPD4_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_INTERRUPT) {
-		tmp = RREG32(DC_HPD5_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_INTERRUPT) {
-		tmp = RREG32(DC_HPD6_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD1_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD2_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD3_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD4_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD5_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD6_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.afmt_status1 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.afmt_status2 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.afmt_status3 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.afmt_status4 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.afmt_status5 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.afmt_status6 & AFMT_AZ_FORMAT_WTRIG) {
-		tmp = RREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET);
-		tmp |= AFMT_AZ_FORMAT_WTRIG_ACK;
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, tmp);
+	for (i = 0; i < 6; i++) {
+		disp_int[i] = RREG32(evergreen_disp_int_status[i]);
+		afmt_status[i] = RREG32(AFMT_STATUS + crtc_offsets[i]);
+		if (i < rdev->num_crtc)
+			grph_int[i] = RREG32(GRPH_INT_STATUS + crtc_offsets[i]);
+	}
+
+	/* We write back each interrupt register in pairs of two */
+	for (i = 0; i < rdev->num_crtc; i += 2) {
+		for (j = i; j < (i + 2); j++) {
+			if (grph_int[j] & GRPH_PFLIP_INT_OCCURRED)
+				WREG32(GRPH_INT_STATUS + crtc_offsets[j],
+				       GRPH_PFLIP_INT_CLEAR);
+		}
+
+		for (j = i; j < (i + 2); j++) {
+			if (disp_int[j] & LB_D1_VBLANK_INTERRUPT)
+				WREG32(VBLANK_STATUS + crtc_offsets[j],
+				       VBLANK_ACK);
+			if (disp_int[j] & LB_D1_VLINE_INTERRUPT)
+				WREG32(VLINE_STATUS + crtc_offsets[j],
+				       VLINE_ACK);
+		}
+	}
+
+	for (i = 0; i < 6; i++) {
+		if (disp_int[i] & DC_HPD1_INTERRUPT)
+			WREG32_OR(DC_HPDx_INT_CONTROL(i), DC_HPDx_INT_ACK);
+	}
+
+	for (i = 0; i < 6; i++) {
+		if (disp_int[i] & DC_HPD1_RX_INTERRUPT)
+			WREG32_OR(DC_HPDx_INT_CONTROL(i), DC_HPDx_RX_INT_ACK);
+	}
+
+	for (i = 0; i < 6; i++) {
+		if (afmt_status[i] & AFMT_AZ_FORMAT_WTRIG)
+			WREG32_OR(AFMT_AUDIO_PACKET_CONTROL + crtc_offsets[i],
+				  AFMT_AZ_FORMAT_WTRIG_ACK);
 	}
 }
 
@@ -5057,6 +4721,10 @@ static u32 evergreen_get_ih_wptr(struct radeon_device *rdev)
 
 int evergreen_irq_process(struct radeon_device *rdev)
 {
+	u32 *disp_int = rdev->irq.stat_regs.evergreen.disp_int;
+	u32 *afmt_status = rdev->irq.stat_regs.evergreen.afmt_status;
+	u32 crtc_idx, hpd_idx, afmt_idx;
+	u32 mask;
 	u32 wptr;
 	u32 rptr;
 	u32 src_id, src_data;
@@ -5066,6 +4734,7 @@ int evergreen_irq_process(struct radeon_device *rdev)
 	bool queue_dp = false;
 	bool queue_thermal = false;
 	u32 status, addr;
+	const char *event_name;
 
 	if (!rdev->ih.enabled || rdev->shutdown)
 		return IRQ_NONE;
@@ -5096,184 +4765,44 @@ restart_ih:
 
 		switch (src_id) {
 		case 1: /* D1 vblank/vline */
-			switch (src_data) {
-			case 0: /* D1 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D1 vblank - IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[0]) {
-					drm_handle_vblank(rdev->ddev, 0);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[0]))
-					radeon_crtc_handle_vblank(rdev, 0);
-				rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D1 vblank\n");
-
-				break;
-			case 1: /* D1 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D1 vline - IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D1 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 2: /* D2 vblank/vline */
-			switch (src_data) {
-			case 0: /* D2 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D2 vblank - IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[1]) {
-					drm_handle_vblank(rdev->ddev, 1);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[1]))
-					radeon_crtc_handle_vblank(rdev, 1);
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D2 vblank\n");
-
-				break;
-			case 1: /* D2 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D2 vline - IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D2 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 3: /* D3 vblank/vline */
-			switch (src_data) {
-			case 0: /* D3 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D3 vblank - IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[2]) {
-					drm_handle_vblank(rdev->ddev, 2);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[2]))
-					radeon_crtc_handle_vblank(rdev, 2);
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D3 vblank\n");
-
-				break;
-			case 1: /* D3 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D3 vline - IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D3 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 4: /* D4 vblank/vline */
-			switch (src_data) {
-			case 0: /* D4 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D4 vblank - IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[3]) {
-					drm_handle_vblank(rdev->ddev, 3);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[3]))
-					radeon_crtc_handle_vblank(rdev, 3);
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D4 vblank\n");
-
-				break;
-			case 1: /* D4 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D4 vline - IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D4 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 5: /* D5 vblank/vline */
-			switch (src_data) {
-			case 0: /* D5 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D5 vblank - IH event w/o asserted irq bit?\n");
+		case 6: /* D6 vblank/vline */
+			crtc_idx = src_id - 1;
+
+			if (src_data == 0) { /* vblank */
+				mask = LB_D1_VBLANK_INTERRUPT;
+				event_name = "vblank";
 
-				if (rdev->irq.crtc_vblank_int[4]) {
-					drm_handle_vblank(rdev->ddev, 4);
+				if (rdev->irq.crtc_vblank_int[crtc_idx]) {
+					drm_handle_vblank(rdev->ddev, crtc_idx);
 					rdev->pm.vblank_sync = true;
 					wake_up(&rdev->irq.vblank_queue);
 				}
-				if (atomic_read(&rdev->irq.pflip[4]))
-					radeon_crtc_handle_vblank(rdev, 4);
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D5 vblank\n");
-
-				break;
-			case 1: /* D5 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D5 vline - IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D5 vline\n");
+				if (atomic_read(&rdev->irq.pflip[crtc_idx])) {
+					radeon_crtc_handle_vblank(rdev,
+								  crtc_idx);
+				}
 
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
+			} else if (src_data == 1) { /* vline */
+				mask = LB_D1_VLINE_INTERRUPT;
+				event_name = "vline";
+			} else {
+				DRM_DEBUG("Unhandled interrupt: %d %d\n",
+					  src_id, src_data);
 				break;
 			}
-			break;
-		case 6: /* D6 vblank/vline */
-			switch (src_data) {
-			case 0: /* D6 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: D6 vblank - IH event w/o asserted irq bit?\n");
 
-				if (rdev->irq.crtc_vblank_int[5]) {
-					drm_handle_vblank(rdev->ddev, 5);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[5]))
-					radeon_crtc_handle_vblank(rdev, 5);
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D6 vblank\n");
-
-				break;
-			case 1: /* D6 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: D6 vline - IH event w/o asserted irq bit?\n");
+			if (!(disp_int[crtc_idx] & mask)) {
+				DRM_DEBUG("IH: D%d %s - IH event w/o asserted irq bit?\n",
+					  crtc_idx + 1, event_name);
+			}
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D6 vline\n");
+			disp_int[crtc_idx] &= ~mask;
+			DRM_DEBUG("IH: D%d %s\n", crtc_idx + 1, event_name);
 
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
 			break;
 		case 8: /* D1 page flip */
 		case 10: /* D2 page flip */
@@ -5286,162 +4815,45 @@ restart_ih:
 				radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
 			break;
 		case 42: /* HPD hotplug */
-			switch (src_data) {
-			case 0:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~DC_HPD1_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD1\n");
-				break;
-			case 1:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~DC_HPD2_INTERRUPT;
+			if (src_data <= 5) {
+				hpd_idx = src_data;
+				mask = DC_HPD1_INTERRUPT;
 				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD2\n");
-				break;
-			case 2:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~DC_HPD3_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD3\n");
-				break;
-			case 3:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+				event_name = "HPD";
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~DC_HPD4_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD4\n");
-				break;
-			case 4:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~DC_HPD5_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD5\n");
-				break;
-			case 5:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~DC_HPD6_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD6\n");
-				break;
-			case 6:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~DC_HPD1_RX_INTERRUPT;
+			} else if (src_data <= 11) {
+				hpd_idx = src_data - 6;
+				mask = DC_HPD1_RX_INTERRUPT;
 				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 1\n");
-				break;
-			case 7:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~DC_HPD2_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 2\n");
-				break;
-			case 8:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+				event_name = "HPD_RX";
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~DC_HPD3_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 3\n");
+			} else {
+				DRM_DEBUG("Unhandled interrupt: %d %d\n",
+					  src_id, src_data);
 				break;
-			case 9:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			}
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~DC_HPD4_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 4\n");
-				break;
-			case 10:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			if (!(disp_int[hpd_idx] & mask))
+				DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~DC_HPD5_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 5\n");
-				break;
-			case 11:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			disp_int[hpd_idx] &= ~mask;
+			DRM_DEBUG("IH: %s%d\n", event_name, hpd_idx + 1);
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~DC_HPD6_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 6\n");
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
 			break;
 		case 44: /* hdmi */
-			switch (src_data) {
-			case 0:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status1 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.afmt_status1 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI0\n");
-				break;
-			case 1:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status2 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.afmt_status2 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI1\n");
-				break;
-			case 2:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status3 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.afmt_status3 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI2\n");
-				break;
-			case 3:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status4 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.afmt_status4 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI3\n");
-				break;
-			case 4:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status5 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.afmt_status5 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI4\n");
-				break;
-			case 5:
-				if (!(rdev->irq.stat_regs.evergreen.afmt_status6 & AFMT_AZ_FORMAT_WTRIG))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			afmt_idx = src_data;
+			if (!(afmt_status[afmt_idx] & AFMT_AZ_FORMAT_WTRIG))
+				DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
 
-				rdev->irq.stat_regs.evergreen.afmt_status6 &= ~AFMT_AZ_FORMAT_WTRIG;
-				queue_hdmi = true;
-				DRM_DEBUG("IH: HDMI5\n");
-				break;
-			default:
-				DRM_ERROR("Unhandled interrupt: %d %d\n", src_id, src_data);
+			if (afmt_idx > 5) {
+				DRM_ERROR("Unhandled interrupt: %d %d\n",
+					  src_id, src_data);
 				break;
 			}
+			afmt_status[afmt_idx] &= ~AFMT_AZ_FORMAT_WTRIG;
+			queue_hdmi = true;
+			DRM_DEBUG("IH: HDMI%d\n", afmt_idx + 1);
+			break;
 		case 96:
 			DRM_ERROR("SRBM_READ_ERROR: 0x%x\n", RREG32(SRBM_READ_ERROR));
 			WREG32(SRBM_INT_ACK, 0x1);
@@ -5540,6 +4952,73 @@ restart_ih:
 	return IRQ_HANDLED;
 }
 
+static void evergreen_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails uvd_v2_2_resume() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void evergreen_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = uvd_v2_2_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void evergreen_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
 static int evergreen_startup(struct radeon_device *rdev)
 {
 	struct radeon_ring *ring;
@@ -5604,16 +5083,7 @@ static int evergreen_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = uvd_v2_2_resume(rdev);
-	if (!r) {
-		r = radeon_fence_driver_start_ring(rdev,
-						   R600_RING_TYPE_UVD_INDEX);
-		if (r)
-			dev_err(rdev->dev, "UVD fences init error (%d).\n", r);
-	}
-
-	if (r)
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+	evergreen_uvd_start(rdev);
 
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
@@ -5652,16 +5122,7 @@ static int evergreen_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-	if (ring->ring_size) {
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     RADEON_CP_PACKET2);
-		if (!r)
-			r = uvd_v1_0_init(rdev);
-
-		if (r)
-			DRM_ERROR("radeon: error initializing UVD (%d).\n", r);
-	}
+	evergreen_uvd_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -5716,8 +5177,10 @@ int evergreen_suspend(struct radeon_device *rdev)
 {
 	radeon_pm_suspend(rdev);
 	radeon_audio_fini(rdev);
-	uvd_v1_0_fini(rdev);
-	radeon_uvd_suspend(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
 	r700_cp_stop(rdev);
 	r600_dma_stop(rdev);
 	evergreen_irq_suspend(rdev);
@@ -5818,12 +5281,7 @@ int evergreen_init(struct radeon_device *rdev)
 	rdev->ring[R600_RING_TYPE_DMA_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX], 64 * 1024);
 
-	r = radeon_uvd_init(rdev);
-	if (!r) {
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
-		r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX],
-			       4096);
-	}
+	evergreen_uvd_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -5890,7 +5348,7 @@ void evergreen_fini(struct radeon_device *rdev)
 void evergreen_pcie_gen2_enable(struct radeon_device *rdev)
 {
 	u32 link_width_cntl, speed_cntl;
-	u32 mask;
+	enum pci_bus_speed max_bus_speed;
 
 	if (radeon_pcie_gen2 == 0)
 		return;
@@ -5905,10 +5363,9 @@ void evergreen_pcie_gen2_enable(struct radeon_device *rdev)
 	if (ASIC_IS_X2(rdev))
 		return;
 
-	if (drm_pcie_get_speed_cap_mask(rdev->ddev, &mask))
-		return;
-
-	if (!(mask & (DRM_PCIE_SPEED_50|DRM_PCIE_SPEED_80)))
+	max_bus_speed = pcie_get_speed_cap(rdev->pdev->bus->self);
+	if ((max_bus_speed != PCIE_SPEED_5_0GT) &&
+		(max_bus_speed != PCIE_SPEED_8_0GT))	
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
diff --git a/sys/dev/pci/drm/radeon/evergreen_blit_shaders.c b/sys/dev/pci/drm/radeon/evergreen_blit_shaders.c
index cf426879fbb..1a96ddb3e5e 100644
--- a/sys/dev/pci/drm/radeon/evergreen_blit_shaders.c
+++ b/sys/dev/pci/drm/radeon/evergreen_blit_shaders.c
@@ -24,7 +24,9 @@
  *     Alex Deucher <alexander.deucher@amd.com>
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/bug.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
 
 /*
  * evergreen cards need to use the 3D engine to blit data which requires
diff --git a/sys/dev/pci/drm/radeon/evergreen_cs.c b/sys/dev/pci/drm/radeon/evergreen_cs.c
index a48d9bd90aa..e08c9b4368e 100644
--- a/sys/dev/pci/drm/radeon/evergreen_cs.c
+++ b/sys/dev/pci/drm/radeon/evergreen_cs.c
@@ -25,8 +25,9 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "evergreend.h"
 #include "evergreen_reg_safe.h"
 #include "cayman_reg_safe.h"
@@ -1060,8 +1061,7 @@ static int evergreen_packet0_check(struct radeon_cs_parser *p,
 		}
 		break;
 	default:
-		printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n",
-		       reg, idx);
+		pr_err("Forbidden register 0x%04X in cs at %d\n", reg, idx);
 		return -EINVAL;
 	}
 	return 0;
@@ -1299,6 +1299,7 @@ static int evergreen_cs_handle_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
+		break;
 	case CB_TARGET_MASK:
 		track->cb_target_mask = radeon_get_ib_value(p, idx);
 		track->cb_dirty = true;
@@ -1816,8 +1817,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         (idx_value & 0xfffffff0) +
-		         ((u64)(tmp & 0xff) << 32);
+			 (idx_value & 0xfffffff0) +
+			 ((u64)(tmp & 0xff) << 32);
 
 		ib[idx + 0] = offset;
 		ib[idx + 1] = (tmp & 0xffffff00) | (upper_32_bits(offset) & 0xff);
@@ -1862,8 +1863,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         idx_value +
-		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
+			 idx_value +
+			 ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
 		ib[idx+0] = offset;
 		ib[idx+1] = upper_32_bits(offset) & 0xff;
@@ -1897,8 +1898,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         idx_value +
-		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
+			 idx_value +
+			 ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
 		ib[idx+0] = offset;
 		ib[idx+1] = upper_32_bits(offset) & 0xff;
@@ -1925,8 +1926,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         radeon_get_ib_value(p, idx+1) +
-		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+			 radeon_get_ib_value(p, idx+1) +
+			 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 		ib[idx+1] = offset;
 		ib[idx+2] = upper_32_bits(offset) & 0xff;
@@ -2098,8 +2099,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			}
 
 			offset = reloc->gpu_offset +
-			         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
-			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+				 (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
+				 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 			ib[idx+1] = (ib[idx+1] & 0x3) | (offset & 0xfffffffc);
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
@@ -2209,6 +2210,12 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 		break;
 	}
+	case PACKET3_PFP_SYNC_ME:
+		if (pkt->count) {
+			DRM_ERROR("bad PFP_SYNC_ME\n");
+			return -EINVAL;
+		}
+		break;
 	case PACKET3_SURFACE_SYNC:
 		if (pkt->count != 3) {
 			DRM_ERROR("bad SURFACE_SYNC\n");
@@ -2239,8 +2246,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				return -EINVAL;
 			}
 			offset = reloc->gpu_offset +
-			         (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
-			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+				 (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
+				 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 			ib[idx+1] = offset & 0xfffffff8;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
@@ -2261,8 +2268,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
-		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+			 (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
+			 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 		ib[idx+1] = offset & 0xfffffffc;
 		ib[idx+2] = (ib[idx+2] & 0xffffff00) | (upper_32_bits(offset) & 0xff);
@@ -2283,8 +2290,8 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
-		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+			 (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
+			 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 		ib[idx+1] = offset & 0xfffffffc;
 		ib[idx+2] = (ib[idx+2] & 0xffffff00) | (upper_32_bits(offset) & 0xff);
@@ -2608,6 +2615,51 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			}
 		}
 		break;
+	case PACKET3_SET_APPEND_CNT:
+	{
+		uint32_t areg;
+		uint32_t allowed_reg_base;
+		uint32_t source_sel;
+		if (pkt->count != 2) {
+			DRM_ERROR("bad SET_APPEND_CNT (invalid count)\n");
+			return -EINVAL;
+		}
+
+		allowed_reg_base = GDS_APPEND_COUNT_0;
+		allowed_reg_base -= PACKET3_SET_CONTEXT_REG_START;
+		allowed_reg_base >>= 2;
+
+		areg = idx_value >> 16;
+		if (areg < allowed_reg_base || areg > (allowed_reg_base + 11)) {
+			dev_warn(p->dev, "forbidden register for append cnt 0x%08x at %d\n",
+				 areg, idx);
+			return -EINVAL;
+		}
+
+		source_sel = G_PACKET3_SET_APPEND_CNT_SRC_SELECT(idx_value);
+		if (source_sel == PACKET3_SAC_SRC_SEL_MEM) {
+			uint64_t offset;
+			uint32_t swap;
+			r = radeon_cs_packet_next_reloc(p, &reloc, 0);
+			if (r) {
+				DRM_ERROR("bad SET_APPEND_CNT (missing reloc)\n");
+				return -EINVAL;
+			}
+			offset = radeon_get_ib_value(p, idx + 1);
+			swap = offset & 0x3;
+			offset &= ~0x3;
+
+			offset += ((u64)(radeon_get_ib_value(p, idx + 2) & 0xff)) << 32;
+
+			offset += reloc->gpu_offset;
+			ib[idx+1] = (offset & 0xfffffffc) | swap;
+			ib[idx+2] = upper_32_bits(offset) & 0xff;
+		} else {
+			DRM_ERROR("bad SET_APPEND_CNT (unsupported operation)\n");
+			return -EINVAL;
+		}
+		break;
+	}
 	case PACKET3_NOP:
 		break;
 	default:
@@ -2724,7 +2776,7 @@ int evergreen_cs_parse(struct radeon_cs_parser *p)
 	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib.length_dw; r++) {
-		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
+		pr_info("%05d  0x%08X\n", r, p->ib.ptr[r]);
 		mdelay(1);
 	}
 #endif
@@ -3163,7 +3215,7 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib->length_dw; r++) {
-		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
+		pr_info("%05d  0x%08X\n", r, p->ib.ptr[r]);
 		mdelay(1);
 	}
 #endif
@@ -3336,6 +3388,7 @@ static int evergreen_vm_packet3_check(struct radeon_device *rdev,
 	case PACKET3_MPEG_INDEX:
 	case PACKET3_WAIT_REG_MEM:
 	case PACKET3_MEM_WRITE:
+	case PACKET3_PFP_SYNC_ME:
 	case PACKET3_SURFACE_SYNC:
 	case PACKET3_EVENT_WRITE:
 	case PACKET3_EVENT_WRITE_EOP:
@@ -3438,6 +3491,27 @@ static int evergreen_vm_packet3_check(struct radeon_device *rdev,
 			}
 		}
 		break;
+	case PACKET3_SET_APPEND_CNT: {
+		uint32_t areg;
+		uint32_t allowed_reg_base;
+
+		if (pkt->count != 2) {
+			DRM_ERROR("bad SET_APPEND_CNT (invalid count)\n");
+			return -EINVAL;
+		}
+
+		allowed_reg_base = GDS_APPEND_COUNT_0;
+		allowed_reg_base -= PACKET3_SET_CONTEXT_REG_START;
+		allowed_reg_base >>= 2;
+
+		areg = idx_value >> 16;
+		if (areg < allowed_reg_base || areg > (allowed_reg_base + 11)) {
+			DRM_ERROR("forbidden register for append cnt 0x%08x at %d\n",
+				  areg, idx);
+			return -EINVAL;
+		}
+		break;
+	}
 	default:
 		return -EINVAL;
 	}
diff --git a/sys/dev/pci/drm/radeon/evergreen_dma.c b/sys/dev/pci/drm/radeon/evergreen_dma.c
index a28b426eb8a..96535aa8659 100644
--- a/sys/dev/pci/drm/radeon/evergreen_dma.c
+++ b/sys/dev/pci/drm/radeon/evergreen_dma.c
@@ -21,7 +21,7 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "evergreend.h"
diff --git a/sys/dev/pci/drm/radeon/evergreen_hdmi.c b/sys/dev/pci/drm/radeon/evergreen_hdmi.c
index 025c30122af..761f9b6deb8 100644
--- a/sys/dev/pci/drm/radeon/evergreen_hdmi.c
+++ b/sys/dev/pci/drm/radeon/evergreen_hdmi.c
@@ -24,9 +24,9 @@
  * Authors: Christian König
  *          Rafał Miłecki
  */
-#include <dev/pci/drm/linux_hdmi.h>
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/hdmi.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -206,7 +206,7 @@ void evergreen_hdmi_write_sad_regs(struct drm_encoder *encoder,
  * build a AVI Info Frame
  */
 void evergreen_set_avi_packet(struct radeon_device *rdev, u32 offset,
-    unsigned char *buffer, size_t size)
+			      unsigned char *buffer, size_t size)
 {
 	uint8_t *frame = buffer + 3;
 
diff --git a/sys/dev/pci/drm/radeon/evergreend.h b/sys/dev/pci/drm/radeon/evergreend.h
index 13b6029d65c..f3d88ca2aa8 100644
--- a/sys/dev/pci/drm/radeon/evergreend.h
+++ b/sys/dev/pci/drm/radeon/evergreend.h
@@ -1523,6 +1523,7 @@
 #define UVD_UDEC_ADDR_CONFIG				0xef4c
 #define UVD_UDEC_DB_ADDR_CONFIG				0xef50
 #define UVD_UDEC_DBW_ADDR_CONFIG			0xef54
+#define UVD_NO_OP					0xeffc
 #define UVD_RBC_RB_RPTR					0xf690
 #define UVD_RBC_RB_WPTR					0xf694
 #define UVD_STATUS					0xf6bc
@@ -1624,6 +1625,7 @@
 		 */
 #              define PACKET3_CP_DMA_CMD_SAIC      (1 << 28)
 #              define PACKET3_CP_DMA_CMD_DAIC      (1 << 29)
+#define	PACKET3_PFP_SYNC_ME				0x42
 #define	PACKET3_SURFACE_SYNC				0x43
 #              define PACKET3_CB0_DEST_BASE_ENA    (1 << 6)
 #              define PACKET3_CB1_DEST_BASE_ENA    (1 << 7)
@@ -1689,6 +1691,36 @@
 #define	PACKET3_SET_CONTEXT_REG_INDIRECT		0x73
 #define	PACKET3_SET_RESOURCE_INDIRECT			0x74
 #define	PACKET3_SET_APPEND_CNT			        0x75
+/* SET_APPEND_CNT - documentation
+ * 1. header
+ * 2. COMMAND
+ *  1:0 - SOURCE SEL
+ *  15:2 - Reserved
+ *  31:16 - WR_REG_OFFSET - context register to write source data to.
+ *          (one of R_02872C_GDS_APPEND_COUNT_0-11)
+ * 3. CONTROL
+ *  (for source == mem)
+ *  31:2 SRC_ADDRESS_LO
+ *  0:1 SWAP
+ *  (for source == GDS)
+ *  31:0 GDS offset
+ *  (for source == DATA)
+ *  31:0 DATA
+ *  (for source == REG)
+ *  31:0 REG
+ * 4. SRC_ADDRESS_HI[7:0]
+ * kernel driver 2.44 only supports SRC == MEM.
+ */
+#define PACKET3_SET_APPEND_CNT_SRC_SELECT(x) ((x) << 0)
+#define G_PACKET3_SET_APPEND_CNT_SRC_SELECT(x) ((x & 0x3) >> 0)
+/* source is from the data in CONTROL */
+#define PACKET3_SAC_SRC_SEL_DATA 0x0
+/* source is from register */
+#define PACKET3_SAC_SRC_SEL_REG  0x1
+/* source is from GDS offset in CONTROL */
+#define PACKET3_SAC_SRC_SEL_GDS  0x2
+/* source is from memory address */
+#define PACKET3_SAC_SRC_SEL_MEM  0x3
 
 #define	SQ_RESOURCE_CONSTANT_WORD7_0				0x3001c
 #define		S__SQ_CONSTANT_TYPE(x)			(((x) & 3) << 30)
@@ -2005,6 +2037,19 @@
 
 #define GDS_ADDR_BASE					0x28720
 
+#define GDS_APPEND_COUNT_0				0x2872C
+#define GDS_APPEND_COUNT_1				0x28730
+#define GDS_APPEND_COUNT_2				0x28734
+#define GDS_APPEND_COUNT_3				0x28738
+#define GDS_APPEND_COUNT_4				0x2873C
+#define GDS_APPEND_COUNT_5				0x28740
+#define GDS_APPEND_COUNT_6				0x28744
+#define GDS_APPEND_COUNT_7				0x28748
+#define GDS_APPEND_COUNT_8				0x2874c
+#define GDS_APPEND_COUNT_9				0x28750
+#define GDS_APPEND_COUNT_10				0x28754
+#define GDS_APPEND_COUNT_11				0x28758
+
 #define	CB_IMMED0_BASE					0x28b9c
 #define	CB_IMMED1_BASE					0x28ba0
 #define	CB_IMMED2_BASE					0x28ba4
diff --git a/sys/dev/pci/drm/radeon/kv_dpm.c b/sys/dev/pci/drm/radeon/kv_dpm.c
index b2b77d65076..f055d6ea352 100644
--- a/sys/dev/pci/drm/radeon/kv_dpm.c
+++ b/sys/dev/pci/drm/radeon/kv_dpm.c
@@ -21,12 +21,13 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "r600_dpm.h"
 #include "kv_dpm.h"
 #include "radeon_asic.h"
+#include <linux/seq_file.h>
 
 #define KV_MAX_DEEPSLEEP_DIVIDER_ID     5
 #define KV_MINIMUM_ENGINE_CLOCK         800
@@ -2163,7 +2164,7 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 	if (pi->caps_stable_p_state) {
 		stable_p_state_sclk = (max_limits->sclk * 75) / 100;
 
-		for (i = table->count - 1; i >= 0; i++) {
+		for (i = table->count - 1; i >= 0; i--) {
 			if (stable_p_state_sclk >= table->entries[i].clk) {
 				stable_p_state_sclk = table->entries[i].clk;
 				break;
@@ -2639,7 +2640,7 @@ static int kv_parse_power_table(struct radeon_device *rdev)
 	struct _NonClockInfoArray *non_clock_info_array;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 	struct kv_ps *ps;
@@ -2659,8 +2660,9 @@ static int kv_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -2737,7 +2739,7 @@ int kv_dpm_init(struct radeon_device *rdev)
 	for (i = 0; i < SUMO_MAX_HARDWARE_POWERLEVELS; i++)
 		pi->at[i] = TRINITY_AT_DFLT;
 
-        pi->sram_end = SMC_RAM_END;
+	pi->sram_end = SMC_RAM_END;
 
 	/* Enabling nb dpm on an asrock system prevents dpm from working */
 	if (rdev->pdev->subsystem_vendor == 0x1849)
diff --git a/sys/dev/pci/drm/radeon/kv_smc.c b/sys/dev/pci/drm/radeon/kv_smc.c
index 6077d426913..af60bd32a28 100644
--- a/sys/dev/pci/drm/radeon/kv_smc.c
+++ b/sys/dev/pci/drm/radeon/kv_smc.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "kv_dpm.h"
diff --git a/sys/dev/pci/drm/radeon/mkregtable.c b/sys/dev/pci/drm/radeon/mkregtable.c
index a7ecfba3ebd..52a7246fed9 100644
--- a/sys/dev/pci/drm/radeon/mkregtable.c
+++ b/sys/dev/pci/drm/radeon/mkregtable.c
@@ -43,10 +43,6 @@ struct list_head {
 	struct list_head *next, *prev;
 };
 
-#define LIST_HEAD_INIT(name) { &(name), &(name) }
-
-#define LIST_HEAD(name) \
-	struct list_head name = LIST_HEAD_INIT(name)
 
 static inline void INIT_LIST_HEAD(struct list_head *list)
 {
@@ -75,19 +71,6 @@ extern void __list_add(struct list_head *new,
 #endif
 
 /**
- * list_add - add a new entry
- * @new: new entry to be added
- * @head: list head to add it after
- *
- * Insert a new entry after the specified head.
- * This is good for implementing stacks.
- */
-static inline void list_add(struct list_head *new, struct list_head *head)
-{
-	__list_add(new, head, head->next);
-}
-
-/**
  * list_add_tail - add a new entry
  * @new: new entry to be added
  * @head: list head to add it before
@@ -100,250 +83,6 @@ static inline void list_add_tail(struct list_head *new, struct list_head *head)
 	__list_add(new, head->prev, head);
 }
 
-/*
- * Delete a list entry by making the prev/next entries
- * point to each other.
- *
- * This is only for internal list manipulation where we know
- * the prev/next entries already!
- */
-static inline void __list_del(struct list_head *prev, struct list_head *next)
-{
-	next->prev = prev;
-	prev->next = next;
-}
-
-/**
- * list_del - deletes entry from list.
- * @entry: the element to delete from the list.
- * Note: list_empty() on entry does not return true after this, the entry is
- * in an undefined state.
- */
-#ifndef CONFIG_DEBUG_LIST
-static inline void list_del(struct list_head *entry)
-{
-	__list_del(entry->prev, entry->next);
-	entry->next = (void *)0xDEADBEEF;
-	entry->prev = (void *)0xBEEFDEAD;
-}
-#else
-extern void list_del(struct list_head *entry);
-#endif
-
-/**
- * list_replace - replace old entry by new one
- * @old : the element to be replaced
- * @new : the new element to insert
- *
- * If @old was empty, it will be overwritten.
- */
-static inline void list_replace(struct list_head *old, struct list_head *new)
-{
-	new->next = old->next;
-	new->next->prev = new;
-	new->prev = old->prev;
-	new->prev->next = new;
-}
-
-static inline void list_replace_init(struct list_head *old,
-				     struct list_head *new)
-{
-	list_replace(old, new);
-	INIT_LIST_HEAD(old);
-}
-
-/**
- * list_del_init - deletes entry from list and reinitialize it.
- * @entry: the element to delete from the list.
- */
-static inline void list_del_init(struct list_head *entry)
-{
-	__list_del(entry->prev, entry->next);
-	INIT_LIST_HEAD(entry);
-}
-
-/**
- * list_move - delete from one list and add as another's head
- * @list: the entry to move
- * @head: the head that will precede our entry
- */
-static inline void list_move(struct list_head *list, struct list_head *head)
-{
-	__list_del(list->prev, list->next);
-	list_add(list, head);
-}
-
-/**
- * list_move_tail - delete from one list and add as another's tail
- * @list: the entry to move
- * @head: the head that will follow our entry
- */
-static inline void list_move_tail(struct list_head *list,
-				  struct list_head *head)
-{
-	__list_del(list->prev, list->next);
-	list_add_tail(list, head);
-}
-
-/**
- * list_is_last - tests whether @list is the last entry in list @head
- * @list: the entry to test
- * @head: the head of the list
- */
-static inline int list_is_last(const struct list_head *list,
-			       const struct list_head *head)
-{
-	return list->next == head;
-}
-
-/**
- * list_empty - tests whether a list is empty
- * @head: the list to test.
- */
-static inline int list_empty(const struct list_head *head)
-{
-	return head->next == head;
-}
-
-/**
- * list_empty_careful - tests whether a list is empty and not being modified
- * @head: the list to test
- *
- * Description:
- * tests whether a list is empty _and_ checks that no other CPU might be
- * in the process of modifying either member (next or prev)
- *
- * NOTE: using list_empty_careful() without synchronization
- * can only be safe if the only activity that can happen
- * to the list entry is list_del_init(). Eg. it cannot be used
- * if another CPU could re-list_add() it.
- */
-static inline int list_empty_careful(const struct list_head *head)
-{
-	struct list_head *next = head->next;
-	return (next == head) && (next == head->prev);
-}
-
-/**
- * list_is_singular - tests whether a list has just one entry.
- * @head: the list to test.
- */
-static inline int list_is_singular(const struct list_head *head)
-{
-	return !list_empty(head) && (head->next == head->prev);
-}
-
-static inline void __list_cut_position(struct list_head *list,
-				       struct list_head *head,
-				       struct list_head *entry)
-{
-	struct list_head *new_first = entry->next;
-	list->next = head->next;
-	list->next->prev = list;
-	list->prev = entry;
-	entry->next = list;
-	head->next = new_first;
-	new_first->prev = head;
-}
-
-/**
- * list_cut_position - cut a list into two
- * @list: a new list to add all removed entries
- * @head: a list with entries
- * @entry: an entry within head, could be the head itself
- *	and if so we won't cut the list
- *
- * This helper moves the initial part of @head, up to and
- * including @entry, from @head to @list. You should
- * pass on @entry an element you know is on @head. @list
- * should be an empty list or a list you do not care about
- * losing its data.
- *
- */
-static inline void list_cut_position(struct list_head *list,
-				     struct list_head *head,
-				     struct list_head *entry)
-{
-	if (list_empty(head))
-		return;
-	if (list_is_singular(head) && (head->next != entry && head != entry))
-		return;
-	if (entry == head)
-		INIT_LIST_HEAD(list);
-	else
-		__list_cut_position(list, head, entry);
-}
-
-static inline void __list_splice(const struct list_head *list,
-				 struct list_head *prev, struct list_head *next)
-{
-	struct list_head *first = list->next;
-	struct list_head *last = list->prev;
-
-	first->prev = prev;
-	prev->next = first;
-
-	last->next = next;
-	next->prev = last;
-}
-
-/**
- * list_splice - join two lists, this is designed for stacks
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- */
-static inline void list_splice(const struct list_head *list,
-			       struct list_head *head)
-{
-	if (!list_empty(list))
-		__list_splice(list, head, head->next);
-}
-
-/**
- * list_splice_tail - join two lists, each list being a queue
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- */
-static inline void list_splice_tail(struct list_head *list,
-				    struct list_head *head)
-{
-	if (!list_empty(list))
-		__list_splice(list, head->prev, head);
-}
-
-/**
- * list_splice_init - join two lists and reinitialise the emptied list.
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- *
- * The list at @list is reinitialised
- */
-static inline void list_splice_init(struct list_head *list,
-				    struct list_head *head)
-{
-	if (!list_empty(list)) {
-		__list_splice(list, head, head->next);
-		INIT_LIST_HEAD(list);
-	}
-}
-
-/**
- * list_splice_tail_init - join two lists and reinitialise the emptied list
- * @list: the new list to add.
- * @head: the place to add it in the first list.
- *
- * Each of the lists is a queue.
- * The list at @list is reinitialised
- */
-static inline void list_splice_tail_init(struct list_head *list,
-					 struct list_head *head)
-{
-	if (!list_empty(list)) {
-		__list_splice(list, head->prev, head);
-		INIT_LIST_HEAD(list);
-	}
-}
-
 /**
  * list_entry - get the struct for this entry
  * @ptr:	the &struct list_head pointer.
@@ -354,56 +93,6 @@ static inline void list_splice_tail_init(struct list_head *list,
 	container_of(ptr, type, member)
 
 /**
- * list_first_entry - get the first element from a list
- * @ptr:	the list head to take the element from.
- * @type:	the type of the struct this is embedded in.
- * @member:	the name of the list_head within the struct.
- *
- * Note, that list is expected to be not empty.
- */
-#define list_first_entry(ptr, type, member) \
-	list_entry((ptr)->next, type, member)
-
-/**
- * list_for_each	-	iterate over a list
- * @pos:	the &struct list_head to use as a loop cursor.
- * @head:	the head for your list.
- */
-#define list_for_each(pos, head) \
-	for (pos = (head)->next; prefetch(pos->next), pos != (head); \
-		pos = pos->next)
-
-/**
- * list_for_each_prev	-	iterate over a list backwards
- * @pos:	the &struct list_head to use as a loop cursor.
- * @head:	the head for your list.
- */
-#define list_for_each_prev(pos, head) \
-	for (pos = (head)->prev; prefetch(pos->prev), pos != (head); \
-		pos = pos->prev)
-
-/**
- * list_for_each_safe - iterate over a list safe against removal of list entry
- * @pos:	the &struct list_head to use as a loop cursor.
- * @n:		another &struct list_head to use as temporary storage
- * @head:	the head for your list.
- */
-#define list_for_each_safe(pos, n, head) \
-	for (pos = (head)->next, n = pos->next; pos != (head); \
-		pos = n, n = pos->next)
-
-/**
- * list_for_each_prev_safe - iterate over a list backwards safe against removal of list entry
- * @pos:	the &struct list_head to use as a loop cursor.
- * @n:		another &struct list_head to use as temporary storage
- * @head:	the head for your list.
- */
-#define list_for_each_prev_safe(pos, n, head) \
-	for (pos = (head)->prev, n = pos->prev; \
-	     prefetch(pos->prev), pos != (head); \
-	     pos = n, n = pos->prev)
-
-/**
  * list_for_each_entry	-	iterate over list of given type
  * @pos:	the type * to use as a loop cursor.
  * @head:	the head for your list.
@@ -414,128 +103,6 @@ static inline void list_splice_tail_init(struct list_head *list,
 	     &pos->member != (head); 	\
 	     pos = list_entry(pos->member.next, typeof(*pos), member))
 
-/**
- * list_for_each_entry_reverse - iterate backwards over list of given type.
- * @pos:	the type * to use as a loop cursor.
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- */
-#define list_for_each_entry_reverse(pos, head, member)			\
-	for (pos = list_entry((head)->prev, typeof(*pos), member);	\
-	     prefetch(pos->member.prev), &pos->member != (head); 	\
-	     pos = list_entry(pos->member.prev, typeof(*pos), member))
-
-/**
- * list_prepare_entry - prepare a pos entry for use in list_for_each_entry_continue()
- * @pos:	the type * to use as a start point
- * @head:	the head of the list
- * @member:	the name of the list_head within the struct.
- *
- * Prepares a pos entry for use as a start point in list_for_each_entry_continue().
- */
-#define list_prepare_entry(pos, head, member) \
-	((pos) ? : list_entry(head, typeof(*pos), member))
-
-/**
- * list_for_each_entry_continue - continue iteration over list of given type
- * @pos:	the type * to use as a loop cursor.
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Continue to iterate over list of given type, continuing after
- * the current position.
- */
-#define list_for_each_entry_continue(pos, head, member) 		\
-	for (pos = list_entry(pos->member.next, typeof(*pos), member);	\
-	     prefetch(pos->member.next), &pos->member != (head);	\
-	     pos = list_entry(pos->member.next, typeof(*pos), member))
-
-/**
- * list_for_each_entry_continue_reverse - iterate backwards from the given point
- * @pos:	the type * to use as a loop cursor.
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Start to iterate over list of given type backwards, continuing after
- * the current position.
- */
-#define list_for_each_entry_continue_reverse(pos, head, member)		\
-	for (pos = list_entry(pos->member.prev, typeof(*pos), member);	\
-	     prefetch(pos->member.prev), &pos->member != (head);	\
-	     pos = list_entry(pos->member.prev, typeof(*pos), member))
-
-/**
- * list_for_each_entry_from - iterate over list of given type from the current point
- * @pos:	the type * to use as a loop cursor.
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Iterate over list of given type, continuing from current position.
- */
-#define list_for_each_entry_from(pos, head, member) 			\
-	for (; prefetch(pos->member.next), &pos->member != (head);	\
-	     pos = list_entry(pos->member.next, typeof(*pos), member))
-
-/**
- * list_for_each_entry_safe - iterate over list of given type safe against removal of list entry
- * @pos:	the type * to use as a loop cursor.
- * @n:		another type * to use as temporary storage
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- */
-#define list_for_each_entry_safe(pos, n, head, member)			\
-	for (pos = list_entry((head)->next, typeof(*pos), member),	\
-		n = list_entry(pos->member.next, typeof(*pos), member);	\
-	     &pos->member != (head); 					\
-	     pos = n, n = list_entry(n->member.next, typeof(*n), member))
-
-/**
- * list_for_each_entry_safe_continue
- * @pos:	the type * to use as a loop cursor.
- * @n:		another type * to use as temporary storage
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Iterate over list of given type, continuing after current point,
- * safe against removal of list entry.
- */
-#define list_for_each_entry_safe_continue(pos, n, head, member) 		\
-	for (pos = list_entry(pos->member.next, typeof(*pos), member), 		\
-		n = list_entry(pos->member.next, typeof(*pos), member);		\
-	     &pos->member != (head);						\
-	     pos = n, n = list_entry(n->member.next, typeof(*n), member))
-
-/**
- * list_for_each_entry_safe_from
- * @pos:	the type * to use as a loop cursor.
- * @n:		another type * to use as temporary storage
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Iterate over list of given type from current point, safe against
- * removal of list entry.
- */
-#define list_for_each_entry_safe_from(pos, n, head, member) 			\
-	for (n = list_entry(pos->member.next, typeof(*pos), member);		\
-	     &pos->member != (head);						\
-	     pos = n, n = list_entry(n->member.next, typeof(*n), member))
-
-/**
- * list_for_each_entry_safe_reverse
- * @pos:	the type * to use as a loop cursor.
- * @n:		another type * to use as temporary storage
- * @head:	the head for your list.
- * @member:	the name of the list_head within the struct.
- *
- * Iterate backwards over list of given type, safe against removal
- * of list entry.
- */
-#define list_for_each_entry_safe_reverse(pos, n, head, member)		\
-	for (pos = list_entry((head)->prev, typeof(*pos), member),	\
-		n = list_entry(pos->member.prev, typeof(*pos), member);	\
-	     &pos->member != (head); 					\
-	     pos = n, n = list_entry(n->member.prev, typeof(*n), member))
-
 struct offset {
 	struct list_head list;
 	unsigned offset;
diff --git a/sys/dev/pci/drm/radeon/ni.c b/sys/dev/pci/drm/radeon/ni.c
index af18279f912..5bb87e8d385 100644
--- a/sys/dev/pci/drm/radeon/ni.c
+++ b/sys/dev/pci/drm/radeon/ni.c
@@ -21,11 +21,14 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "nid.h"
 #include "atom.h"
 #include "ni_reg.h"
@@ -771,8 +774,7 @@ int ni_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->pfp_fw->size != pfp_req_size) {
-		printk(KERN_ERR
-		       "ni_cp: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("ni_cp: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->pfp_fw->size, fw_name);
 		err = -EINVAL;
 		goto out;
@@ -783,8 +785,7 @@ int ni_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->me_fw->size != me_req_size) {
-		printk(KERN_ERR
-		       "ni_cp: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("ni_cp: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->me_fw->size, fw_name);
 		err = -EINVAL;
 	}
@@ -794,8 +795,7 @@ int ni_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->rlc_fw->size != rlc_req_size) {
-		printk(KERN_ERR
-		       "ni_rlc: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("ni_rlc: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->rlc_fw->size, fw_name);
 		err = -EINVAL;
 	}
@@ -807,8 +807,7 @@ int ni_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->mc_fw->size != mc_req_size) {
-			printk(KERN_ERR
-			       "ni_mc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("ni_mc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->mc_fw->size, fw_name);
 			err = -EINVAL;
 		}
@@ -818,15 +817,12 @@ int ni_init_microcode(struct radeon_device *rdev)
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
 		err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 		if (err) {
-			printk(KERN_ERR
-			       "smc: error loading firmware \"%s\"\n",
-			       fw_name);
+			pr_err("smc: error loading firmware \"%s\"\n", fw_name);
 			release_firmware(rdev->smc_fw);
 			rdev->smc_fw = NULL;
 			err = 0;
 		} else if (rdev->smc_fw->size != smc_req_size) {
-			printk(KERN_ERR
-			       "ni_mc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("ni_mc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->mc_fw->size, fw_name);
 			err = -EINVAL;
 		}
@@ -835,8 +831,7 @@ int ni_init_microcode(struct radeon_device *rdev)
 out:
 	if (err) {
 		if (err != -EINVAL)
-			printk(KERN_ERR
-			       "ni_cp: Failed to load firmware \"%s\"\n",
+			pr_err("ni_cp: Failed to load firmware \"%s\"\n",
 			       fw_name);
 		release_firmware(rdev->pfp_fw);
 		rdev->pfp_fw = NULL;
@@ -1153,6 +1148,7 @@ static void cayman_gpu_init(struct radeon_device *rdev)
 						rdev->config.cayman.max_shader_engines,
 						CAYMAN_MAX_BACKENDS, disabled_rb_mask);
 	}
+	rdev->config.cayman.backend_map = tmp;
 	WREG32(GB_BACKEND_MAP, tmp);
 
 	cgts_tcc_disable = 0xffff0000;
@@ -1254,7 +1250,7 @@ static void cayman_gpu_init(struct radeon_device *rdev)
 		tmp = RREG32_CG(CG_CGTT_LOCAL_0);
 		tmp &= ~0x00380000;
 		WREG32_CG(CG_CGTT_LOCAL_0, tmp);
-                tmp = RREG32_CG(CG_CGTT_LOCAL_1);
+		tmp = RREG32_CG(CG_CGTT_LOCAL_1);
 		tmp &= ~0x0e000000;
 		WREG32_CG(CG_CGTT_LOCAL_1, tmp);
 	}
@@ -1954,10 +1950,15 @@ static void cayman_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	evergreen_print_gpu_status_regs(rdev);
 }
 
-int cayman_asic_reset(struct radeon_device *rdev)
+int cayman_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	u32 reset_mask;
 
+	if (hard) {
+		evergreen_gpu_pci_config_reset(rdev);
+		return 0;
+	}
+
 	reset_mask = cayman_gpu_check_soft_reset(rdev);
 
 	if (reset_mask)
@@ -1997,6 +1998,160 @@ bool cayman_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
+static void cayman_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails uvd_v2_2_resume() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void cayman_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = uvd_v2_2_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void cayman_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
+static void cayman_vce_init(struct radeon_device *rdev)
+{
+	int r;
+
+	/* Only set for CHIP_ARUBA */
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE (%d) init.\n", r);
+		/*
+		 * At this point rdev->vce.vcpu_bo is NULL which trickles down
+		 * to early fails cayman_vce_start() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable vce here.
+		 */
+		rdev->has_vce = 0;
+		return;
+	}
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE1_INDEX], 4096);
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE2_INDEX], 4096);
+}
+
+static void cayman_vce_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = vce_v1_0_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE1_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 fences (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE2_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE2 fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
+}
+
+static void cayman_vce_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_vce || !rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, 0x0);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, 0x0);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	r = vce_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE (%d).\n", r);
+		return;
+	}
+}
+
 static int cayman_startup(struct radeon_device *rdev)
 {
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
@@ -2051,34 +2206,8 @@ static int cayman_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = uvd_v2_2_resume(rdev);
-	if (!r) {
-		r = radeon_fence_driver_start_ring(rdev,
-						   R600_RING_TYPE_UVD_INDEX);
-		if (r)
-			dev_err(rdev->dev, "UVD fences init error (%d).\n", r);
-	}
-	if (r)
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
-
-	if (rdev->family == CHIP_ARUBA) {
-		r = radeon_vce_resume(rdev);
-		if (!r)
-			r = vce_v1_0_resume(rdev);
-
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE1_INDEX);
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE2_INDEX);
-
-		if (r) {
-			dev_err(rdev->dev, "VCE init error (%d).\n", r);
-			rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
-			rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
-		}
-	}
+	cayman_uvd_start(rdev);
+	cayman_vce_start(rdev);
 
 	r = radeon_fence_driver_start_ring(rdev, CAYMAN_RING_TYPE_CP1_INDEX);
 	if (r) {
@@ -2147,30 +2276,8 @@ static int cayman_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-	if (ring->ring_size) {
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     RADEON_CP_PACKET2);
-		if (!r)
-			r = uvd_v1_0_init(rdev);
-		if (r)
-			DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
-	}
-
-	if (rdev->family == CHIP_ARUBA) {
-		ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-		if (ring->ring_size)
-			r = radeon_ring_init(rdev, ring, ring->ring_size, 0, 0x0);
-
-		ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-		if (ring->ring_size)
-			r = radeon_ring_init(rdev, ring, ring->ring_size, 0, 0x0);
-
-		if (!r)
-			r = vce_v1_0_init(rdev);
-		if (r)
-			DRM_ERROR("radeon: failed initializing VCE (%d).\n", r);
-	}
+	cayman_uvd_resume(rdev);
+	cayman_vce_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -2225,8 +2332,10 @@ int cayman_suspend(struct radeon_device *rdev)
 	radeon_vm_manager_fini(rdev);
 	cayman_cp_enable(rdev, false);
 	cayman_dma_stop(rdev);
-	uvd_v1_0_fini(rdev);
-	radeon_uvd_suspend(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
 	evergreen_irq_suspend(rdev);
 	radeon_wb_disable(rdev);
 	cayman_pcie_gart_disable(rdev);
@@ -2320,25 +2429,8 @@ int cayman_init(struct radeon_device *rdev)
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 64 * 1024);
 
-	r = radeon_uvd_init(rdev);
-	if (!r) {
-		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-	}
-
-	if (rdev->family == CHIP_ARUBA) {
-		r = radeon_vce_init(rdev);
-		if (!r) {
-			ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-			ring->ring_obj = NULL;
-			r600_ring_init(rdev, ring, 4096);
-
-			ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-			ring->ring_obj = NULL;
-			r600_ring_init(rdev, ring, 4096);
-		}
-	}
+	cayman_uvd_init(rdev);
+	cayman_vce_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -2393,7 +2485,7 @@ void cayman_fini(struct radeon_device *rdev)
 	radeon_irq_kms_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_fini(rdev);
-	if (rdev->family == CHIP_ARUBA)
+	if (rdev->has_vce)
 		radeon_vce_fini(rdev);
 	cayman_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
@@ -2629,7 +2721,7 @@ int tn_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk)
 	struct atom_clock_dividers dividers;
 	int r, i;
 
-        r = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+	r = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
 					   ecclk, false, &dividers);
 	if (r)
 		return r;
diff --git a/sys/dev/pci/drm/radeon/ni_dma.c b/sys/dev/pci/drm/radeon/ni_dma.c
index 146656689a2..ce787a9f12c 100644
--- a/sys/dev/pci/drm/radeon/ni_dma.c
+++ b/sys/dev/pci/drm/radeon/ni_dma.c
@@ -21,7 +21,7 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_trace.h"
diff --git a/sys/dev/pci/drm/radeon/ni_dpm.c b/sys/dev/pci/drm/radeon/ni_dpm.c
index b30db1be6d0..0fd8d6ba982 100644
--- a/sys/dev/pci/drm/radeon/ni_dpm.c
+++ b/sys/dev/pci/drm/radeon/ni_dpm.c
@@ -21,13 +21,15 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "nid.h"
 #include "r600_dpm.h"
 #include "ni_dpm.h"
 #include "atom.h"
+#include <linux/math64.h>
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -723,9 +725,9 @@ extern int ni_mc_load_microcode(struct radeon_device *rdev);
 
 struct ni_power_info *ni_get_pi(struct radeon_device *rdev)
 {
-        struct ni_power_info *pi = rdev->pm.dpm.priv;
+	struct ni_power_info *pi = rdev->pm.dpm.priv;
 
-        return pi;
+	return pi;
 }
 
 struct ni_ps *ni_get_ps(struct radeon_ps *rps)
@@ -1094,9 +1096,9 @@ static void ni_stop_smc(struct radeon_device *rdev)
 
 static int ni_process_firmware_header(struct radeon_device *rdev)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
-        struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	u32 tmp;
 	int ret;
 
@@ -1200,14 +1202,14 @@ static int ni_enter_ulp_state(struct radeon_device *rdev)
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 
 	if (pi->gfx_clock_gating) {
-                WREG32_P(SCLK_PWRMGT_CNTL, 0, ~DYN_GFX_CLK_OFF_EN);
+		WREG32_P(SCLK_PWRMGT_CNTL, 0, ~DYN_GFX_CLK_OFF_EN);
 		WREG32_P(SCLK_PWRMGT_CNTL, GFX_CLK_FORCE_ON, ~GFX_CLK_FORCE_ON);
-                WREG32_P(SCLK_PWRMGT_CNTL, 0, ~GFX_CLK_FORCE_ON);
+		WREG32_P(SCLK_PWRMGT_CNTL, 0, ~GFX_CLK_FORCE_ON);
 		RREG32(GB_ADDR_CONFIG);
-        }
+	}
 
 	WREG32_P(SMC_MSG, HOST_SMC_MSG(PPSMC_MSG_SwitchToMinimumPower),
-                 ~HOST_SMC_MSG_MASK);
+		 ~HOST_SMC_MSG_MASK);
 
 	udelay(25000);
 
@@ -1319,12 +1321,12 @@ static void ni_populate_mvdd_value(struct radeon_device *rdev,
 				   u32 mclk,
 				   NISLANDS_SMC_VOLTAGE_VALUE *voltage)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 
 	if (!pi->mvdd_control) {
 		voltage->index = eg_pi->mvdd_high_index;
-                voltage->value = cpu_to_be16(MVDD_HIGH_VALUE);
+		voltage->value = cpu_to_be16(MVDD_HIGH_VALUE);
 		return;
 	}
 
@@ -1508,47 +1510,47 @@ int ni_copy_and_switch_arb_sets(struct radeon_device *rdev,
 	u32 mc_cg_config;
 
 	switch (arb_freq_src) {
-        case MC_CG_ARB_FREQ_F0:
+	case MC_CG_ARB_FREQ_F0:
 		mc_arb_dram_timing  = RREG32(MC_ARB_DRAM_TIMING);
 		mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
 		burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE0_MASK) >> STATE0_SHIFT;
 		break;
-        case MC_CG_ARB_FREQ_F1:
+	case MC_CG_ARB_FREQ_F1:
 		mc_arb_dram_timing  = RREG32(MC_ARB_DRAM_TIMING_1);
 		mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_1);
 		burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE1_MASK) >> STATE1_SHIFT;
 		break;
-        case MC_CG_ARB_FREQ_F2:
+	case MC_CG_ARB_FREQ_F2:
 		mc_arb_dram_timing  = RREG32(MC_ARB_DRAM_TIMING_2);
 		mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_2);
 		burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE2_MASK) >> STATE2_SHIFT;
 		break;
-        case MC_CG_ARB_FREQ_F3:
+	case MC_CG_ARB_FREQ_F3:
 		mc_arb_dram_timing  = RREG32(MC_ARB_DRAM_TIMING_3);
 		mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_3);
 		burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE3_MASK) >> STATE3_SHIFT;
 		break;
-        default:
+	default:
 		return -EINVAL;
 	}
 
 	switch (arb_freq_dest) {
-        case MC_CG_ARB_FREQ_F0:
+	case MC_CG_ARB_FREQ_F0:
 		WREG32(MC_ARB_DRAM_TIMING, mc_arb_dram_timing);
 		WREG32(MC_ARB_DRAM_TIMING2, mc_arb_dram_timing2);
 		WREG32_P(MC_ARB_BURST_TIME, STATE0(burst_time), ~STATE0_MASK);
 		break;
-        case MC_CG_ARB_FREQ_F1:
+	case MC_CG_ARB_FREQ_F1:
 		WREG32(MC_ARB_DRAM_TIMING_1, mc_arb_dram_timing);
 		WREG32(MC_ARB_DRAM_TIMING2_1, mc_arb_dram_timing2);
 		WREG32_P(MC_ARB_BURST_TIME, STATE1(burst_time), ~STATE1_MASK);
 		break;
-        case MC_CG_ARB_FREQ_F2:
+	case MC_CG_ARB_FREQ_F2:
 		WREG32(MC_ARB_DRAM_TIMING_2, mc_arb_dram_timing);
 		WREG32(MC_ARB_DRAM_TIMING2_2, mc_arb_dram_timing2);
 		WREG32_P(MC_ARB_BURST_TIME, STATE2(burst_time), ~STATE2_MASK);
 		break;
-        case MC_CG_ARB_FREQ_F3:
+	case MC_CG_ARB_FREQ_F3:
 		WREG32(MC_ARB_DRAM_TIMING_3, mc_arb_dram_timing);
 		WREG32(MC_ARB_DRAM_TIMING2_3, mc_arb_dram_timing2);
 		WREG32_P(MC_ARB_BURST_TIME, STATE3(burst_time), ~STATE3_MASK);
@@ -1619,9 +1621,7 @@ static int ni_populate_memory_timing_parameters(struct radeon_device *rdev,
 		(u8)rv770_calculate_memory_refresh_rate(rdev, pl->sclk);
 
 
-	radeon_atom_set_engine_dram_timings(rdev,
-                                            pl->sclk,
-                                            pl->mclk);
+	radeon_atom_set_engine_dram_timings(rdev, pl->sclk, pl->mclk);
 
 	dram_timing = RREG32(MC_ARB_DRAM_TIMING);
 	dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
@@ -1865,9 +1865,9 @@ static int ni_populate_smc_acpi_state(struct radeon_device *rdev,
 
 	mpll_ad_func_cntl_2 |= BIAS_GEN_PDNB | RESET_EN;
 
-        if (pi->mem_gddr5)
-                mpll_dq_func_cntl &= ~PDNB;
-        mpll_dq_func_cntl_2 |= BIAS_GEN_PDNB | RESET_EN | BYPASS;
+	if (pi->mem_gddr5)
+		mpll_dq_func_cntl &= ~PDNB;
+	mpll_dq_func_cntl_2 |= BIAS_GEN_PDNB | RESET_EN | BYPASS;
 
 
 	mclk_pwrmgt_cntl |= (MRDCKA0_RESET |
@@ -1889,15 +1889,15 @@ static int ni_populate_smc_acpi_state(struct radeon_device *rdev,
 			      MRDCKD1_PDNB);
 
 	dll_cntl |= (MRDCKA0_BYPASS |
-                     MRDCKA1_BYPASS |
-                     MRDCKB0_BYPASS |
-                     MRDCKB1_BYPASS |
-                     MRDCKC0_BYPASS |
-                     MRDCKC1_BYPASS |
-                     MRDCKD0_BYPASS |
-                     MRDCKD1_BYPASS);
-
-        spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK;
+		     MRDCKA1_BYPASS |
+		     MRDCKB0_BYPASS |
+		     MRDCKB1_BYPASS |
+		     MRDCKC0_BYPASS |
+		     MRDCKC1_BYPASS |
+		     MRDCKD0_BYPASS |
+		     MRDCKD1_BYPASS);
+
+	spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK;
 	spll_func_cntl_2 |= SCLK_MUX_SEL(4);
 
 	table->ACPIState.levels[0].mclk.vMPLL_AD_FUNC_CNTL = cpu_to_be32(mpll_ad_func_cntl);
@@ -2087,7 +2087,7 @@ static int ni_populate_sclk_value(struct radeon_device *rdev,
 
 static int ni_init_smc_spll_table(struct radeon_device *rdev)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	SMC_NISLANDS_SPLL_DIV_TABLE *spll_table;
 	NISLANDS_SMC_SCLK_VALUE sclk_params;
@@ -2309,8 +2309,8 @@ static int ni_convert_power_level_to_smc(struct radeon_device *rdev,
 					 NISLANDS_SMC_HW_PERFORMANCE_LEVEL *level)
 {
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
-        struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	int ret;
 	bool dll_state_on;
 	u16 std_vddc;
@@ -2389,8 +2389,8 @@ static int ni_populate_smc_t(struct radeon_device *rdev,
 			     struct radeon_ps *radeon_state,
 			     NISLANDS_SMC_SWSTATE *smc_state)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
-        struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 	struct ni_ps *state = ni_get_ps(radeon_state);
 	u32 a_t;
 	u32 t_l, t_h;
@@ -2449,8 +2449,8 @@ static int ni_populate_power_containment_values(struct radeon_device *rdev,
 						struct radeon_ps *radeon_state,
 						NISLANDS_SMC_SWSTATE *smc_state)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
-        struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	struct ni_ps *state = ni_get_ps(radeon_state);
 	u32 prev_sclk;
@@ -2593,7 +2593,7 @@ static int ni_enable_power_containment(struct radeon_device *rdev,
 				       struct radeon_ps *radeon_new_state,
 				       bool enable)
 {
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	PPSMC_Result smc_result;
 	int ret = 0;
 
@@ -2623,7 +2623,7 @@ static int ni_convert_power_state_to_smc(struct radeon_device *rdev,
 					 struct radeon_ps *radeon_state,
 					 NISLANDS_SMC_SWSTATE *smc_state)
 {
-        struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	struct ni_ps *state = ni_get_ps(radeon_state);
 	int i, ret;
@@ -2768,46 +2768,46 @@ static bool ni_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg)
 	bool result = true;
 
 	switch (in_reg) {
-        case  MC_SEQ_RAS_TIMING >> 2:
+	case  MC_SEQ_RAS_TIMING >> 2:
 		*out_reg = MC_SEQ_RAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_CAS_TIMING >> 2:
+	case MC_SEQ_CAS_TIMING >> 2:
 		*out_reg = MC_SEQ_CAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING >> 2:
+	case MC_SEQ_MISC_TIMING >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING2 >> 2:
+	case MC_SEQ_MISC_TIMING2 >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING2_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D0 >> 2:
+	case MC_SEQ_RD_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D1 >> 2:
+	case MC_SEQ_RD_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D1_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D0 >> 2:
+	case MC_SEQ_WR_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D1 >> 2:
+	case MC_SEQ_WR_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D1_LP >> 2;
 		break;
-        case MC_PMG_CMD_EMRS >> 2:
+	case MC_PMG_CMD_EMRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_EMRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS >> 2:
+	case MC_PMG_CMD_MRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS1 >> 2:
+	case MC_PMG_CMD_MRS1 >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS1_LP >> 2;
 		break;
-        case MC_SEQ_PMG_TIMING >> 2:
+	case MC_SEQ_PMG_TIMING >> 2:
 		*out_reg = MC_SEQ_PMG_TIMING_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS2 >> 2:
+	case MC_PMG_CMD_MRS2 >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS2_LP >> 2;
 		break;
-        default:
+	default:
 		result = false;
 		break;
 	}
@@ -2874,9 +2874,9 @@ static int ni_initialize_mc_reg_table(struct radeon_device *rdev)
 	struct ni_mc_reg_table *ni_table = &ni_pi->mc_reg_table;
 	u8 module_index = rv770_get_memory_module_index(rdev);
 
-        table = kzalloc(sizeof(struct atom_mc_reg_table), GFP_KERNEL);
-        if (!table)
-                return -ENOMEM;
+	table = kzalloc(sizeof(struct atom_mc_reg_table), GFP_KERNEL);
+	if (!table)
+		return -ENOMEM;
 
 	WREG32(MC_SEQ_RAS_TIMING_LP, RREG32(MC_SEQ_RAS_TIMING));
 	WREG32(MC_SEQ_CAS_TIMING_LP, RREG32(MC_SEQ_CAS_TIMING));
@@ -2894,25 +2894,25 @@ static int ni_initialize_mc_reg_table(struct radeon_device *rdev)
 
 	ret = radeon_atom_init_mc_reg_table(rdev, module_index, table);
 
-        if (ret)
-                goto init_mc_done;
+	if (ret)
+		goto init_mc_done;
 
 	ret = ni_copy_vbios_mc_reg_table(table, ni_table);
 
-        if (ret)
-                goto init_mc_done;
+	if (ret)
+		goto init_mc_done;
 
 	ni_set_s0_mc_reg_index(ni_table);
 
 	ret = ni_set_mc_special_registers(rdev, ni_table);
 
-        if (ret)
-                goto init_mc_done;
+	if (ret)
+		goto init_mc_done;
 
 	ni_set_valid_flag(ni_table);
 
 init_mc_done:
-        kfree(table);
+	kfree(table);
 
 	return ret;
 }
@@ -2992,7 +2992,7 @@ static int ni_populate_mc_reg_table(struct radeon_device *rdev,
 {
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	struct ni_ps *boot_state = ni_get_ps(radeon_boot_state);
 	SMC_NIslands_MCRegisters *mc_reg_table = &ni_pi->smc_mc_reg_table;
 
@@ -3023,7 +3023,7 @@ static int ni_upload_mc_reg_table(struct radeon_device *rdev,
 {
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	struct ni_ps *ni_new_state = ni_get_ps(radeon_new_state);
 	SMC_NIslands_MCRegisters *mc_reg_table = &ni_pi->smc_mc_reg_table;
 	u16 address;
@@ -3140,7 +3140,7 @@ static int ni_initialize_smc_cac_tables(struct radeon_device *rdev)
 	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 	PP_NIslands_CACTABLES *cac_tables = NULL;
 	int i, ret;
-        u32 reg;
+	u32 reg;
 
 	if (ni_pi->enable_cac == false)
 		return 0;
@@ -3420,13 +3420,13 @@ static int ni_pcie_performance_request(struct radeon_device *rdev,
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 
 	if ((perf_req == PCIE_PERF_REQ_PECI_GEN1) ||
-            (perf_req == PCIE_PERF_REQ_PECI_GEN2)) {
+	    (perf_req == PCIE_PERF_REQ_PECI_GEN2)) {
 		if (eg_pi->pcie_performance_request_registered == false)
 			radeon_acpi_pcie_notify_device_ready(rdev);
 		eg_pi->pcie_performance_request_registered = true;
 		return radeon_acpi_pcie_performance_request(rdev, perf_req, advertise);
 	} else if ((perf_req == PCIE_PERF_REQ_REMOVE_REGISTRY) &&
-                   eg_pi->pcie_performance_request_registered) {
+		    eg_pi->pcie_performance_request_registered) {
 		eg_pi->pcie_performance_request_registered = false;
 		return radeon_acpi_pcie_performance_request(rdev, perf_req, advertise);
 	}
@@ -3439,12 +3439,12 @@ static int ni_advertise_gen2_capability(struct radeon_device *rdev)
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	u32 tmp;
 
-        tmp = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+	tmp = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 
-        if ((tmp & LC_OTHER_SIDE_EVER_SENT_GEN2) &&
-            (tmp & LC_OTHER_SIDE_SUPPORTS_GEN2))
-                pi->pcie_gen2 = true;
-        else
+	if ((tmp & LC_OTHER_SIDE_EVER_SENT_GEN2) &&
+	    (tmp & LC_OTHER_SIDE_SUPPORTS_GEN2))
+		pi->pcie_gen2 = true;
+	else
 		pi->pcie_gen2 = false;
 
 	if (!pi->pcie_gen2)
@@ -3456,8 +3456,8 @@ static int ni_advertise_gen2_capability(struct radeon_device *rdev)
 static void ni_enable_bif_dynamic_pcie_gen2(struct radeon_device *rdev,
 					    bool enable)
 {
-        struct rv7xx_power_info *pi = rv770_get_pi(rdev);
-        u32 tmp, bif;
+	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
+	u32 tmp, bif;
 
 	tmp = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 
@@ -3500,7 +3500,7 @@ static void ni_enable_dynamic_pcie_gen2(struct radeon_device *rdev,
 	if (enable)
 		WREG32_P(GENERAL_PWRMGT, ENABLE_GEN2PCIE, ~ENABLE_GEN2PCIE);
 	else
-                WREG32_P(GENERAL_PWRMGT, 0, ~ENABLE_GEN2PCIE);
+		WREG32_P(GENERAL_PWRMGT, 0, ~ENABLE_GEN2PCIE);
 }
 
 void ni_set_uvd_clock_before_set_eng_clock(struct radeon_device *rdev,
@@ -3561,7 +3561,7 @@ void ni_update_current_ps(struct radeon_device *rdev,
 {
 	struct ni_ps *new_ps = ni_get_ps(rps);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 
 	eg_pi->current_rps = *rps;
 	ni_pi->current_ps = *new_ps;
@@ -3573,7 +3573,7 @@ void ni_update_requested_ps(struct radeon_device *rdev,
 {
 	struct ni_ps *new_ps = ni_get_ps(rps);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
-        struct ni_power_info *ni_pi = ni_get_pi(rdev);
+	struct ni_power_info *ni_pi = ni_get_pi(rdev);
 
 	eg_pi->requested_rps = *rps;
 	ni_pi->requested_ps = *new_ps;
@@ -3589,8 +3589,8 @@ int ni_dpm_enable(struct radeon_device *rdev)
 
 	if (pi->gfx_clock_gating)
 		ni_cg_clockgating_default(rdev);
-        if (btc_dpm_enabled(rdev))
-                return -EINVAL;
+	if (btc_dpm_enabled(rdev))
+		return -EINVAL;
 	if (pi->mg_clock_gating)
 		ni_mg_clockgating_default(rdev);
 	if (eg_pi->ls_clock_gating)
@@ -3989,7 +3989,7 @@ static int ni_parse_power_table(struct radeon_device *rdev)
 	union pplib_clock_info *clock_info;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	struct ni_ps *ps;
 
@@ -3998,8 +3998,9 @@ static int ni_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
@@ -4075,7 +4076,9 @@ int ni_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/sys/dev/pci/drm/radeon/ni_reg.h b/sys/dev/pci/drm/radeon/ni_reg.h
index da310a70c0f..827ccc87cbc 100644
--- a/sys/dev/pci/drm/radeon/ni_reg.h
+++ b/sys/dev/pci/drm/radeon/ni_reg.h
@@ -109,6 +109,8 @@
 #define NI_DP_MSE_SAT2                                 0x7398
 
 #define NI_DP_MSE_SAT_UPDATE                           0x739c
+#       define NI_DP_MSE_SAT_UPDATE_MASK               0x3
+#       define NI_DP_MSE_16_MTP_KEEPOUT                0x100
 
 #define NI_DIG_BE_CNTL                                 0x7140
 #       define NI_DIG_FE_SOURCE_SELECT(x)              (((x) & 0x7f) << 8)
diff --git a/sys/dev/pci/drm/radeon/nid.h b/sys/dev/pci/drm/radeon/nid.h
index 47eb49b77d3..3c9fec88ea4 100644
--- a/sys/dev/pci/drm/radeon/nid.h
+++ b/sys/dev/pci/drm/radeon/nid.h
@@ -1137,6 +1137,7 @@
 #define UVD_UDEC_ADDR_CONFIG				0xEF4C
 #define UVD_UDEC_DB_ADDR_CONFIG				0xEF50
 #define UVD_UDEC_DBW_ADDR_CONFIG			0xEF54
+#define UVD_NO_OP					0xEFFC
 #define UVD_RBC_RB_RPTR					0xF690
 #define UVD_RBC_RB_WPTR					0xF694
 #define UVD_STATUS					0xf6bc
diff --git a/sys/dev/pci/drm/radeon/r100.c b/sys/dev/pci/drm/radeon/r100.c
index 9fcac11ad7c..a1e8f7a716f 100644
--- a/sys/dev/pci/drm/radeon/r100.c
+++ b/sys/dev/pci/drm/radeon/r100.c
@@ -25,8 +25,10 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -36,6 +38,9 @@
 #include "rv250d.h"
 #include "atom.h"
 
+#include <linux/firmware.h>
+#include <linux/module.h>
+
 #include "r100_reg_safe.h"
 #include "rn50_reg_safe.h"
 
@@ -148,7 +153,7 @@ void r100_wait_for_vblank(struct radeon_device *rdev, int crtc)
  * bit to go high, when it does, we release the lock, and allow the
  * double buffered update to take place.
  */
-void r100_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void r100_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base, bool async)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = ((u32)crtc_base) | RADEON_CRTC_OFFSET__OFFSET_LOCK;
@@ -587,7 +592,8 @@ void r100_hpd_init(struct radeon_device *rdev)
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		struct radeon_connector *radeon_connector = to_radeon_connector(connector);
-		enable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			enable |= 1 << radeon_connector->hpd.hpd;
 		radeon_hpd_set_polarity(rdev, radeon_connector->hpd.hpd);
 	}
 	radeon_irq_kms_enable_hpd(rdev, enable);
@@ -609,7 +615,8 @@ void r100_hpd_fini(struct radeon_device *rdev)
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		struct radeon_connector *radeon_connector = to_radeon_connector(connector);
-		disable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			disable |= 1 << radeon_connector->hpd.hpd;
 	}
 	radeon_irq_kms_disable_hpd(rdev, disable);
 }
@@ -1035,11 +1042,9 @@ static int r100_cp_init_microcode(struct radeon_device *rdev)
 
 	err = request_firmware(&rdev->me_fw, fw_name, rdev->dev);
 	if (err) {
-		printk(KERN_ERR "radeon_cp: Failed to load firmware \"%s\"\n",
-		       fw_name);
+		pr_err("radeon_cp: Failed to load firmware \"%s\"\n", fw_name);
 	} else if (rdev->me_fw->size % 8) {
-		printk(KERN_ERR
-		       "radeon_cp: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("radeon_cp: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->me_fw->size, fw_name);
 		err = -EINVAL;
 		release_firmware(rdev->me_fw);
@@ -1064,11 +1069,7 @@ u32 r100_gfx_get_rptr(struct radeon_device *rdev,
 u32 r100_gfx_get_wptr(struct radeon_device *rdev,
 		      struct radeon_ring *ring)
 {
-	u32 wptr;
-
-	wptr = RREG32(RADEON_CP_RB_WPTR);
-
-	return wptr;
+	return RREG32(RADEON_CP_RB_WPTR);
 }
 
 void r100_gfx_set_wptr(struct radeon_device *rdev,
@@ -1084,8 +1085,7 @@ static void r100_cp_load_microcode(struct radeon_device *rdev)
 	int i, size;
 
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 
 	if (rdev->me_fw) {
@@ -1243,8 +1243,7 @@ void r100_cp_disable(struct radeon_device *rdev)
 	WREG32(RADEON_CP_CSQ_CNTL, 0);
 	WREG32(R_000770_SCRATCH_UMSK, 0);
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 }
 
@@ -1457,7 +1456,7 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p)
 	header = radeon_get_ib_value(p, h_idx);
 	crtc_id = radeon_get_ib_value(p, h_idx + 5);
 	reg = R100_CP_PACKET0_GET_REG(header);
-	crtc = drm_crtc_find(p->rdev->ddev, crtc_id);
+	crtc = drm_crtc_find(p->rdev->ddev, p->filp, crtc_id);
 	if (!crtc) {
 		DRM_ERROR("cannot find crtc %d\n", crtc_id);
 		return -ENOENT;
@@ -1878,8 +1877,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 		track->tex_dirty = true;
 		break;
 	default:
-		printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n",
-		       reg, idx);
+		pr_err("Forbidden register 0x%04X in cs at %d\n", reg, idx);
 		return -EINVAL;
 	}
 	return 0;
@@ -2483,8 +2481,7 @@ int r100_gui_wait_for_idle(struct radeon_device *rdev)
 	uint32_t tmp;
 
 	if (r100_rbbm_fifo_wait_for_entry(rdev, 64)) {
-		printk(KERN_WARNING "radeon: wait for empty RBBM fifo failed !"
-		       " Bad things might happen.\n");
+		pr_warn("radeon: wait for empty RBBM fifo failed! Bad things might happen.\n");
 	}
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = RREG32(RADEON_RBBM_STATUS);
@@ -2550,7 +2547,7 @@ void r100_bm_disable(struct radeon_device *rdev)
 	mdelay(1);
 }
 
-int r100_asic_reset(struct radeon_device *rdev)
+int r100_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	struct r100_mc_save save;
 	u32 status, tmp;
@@ -3145,7 +3142,8 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 {
 	fixed20_12 trcd_ff, trp_ff, tras_ff, trbs_ff, tcas_ff;
 	fixed20_12 sclk_ff, mclk_ff, sclk_eff_ff, sclk_delay_ff;
-	fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff, crit_point_ff;
+	fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff;
+	fixed20_12 crit_point_ff = {0};
 	uint32_t temp, data, mem_trcd, mem_trp, mem_tras;
 	fixed20_12 memtcas_ff[8] = {
 		dfixed_init(1),
@@ -3199,7 +3197,7 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 	fixed20_12 min_mem_eff;
 	fixed20_12 mc_latency_sclk, mc_latency_mclk, k1;
 	fixed20_12 cur_latency_mclk, cur_latency_sclk;
-	fixed20_12 disp_latency, disp_latency_overhead, disp_drain_rate,
+	fixed20_12 disp_latency, disp_latency_overhead, disp_drain_rate = {0},
 		disp_drain_rate2, read_return_rate;
 	fixed20_12 time_disp1_drop_priority;
 	int c;
@@ -3221,13 +3219,19 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 	radeon_update_display_priority(rdev);
 
 	if (rdev->mode_info.crtcs[0]->base.enabled) {
+		const struct drm_framebuffer *fb =
+			rdev->mode_info.crtcs[0]->base.primary->fb;
+
 		mode1 = &rdev->mode_info.crtcs[0]->base.mode;
-		pixel_bytes1 = rdev->mode_info.crtcs[0]->base.primary->fb->bits_per_pixel / 8;
+		pixel_bytes1 = fb->format->cpp[0];
 	}
 	if (!(rdev->flags & RADEON_SINGLE_CRTC)) {
 		if (rdev->mode_info.crtcs[1]->base.enabled) {
+			const struct drm_framebuffer *fb =
+				rdev->mode_info.crtcs[1]->base.primary->fb;
+
 			mode2 = &rdev->mode_info.crtcs[1]->base.mode;
-			pixel_bytes2 = rdev->mode_info.crtcs[1]->base.primary->fb->bits_per_pixel / 8;
+			pixel_bytes2 = fb->format->cpp[0];
 		}
 	}
 
@@ -3291,7 +3295,7 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 		mem_trp = ((temp >> 8) & 0x7) + 1;
 		mem_tras = ((temp >> 11) & 0xf) + 4;
 	} else if (rdev->family == CHIP_RV350 ||
-		   rdev->family <= CHIP_RV380) {
+		   rdev->family == CHIP_RV380) {
 		/* rv3x0 */
 		mem_trcd = (temp & 0x7) + 3;
 		mem_trp = ((temp >> 8) & 0x7) + 3;
@@ -3726,11 +3730,17 @@ int r100_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 		goto free_ib;
 	}
-	r = radeon_fence_wait(ib.fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(ib.fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		goto free_ib;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		r = -ETIMEDOUT;
+		goto free_ib;
 	}
+	r = 0;
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = RREG32(scratch);
 		if (tmp == 0xDEADBEEF) {
diff --git a/sys/dev/pci/drm/radeon/r200.c b/sys/dev/pci/drm/radeon/r200.c
index 8267fd0f4be..c22321cc5a4 100644
--- a/sys/dev/pci/drm/radeon/r200.c
+++ b/sys/dev/pci/drm/radeon/r200.c
@@ -25,8 +25,8 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -537,8 +537,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		track->tex_dirty = true;
 		break;
 	default:
-		printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n",
-		       reg, idx);
+		pr_err("Forbidden register 0x%04X in cs at %d\n", reg, idx);
 		return -EINVAL;
 	}
 	return 0;
diff --git a/sys/dev/pci/drm/radeon/r300.c b/sys/dev/pci/drm/radeon/r300.c
index 4ebb653cf65..21161aa8acb 100644
--- a/sys/dev/pci/drm/radeon/r300.c
+++ b/sys/dev/pci/drm/radeon/r300.c
@@ -25,13 +25,15 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
+#include <drm/drm_crtc_helper.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_asic.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "r100_track.h"
 #include "r300d.h"
 #include "rv350d.h"
@@ -385,8 +387,7 @@ static void r300_gpu_init(struct radeon_device *rdev)
 	WREG32(R300_GB_TILE_CONFIG, gb_tile_config);
 
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 
 	tmp = RREG32(R300_DST_PIPE_CONFIG);
@@ -397,18 +398,16 @@ static void r300_gpu_init(struct radeon_device *rdev)
 	       R300_DC_DC_DISABLE_IGNORE_PE);
 
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 	if (r300_mc_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait MC idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait MC idle while programming pipes. Bad things might happen.\n");
 	}
-	DRM_INFO("radeon: %d quad pipes, %d Z pipes initialized.\n",
+	DRM_INFO("radeon: %d quad pipes, %d Z pipes initialized\n",
 		 rdev->num_gb_pipes, rdev->num_z_pipes);
 }
 
-int r300_asic_reset(struct radeon_device *rdev)
+int r300_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	struct r100_mc_save save;
 	u32 status, tmp;
@@ -1163,7 +1162,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 	}
 	return 0;
 fail:
-	printk(KERN_ERR "Forbidden register 0x%04X in cs at %d (val=%08x)\n",
+	pr_err("Forbidden register 0x%04X in cs at %d (val=%08x)\n",
 	       reg, idx, idx_value);
 	return -EINVAL;
 }
diff --git a/sys/dev/pci/drm/radeon/r420.c b/sys/dev/pci/drm/radeon/r420.c
index 1089aebee66..45e1d4e6075 100644
--- a/sys/dev/pci/drm/radeon/r420.c
+++ b/sys/dev/pci/drm/radeon/r420.c
@@ -25,7 +25,9 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -90,8 +92,7 @@ void r420_pipes_init(struct radeon_device *rdev)
 	       (1 << 2) | (1 << 3));
 	/* add idle wait as per freedesktop.org bug 24041 */
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 	/* get max number of pipes */
 	gb_pipe_select = RREG32(R400_GB_PIPE_SELECT);
@@ -126,8 +127,7 @@ void r420_pipes_init(struct radeon_device *rdev)
 	tmp |= R300_TILE_SIZE_16 | R300_ENABLE_TILING;
 	WREG32(R300_GB_TILE_CONFIG, tmp);
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 
 	tmp = RREG32(R300_DST_PIPE_CONFIG);
@@ -139,8 +139,7 @@ void r420_pipes_init(struct radeon_device *rdev)
 	       R300_DC_DC_DISABLE_IGNORE_PE);
 
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while programming pipes. Bad things might happen.\n");
 	}
 
 	if (rdev->family == CHIP_RV530) {
@@ -204,6 +203,7 @@ static void r420_clock_resume(struct radeon_device *rdev)
 
 static void r420_cp_errata_init(struct radeon_device *rdev)
 {
+	int r;
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 
 	/* RV410 and R420 can lock up if CP DMA to host memory happens
@@ -213,7 +213,8 @@ static void r420_cp_errata_init(struct radeon_device *rdev)
 	 * of the CP init, apparently.
 	 */
 	radeon_scratch_get(rdev, &rdev->config.r300.resync_scratch);
-	radeon_ring_lock(rdev, ring, 8);
+	r = radeon_ring_lock(rdev, ring, 8);
+	WARN_ON(r);
 	radeon_ring_write(ring, PACKET0(R300_CP_RESYNC_ADDR, 1));
 	radeon_ring_write(ring, rdev->config.r300.resync_scratch);
 	radeon_ring_write(ring, 0xDEADBEEF);
@@ -222,12 +223,14 @@ static void r420_cp_errata_init(struct radeon_device *rdev)
 
 static void r420_cp_errata_fini(struct radeon_device *rdev)
 {
+	int r;
 	struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 
 	/* Catch the RESYNC we dispatched all the way back,
 	 * at the very beginning of the CP init.
 	 */
-	radeon_ring_lock(rdev, ring, 8);
+	r = radeon_ring_lock(rdev, ring, 8);
+	WARN_ON(r);
 	radeon_ring_write(ring, PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
 	radeon_ring_write(ring, R300_RB3D_DC_FINISH);
 	radeon_ring_unlock_commit(rdev, ring, false);
diff --git a/sys/dev/pci/drm/radeon/r520.c b/sys/dev/pci/drm/radeon/r520.c
index 3f51ef2028b..074cf752fae 100644
--- a/sys/dev/pci/drm/radeon/r520.c
+++ b/sys/dev/pci/drm/radeon/r520.c
@@ -25,7 +25,7 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "atom.h"
@@ -86,8 +86,7 @@ static void r520_gpu_init(struct radeon_device *rdev)
 	      (((gb_pipe_select >> 8) & 0xF) << 4);
 	WREG32_PLL(0x000D, tmp);
 	if (r520_mc_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait MC idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait MC idle while programming pipes. Bad things might happen.\n");
 	}
 }
 
diff --git a/sys/dev/pci/drm/radeon/r600.c b/sys/dev/pci/drm/radeon/r600.c
index ddfcbbd1e7a..2fa6028912d 100644
--- a/sys/dev/pci/drm/radeon/r600.c
+++ b/sys/dev/pci/drm/radeon/r600.c
@@ -25,8 +25,12 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/slab.h>
+#include <linux/seq_file.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -231,8 +235,8 @@ int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 		fb_div |= 1;
 
 	r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL);
-        if (r)
-                return r;
+	if (r)
+		return r;
 
 	/* assert PLL_RESET */
 	WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_RESET_MASK, ~UPLL_RESET_MASK);
@@ -998,7 +1002,8 @@ void r600_hpd_init(struct radeon_device *rdev)
 				break;
 			}
 		}
-		enable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			enable |= 1 << radeon_connector->hpd.hpd;
 		radeon_hpd_set_polarity(rdev, radeon_connector->hpd.hpd);
 	}
 	radeon_irq_kms_enable_hpd(rdev, enable);
@@ -1051,7 +1056,8 @@ void r600_hpd_fini(struct radeon_device *rdev)
 				break;
 			}
 		}
-		disable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			disable |= 1 << radeon_connector->hpd.hpd;
 	}
 	radeon_irq_kms_disable_hpd(rdev, disable);
 }
@@ -1088,7 +1094,7 @@ void r600_pcie_gart_tlb_flush(struct radeon_device *rdev)
 		tmp = RREG32(VM_CONTEXT0_REQUEST_RESPONSE);
 		tmp = (tmp & RESPONSE_TYPE_MASK) >> RESPONSE_TYPE_SHIFT;
 		if (tmp == 2) {
-			printk(KERN_WARNING "[drm] r600 flush TLB failed\n");
+			pr_warn("[drm] r600 flush TLB failed\n");
 			return;
 		}
 		if (tmp) {
@@ -1486,7 +1492,7 @@ static int r600_mc_init(struct radeon_device *rdev)
 					rdev->fastfb_working = true;
 				}
 			}
-  		}
+		}
 	}
 
 	radeon_update_bandwidth_info(rdev);
@@ -1867,10 +1873,15 @@ static void r600_gpu_pci_config_reset(struct radeon_device *rdev)
 	}
 }
 
-int r600_asic_reset(struct radeon_device *rdev)
+int r600_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	u32 reset_mask;
 
+	if (hard) {
+		r600_gpu_pci_config_reset(rdev);
+		return 0;
+	}
+
 	reset_mask = r600_gpu_check_soft_reset(rdev);
 
 	if (reset_mask)
@@ -2539,8 +2550,7 @@ int r600_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->pfp_fw->size != pfp_req_size) {
-		printk(KERN_ERR
-		       "r600_cp: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("r600_cp: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->pfp_fw->size, fw_name);
 		err = -EINVAL;
 		goto out;
@@ -2551,8 +2561,7 @@ int r600_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->me_fw->size != me_req_size) {
-		printk(KERN_ERR
-		       "r600_cp: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("r600_cp: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->me_fw->size, fw_name);
 		err = -EINVAL;
 	}
@@ -2562,8 +2571,7 @@ int r600_init_microcode(struct radeon_device *rdev)
 	if (err)
 		goto out;
 	if (rdev->rlc_fw->size != rlc_req_size) {
-		printk(KERN_ERR
-		       "r600_rlc: Bogus length %zu in firmware \"%s\"\n",
+		pr_err("r600_rlc: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->rlc_fw->size, fw_name);
 		err = -EINVAL;
 	}
@@ -2572,15 +2580,12 @@ int r600_init_microcode(struct radeon_device *rdev)
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", smc_chip_name);
 		err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 		if (err) {
-			printk(KERN_ERR
-			       "smc: error loading firmware \"%s\"\n",
-			       fw_name);
+			pr_err("smc: error loading firmware \"%s\"\n", fw_name);
 			release_firmware(rdev->smc_fw);
 			rdev->smc_fw = NULL;
 			err = 0;
 		} else if (rdev->smc_fw->size != smc_req_size) {
-			printk(KERN_ERR
-			       "smc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("smc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->smc_fw->size, fw_name);
 			err = -EINVAL;
 		}
@@ -2589,8 +2594,7 @@ int r600_init_microcode(struct radeon_device *rdev)
 out:
 	if (err) {
 		if (err != -EINVAL)
-			printk(KERN_ERR
-			       "r600_cp: Failed to load firmware \"%s\"\n",
+			pr_err("r600_cp: Failed to load firmware \"%s\"\n",
 			       fw_name);
 		release_firmware(rdev->pfp_fw);
 		rdev->pfp_fw = NULL;
@@ -2620,11 +2624,7 @@ u32 r600_gfx_get_rptr(struct radeon_device *rdev,
 u32 r600_gfx_get_wptr(struct radeon_device *rdev,
 		      struct radeon_ring *ring)
 {
-	u32 wptr;
-
-	wptr = RREG32(R600_CP_RB_WPTR);
-
-	return wptr;
+	return RREG32(R600_CP_RB_WPTR);
 }
 
 void r600_gfx_set_wptr(struct radeon_device *rdev,
@@ -3031,6 +3031,73 @@ void r600_clear_surface_reg(struct radeon_device *rdev, int reg)
 	/* FIXME: implement */
 }
 
+static void r600_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails uvd_v1_0_resume() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void r600_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = uvd_v1_0_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void r600_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
 static int r600_startup(struct radeon_device *rdev)
 {
 	struct radeon_ring *ring;
@@ -3066,17 +3133,7 @@ static int r600_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	if (rdev->has_uvd) {
-		r = uvd_v1_0_resume(rdev);
-		if (!r) {
-			r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
-			if (r) {
-				dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
-			}
-		}
-		if (r)
-			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
-	}
+	r600_uvd_start(rdev);
 
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
@@ -3106,17 +3163,7 @@ static int r600_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	if (rdev->has_uvd) {
-		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-		if (ring->ring_size) {
-			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-					     RADEON_CP_PACKET2);
-			if (!r)
-				r = uvd_v1_0_init(rdev);
-			if (r)
-				DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
-		}
-	}
+	r600_uvd_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -3260,13 +3307,7 @@ int r600_init(struct radeon_device *rdev)
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
-	if (rdev->has_uvd) {
-		r = radeon_uvd_init(rdev);
-		if (!r) {
-			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
-			r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
-		}
-	}
+	r600_uvd_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -3377,11 +3418,17 @@ int r600_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 		goto free_ib;
 	}
-	r = radeon_fence_wait(ib.fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(ib.fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		goto free_ib;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		r = -ETIMEDOUT;
+		goto free_ib;
 	}
+	r = 0;
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = RREG32(scratch);
 		if (tmp == 0xDEADBEEF)
@@ -4441,7 +4488,7 @@ static void r600_pcie_gen2_enable(struct radeon_device *rdev)
 {
 	u32 link_width_cntl, lanes, speed_cntl, training_cntl, tmp;
 	u16 link_cntl2;
-	u32 mask;
+	enum pci_bus_speed max_bus_speed;
 
 	if (radeon_pcie_gen2 == 0)
 		return;
@@ -4460,10 +4507,9 @@ static void r600_pcie_gen2_enable(struct radeon_device *rdev)
 	if (rdev->family <= CHIP_R600)
 		return;
 
-	if (drm_pcie_get_speed_cap_mask(rdev->ddev, &mask))
-		return;
-
-	if (!(mask & (DRM_PCIE_SPEED_50|DRM_PCIE_SPEED_80)))
+	max_bus_speed = pcie_get_speed_cap(rdev->pdev->bus->self);
+	if ((max_bus_speed != PCIE_SPEED_5_0GT) &&
+		(max_bus_speed != PCIE_SPEED_8_0GT))
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
@@ -4569,7 +4615,7 @@ uint64_t r600_get_gpu_clock_counter(struct radeon_device *rdev)
 	mutex_lock(&rdev->gpu_clock_mutex);
 	WREG32(RLC_CAPTURE_GPU_CLOCK_COUNT, 1);
 	clock = (uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_LSB) |
-	        ((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+		((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
 	mutex_unlock(&rdev->gpu_clock_mutex);
 	return clock;
 }
diff --git a/sys/dev/pci/drm/radeon/r600_blit_shaders.c b/sys/dev/pci/drm/radeon/r600_blit_shaders.c
index 2497bad5fec..443cbe59b27 100644
--- a/sys/dev/pci/drm/radeon/r600_blit_shaders.c
+++ b/sys/dev/pci/drm/radeon/r600_blit_shaders.c
@@ -24,7 +24,9 @@
  *     Alex Deucher <alexander.deucher@amd.com>
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/bug.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
 
 /*
  * R6xx+ cards need to use the 3D engine to blit data which requires
diff --git a/sys/dev/pci/drm/radeon/r600_cs.c b/sys/dev/pci/drm/radeon/r600_cs.c
index 8004b49f542..c96b31950ca 100644
--- a/sys/dev/pci/drm/radeon/r600_cs.c
+++ b/sys/dev/pci/drm/radeon/r600_cs.c
@@ -25,8 +25,10 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/kernel.h>
+#include <drm/drmP.h>
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "r600d.h"
 #include "r600_reg_safe.h"
 
@@ -885,7 +887,7 @@ int r600_cs_common_vline_parse(struct radeon_cs_parser *p,
 	crtc_id = radeon_get_ib_value(p, h_idx + 2 + 7 + 1);
 	reg = R600_CP_PACKET0_GET_REG(header);
 
-	crtc = drm_crtc_find(p->rdev->ddev, crtc_id);
+	crtc = drm_crtc_find(p->rdev->ddev, p->filp, crtc_id);
 	if (!crtc) {
 		DRM_ERROR("cannot find crtc %d\n", crtc_id);
 		return -ENOENT;
@@ -930,8 +932,7 @@ static int r600_packet0_check(struct radeon_cs_parser *p,
 		}
 		break;
 	default:
-		printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n",
-		       reg, idx);
+		pr_err("Forbidden register 0x%04X in cs at %d\n", reg, idx);
 		return -EINVAL;
 	}
 	return 0;
@@ -1670,8 +1671,8 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         (idx_value & 0xfffffff0) +
-		         ((u64)(tmp & 0xff) << 32);
+			 (idx_value & 0xfffffff0) +
+			 ((u64)(tmp & 0xff) << 32);
 
 		ib[idx + 0] = offset;
 		ib[idx + 1] = (tmp & 0xffffff00) | (upper_32_bits(offset) & 0xff);
@@ -1711,8 +1712,8 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         idx_value +
-		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
+			 idx_value +
+			 ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
 		ib[idx+0] = offset;
 		ib[idx+1] = upper_32_bits(offset) & 0xff;
@@ -1763,8 +1764,8 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			}
 
 			offset = reloc->gpu_offset +
-			         (radeon_get_ib_value(p, idx+1) & 0xfffffff0) +
-			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+				 (radeon_get_ib_value(p, idx+1) & 0xfffffff0) +
+				 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 			ib[idx+1] = (ib[idx+1] & 0x3) | (offset & 0xfffffff0);
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
@@ -1875,8 +1876,8 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				return -EINVAL;
 			}
 			offset = reloc->gpu_offset +
-			         (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
-			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+				 (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
+				 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 			ib[idx+1] = offset & 0xfffffff8;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
@@ -1897,8 +1898,8 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 		}
 
 		offset = reloc->gpu_offset +
-		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
-		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
+			 (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
+			 ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
 		ib[idx+1] = offset & 0xfffffffc;
 		ib[idx+2] = (ib[idx+2] & 0xffffff00) | (upper_32_bits(offset) & 0xff);
@@ -2318,7 +2319,7 @@ int r600_cs_parse(struct radeon_cs_parser *p)
 	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib.length_dw; r++) {
-		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
+		pr_info("%05d  0x%08X\n", r, p->ib.ptr[r]);
 		mdelay(1);
 	}
 #endif
@@ -2327,101 +2328,6 @@ int r600_cs_parse(struct radeon_cs_parser *p)
 	return 0;
 }
 
-#ifdef CONFIG_DRM_RADEON_UMS
-
-/**
- * cs_parser_fini() - clean parser states
- * @parser:	parser structure holding parsing context.
- * @error:	error number
- *
- * If error is set than unvalidate buffer, otherwise just free memory
- * used by parsing context.
- **/
-static void r600_cs_parser_fini(struct radeon_cs_parser *parser, int error)
-{
-	unsigned i;
-
-	kfree(parser->relocs);
-	for (i = 0; i < parser->nchunks; i++)
-		drm_free_large(parser->chunks[i].kdata);
-	kfree(parser->chunks);
-	kfree(parser->chunks_array);
-}
-
-static int r600_cs_parser_relocs_legacy(struct radeon_cs_parser *p)
-{
-	if (p->chunk_relocs == NULL) {
-		return 0;
-	}
-	p->relocs = kzalloc(sizeof(struct radeon_bo_list), GFP_KERNEL);
-	if (p->relocs == NULL) {
-		return -ENOMEM;
-	}
-	return 0;
-}
-
-int r600_cs_legacy(struct drm_device *dev, void *data, struct drm_file *filp,
-			unsigned family, u32 *ib, int *l)
-{
-	struct radeon_cs_parser parser;
-	struct radeon_cs_chunk *ib_chunk;
-	struct r600_cs_track *track;
-	int r;
-
-	/* initialize tracker */
-	track = kzalloc(sizeof(*track), GFP_KERNEL);
-	if (track == NULL)
-		return -ENOMEM;
-	r600_cs_track_init(track);
-	r600_cs_legacy_get_tiling_conf(dev, &track->npipes, &track->nbanks, &track->group_size);
-	/* initialize parser */
-	memset(&parser, 0, sizeof(struct radeon_cs_parser));
-	parser.filp = filp;
-	parser.dev = &dev->pdev->dev;
-	parser.rdev = NULL;
-	parser.family = family;
-	parser.track = track;
-	parser.ib.ptr = ib;
-	r = radeon_cs_parser_init(&parser, data);
-	if (r) {
-		DRM_ERROR("Failed to initialize parser !\n");
-		r600_cs_parser_fini(&parser, r);
-		return r;
-	}
-	r = r600_cs_parser_relocs_legacy(&parser);
-	if (r) {
-		DRM_ERROR("Failed to parse relocation !\n");
-		r600_cs_parser_fini(&parser, r);
-		return r;
-	}
-	/* Copy the packet into the IB, the parser will read from the
-	 * input memory (cached) and write to the IB (which can be
-	 * uncached). */
-	ib_chunk = parser.chunk_ib;
-	parser.ib.length_dw = ib_chunk->length_dw;
-	*l = parser.ib.length_dw;
-	if (copy_from_user(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
-		r = -EFAULT;
-		r600_cs_parser_fini(&parser, r);
-		return r;
-	}
-	r = r600_cs_parse(&parser);
-	if (r) {
-		DRM_ERROR("Invalid command stream !\n");
-		r600_cs_parser_fini(&parser, r);
-		return r;
-	}
-	r600_cs_parser_fini(&parser, r);
-	return r;
-}
-
-void r600_cs_legacy_init(void)
-{
-	r600_nomm = 1;
-}
-
-#endif
-
 /*
  *  DMA
  */
@@ -2621,7 +2527,7 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 	} while (p->idx < p->chunk_ib->length_dw);
 #if 0
 	for (r = 0; r < p->ib->length_dw; r++) {
-		printk(KERN_INFO "%05d  0x%08X\n", r, p->ib.ptr[r]);
+		pr_info("%05d  0x%08X\n", r, p->ib.ptr[r]);
 		mdelay(1);
 	}
 #endif
diff --git a/sys/dev/pci/drm/radeon/r600_dma.c b/sys/dev/pci/drm/radeon/r600_dma.c
index ee1d48ad1e4..fb65e6fb5c4 100644
--- a/sys/dev/pci/drm/radeon/r600_dma.c
+++ b/sys/dev/pci/drm/radeon/r600_dma.c
@@ -21,7 +21,7 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "r600d.h"
@@ -368,11 +368,16 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 		return r;
 	}
-	r = radeon_fence_wait(ib.fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(ib.fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		return r;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		return -ETIMEDOUT;
 	}
+	r = 0;
 	for (i = 0; i < rdev->usec_timeout; i++) {
 		tmp = le32_to_cpu(rdev->wb.wb[index/4]);
 		if (tmp == 0xDEADBEEF)
diff --git a/sys/dev/pci/drm/radeon/r600_dpm.c b/sys/dev/pci/drm/radeon/r600_dpm.c
index c64fb66085c..5e044c98fca 100644
--- a/sys/dev/pci/drm/radeon/r600_dpm.c
+++ b/sys/dev/pci/drm/radeon/r600_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "r600d.h"
@@ -69,86 +69,89 @@ const u32 r600_dtc[R600_PM_NUMBER_OF_TC] =
 
 void r600_dpm_print_class_info(u32 class, u32 class2)
 {
-	printk("\tui class: ");
+	const char *s;
+
 	switch (class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) {
 	case ATOM_PPLIB_CLASSIFICATION_UI_NONE:
 	default:
-		printk("none\n");
+		s = "none";
 		break;
 	case ATOM_PPLIB_CLASSIFICATION_UI_BATTERY:
-		printk("battery\n");
+		s = "battery";
 		break;
 	case ATOM_PPLIB_CLASSIFICATION_UI_BALANCED:
-		printk("balanced\n");
+		s = "balanced";
 		break;
 	case ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE:
-		printk("performance\n");
+		s = "performance";
 		break;
 	}
-	printk("\tinternal class: ");
+	printk("\tui class: %s\n", s);
+
+	printk("\tinternal class:");
 	if (((class & ~ATOM_PPLIB_CLASSIFICATION_UI_MASK) == 0) &&
 	    (class2 == 0))
-		printk("none");
+		pr_cont(" none");
 	else {
 		if (class & ATOM_PPLIB_CLASSIFICATION_BOOT)
-			printk("boot ");
+			pr_cont(" boot");
 		if (class & ATOM_PPLIB_CLASSIFICATION_THERMAL)
-			printk("thermal ");
+			pr_cont(" thermal");
 		if (class & ATOM_PPLIB_CLASSIFICATION_LIMITEDPOWERSOURCE)
-			printk("limited_pwr ");
+			pr_cont(" limited_pwr");
 		if (class & ATOM_PPLIB_CLASSIFICATION_REST)
-			printk("rest ");
+			pr_cont(" rest");
 		if (class & ATOM_PPLIB_CLASSIFICATION_FORCED)
-			printk("forced ");
+			pr_cont(" forced");
 		if (class & ATOM_PPLIB_CLASSIFICATION_3DPERFORMANCE)
-			printk("3d_perf ");
+			pr_cont(" 3d_perf");
 		if (class & ATOM_PPLIB_CLASSIFICATION_OVERDRIVETEMPLATE)
-			printk("ovrdrv ");
+			pr_cont(" ovrdrv");
 		if (class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE)
-			printk("uvd ");
+			pr_cont(" uvd");
 		if (class & ATOM_PPLIB_CLASSIFICATION_3DLOW)
-			printk("3d_low ");
+			pr_cont(" 3d_low");
 		if (class & ATOM_PPLIB_CLASSIFICATION_ACPI)
-			printk("acpi ");
+			pr_cont(" acpi");
 		if (class & ATOM_PPLIB_CLASSIFICATION_HD2STATE)
-			printk("uvd_hd2 ");
+			pr_cont(" uvd_hd2");
 		if (class & ATOM_PPLIB_CLASSIFICATION_HDSTATE)
-			printk("uvd_hd ");
+			pr_cont(" uvd_hd");
 		if (class & ATOM_PPLIB_CLASSIFICATION_SDSTATE)
-			printk("uvd_sd ");
+			pr_cont(" uvd_sd");
 		if (class2 & ATOM_PPLIB_CLASSIFICATION2_LIMITEDPOWERSOURCE_2)
-			printk("limited_pwr2 ");
+			pr_cont(" limited_pwr2");
 		if (class2 & ATOM_PPLIB_CLASSIFICATION2_ULV)
-			printk("ulv ");
+			pr_cont(" ulv");
 		if (class2 & ATOM_PPLIB_CLASSIFICATION2_MVC)
-			printk("uvd_mvc ");
+			pr_cont(" uvd_mvc");
 	}
-	printk("\n");
+	pr_cont("\n");
 }
 
 void r600_dpm_print_cap_info(u32 caps)
 {
-	printk("\tcaps: ");
+	printk("\tcaps:");
 	if (caps & ATOM_PPLIB_SINGLE_DISPLAY_ONLY)
-		printk("single_disp ");
+		pr_cont(" single_disp");
 	if (caps & ATOM_PPLIB_SUPPORTS_VIDEO_PLAYBACK)
-		printk("video ");
+		pr_cont(" video");
 	if (caps & ATOM_PPLIB_DISALLOW_ON_DC)
-		printk("no_dc ");
-	printk("\n");
+		pr_cont(" no_dc");
+	pr_cont("\n");
 }
 
 void r600_dpm_print_ps_status(struct radeon_device *rdev,
 			      struct radeon_ps *rps)
 {
-	printk("\tstatus: ");
+	printk("\tstatus:");
 	if (rps == rdev->pm.dpm.current_ps)
-		printk("c ");
+		pr_cont(" c");
 	if (rps == rdev->pm.dpm.requested_ps)
-		printk("r ");
+		pr_cont(" r");
 	if (rps == rdev->pm.dpm.boot_ps)
-		printk("b ");
-	printk("\n");
+		pr_cont(" b");
+	pr_cont("\n");
 }
 
 u32 r600_dpm_get_vblank_time(struct radeon_device *rdev)
@@ -845,7 +848,7 @@ int r600_get_platform_caps(struct radeon_device *rdev)
 	struct radeon_mode_info *mode_info = &rdev->mode_info;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 
 	if (!atom_parse_data_header(mode_info->atom_context, index, NULL,
@@ -875,7 +878,7 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 	union fan_info *fan_info;
 	ATOM_PPLIB_Clock_Voltage_Dependency_Table *dep_table;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	int ret, i;
 
@@ -988,7 +991,7 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 			ATOM_PPLIB_PhaseSheddingLimits_Record *entry;
 
 			rdev->pm.dpm.dyn_state.phase_shedding_limits_table.entries =
-				kzalloc(psl->ucNumEntries *
+				kcalloc(psl->ucNumEntries,
 					sizeof(struct radeon_phase_shedding_limits_entry),
 					GFP_KERNEL);
 			if (!rdev->pm.dpm.dyn_state.phase_shedding_limits_table.entries) {
@@ -1071,7 +1074,7 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 			ext_hdr->usVCETableOffset) {
 			VCEClockInfoArray *array = (VCEClockInfoArray *)
 				(mode_info->atom_context->bios + data_offset +
-                                 le16_to_cpu(ext_hdr->usVCETableOffset) + 1);
+				 le16_to_cpu(ext_hdr->usVCETableOffset) + 1);
 			ATOM_PPLIB_VCE_Clock_Voltage_Limit_Table *limits =
 				(ATOM_PPLIB_VCE_Clock_Voltage_Limit_Table *)
 				(mode_info->atom_context->bios + data_offset +
@@ -1324,9 +1327,9 @@ enum radeon_pcie_gen r600_get_pcie_gen_support(struct radeon_device *rdev,
 	case RADEON_PCIE_GEN3:
 		return RADEON_PCIE_GEN3;
 	default:
-		if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == RADEON_PCIE_GEN3))
+		if ((sys_mask & RADEON_PCIE_SPEED_80) && (default_gen == RADEON_PCIE_GEN3))
 			return RADEON_PCIE_GEN3;
-		else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == RADEON_PCIE_GEN2))
+		else if ((sys_mask & RADEON_PCIE_SPEED_50) && (default_gen == RADEON_PCIE_GEN2))
 			return RADEON_PCIE_GEN2;
 		else
 			return RADEON_PCIE_GEN1;
diff --git a/sys/dev/pci/drm/radeon/r600_hdmi.c b/sys/dev/pci/drm/radeon/r600_hdmi.c
index 07df68f80b8..ab32830c4e2 100644
--- a/sys/dev/pci/drm/radeon/r600_hdmi.c
+++ b/sys/dev/pci/drm/radeon/r600_hdmi.c
@@ -23,9 +23,10 @@
  *
  * Authors: Christian König
  */
-#include <dev/pci/drm/linux_hdmi.h>
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/hdmi.h>
+#include <linux/gcd.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -57,7 +58,7 @@ enum r600_hdmi_iec_status_bits {
 
 static struct r600_audio_pin r600_audio_status(struct radeon_device *rdev)
 {
-	struct r600_audio_pin status;
+	struct r600_audio_pin status = {};
 	uint32_t value;
 
 	value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL);
@@ -214,7 +215,7 @@ void r600_hdmi_update_acr(struct drm_encoder *encoder, long offset,
  * build a HDMI Video Info Frame
  */
 void r600_set_avi_packet(struct radeon_device *rdev, u32 offset,
-    unsigned char *buffer, size_t size)
+			 unsigned char *buffer, size_t size)
 {
 	uint8_t *frame = buffer + 3;
 
@@ -311,7 +312,7 @@ void r600_hdmi_audio_workaround(struct drm_encoder *encoder)
 }
 
 void r600_hdmi_audio_set_dto(struct radeon_device *rdev,
-    struct radeon_crtc *crtc, unsigned int clock)
+			     struct radeon_crtc *crtc, unsigned int clock)
 {
 	struct radeon_encoder *radeon_encoder;
 	struct radeon_encoder_atom_dig *dig;
diff --git a/sys/dev/pci/drm/radeon/r600d.h b/sys/dev/pci/drm/radeon/r600d.h
index 1e8495cca41..2e00a5287bd 100644
--- a/sys/dev/pci/drm/radeon/r600d.h
+++ b/sys/dev/pci/drm/radeon/r600d.h
@@ -1490,6 +1490,7 @@
 #define UVD_GPCOM_VCPU_DATA0				0xef10
 #define UVD_GPCOM_VCPU_DATA1				0xef14
 #define UVD_ENGINE_CNTL					0xef18
+#define UVD_NO_OP					0xeffc
 
 #define UVD_SEMA_CNTL					0xf400
 #define UVD_RB_ARB_CTRL					0xf480
diff --git a/sys/dev/pci/drm/radeon/radeon.h b/sys/dev/pci/drm/radeon/radeon.h
index 2acad16f0e9..595543c4fbe 100644
--- a/sys/dev/pci/drm/radeon/radeon.h
+++ b/sys/dev/pci/drm/radeon/radeon.h
@@ -60,15 +60,23 @@
  *                          are considered as fatal)
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/atomic.h>
+#include <linux/wait.h>
+#include <linux/list.h>
+#include <linux/kref.h>
+#include <linux/interval_tree.h>
+#include <linux/hashtable.h>
+#include <linux/dma-fence.h>
 
-#include <dev/pci/drm/ttm/ttm_bo_api.h>
-#include <dev/pci/drm/ttm/ttm_bo_driver.h>
-#include <dev/pci/drm/ttm/ttm_placement.h>
-#include <dev/pci/drm/ttm/ttm_module.h>
-#include <dev/pci/drm/ttm/ttm_execbuf_util.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_execbuf_util.h>
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drm_gem.h>
+
+#include <drm/drmP.h>
 
 #include <dev/wscons/wsconsio.h>
 #include <dev/wscons/wsdisplayvar.h>
@@ -115,6 +123,10 @@ extern int radeon_bapm;
 extern int radeon_backlight;
 extern int radeon_auxch;
 extern int radeon_mst;
+extern int radeon_uvd;
+extern int radeon_vce;
+extern int radeon_si_support;
+extern int radeon_cik_support;
 
 /*
  * Copy from radeon_drv.h so we don't have to include both and have conflicting
@@ -122,6 +134,7 @@ extern int radeon_mst;
  */
 #define RADEON_MAX_USEC_TIMEOUT			100000	/* 100 ms */
 #define RADEON_FENCE_JIFFIES_TIMEOUT		(HZ / 2)
+#define RADEON_USEC_IB_TEST_TIMEOUT		1000000 /* 1s */
 /* RADEON_IB_POOL_SIZE must be a power of 2 */
 #define RADEON_IB_POOL_SIZE			16
 #define RADEON_DEBUGFS_MAX_COMPONENTS		32
@@ -366,7 +379,7 @@ struct radeon_fence_driver {
 };
 
 struct radeon_fence {
-	struct fence		base;
+	struct dma_fence		base;
 
 	struct radeon_device	*rdev;
 	uint64_t		seq;
@@ -374,7 +387,7 @@ struct radeon_fence {
 	unsigned		ring;
 	bool			is_vm_update;
 
-	wait_queue_t		fence_wake;
+	wait_queue_entry_t		fence_wake;
 };
 
 int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring);
@@ -384,6 +397,7 @@ void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring);
 int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, int ring);
 void radeon_fence_process(struct radeon_device *rdev, int ring);
 bool radeon_fence_signaled(struct radeon_fence *fence);
+long radeon_fence_wait_timeout(struct radeon_fence *fence, bool interruptible, long timeout);
 int radeon_fence_wait(struct radeon_fence *fence, bool interruptible);
 int radeon_fence_wait_next(struct radeon_device *rdev, int ring);
 int radeon_fence_wait_empty(struct radeon_device *rdev, int ring);
@@ -460,7 +474,7 @@ struct radeon_bo_list {
 	struct radeon_bo		*robj;
 	struct ttm_validate_buffer	tv;
 	uint64_t			gpu_offset;
-	unsigned			prefered_domains;
+	unsigned			preferred_domains;
 	unsigned			allowed_domains;
 	uint32_t			tiling_flags;
 };
@@ -497,6 +511,7 @@ struct radeon_bo {
 	u32				tiling_flags;
 	u32				pitch;
 	int				surface_reg;
+	unsigned			prime_shared_count;
 	/* list of all virtual address to which this bo
 	 * is associated to
 	 */
@@ -727,10 +742,6 @@ struct radeon_doorbell {
 
 int radeon_doorbell_get(struct radeon_device *rdev, u32 *page);
 void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell);
-void radeon_doorbell_get_kfd_info(struct radeon_device *rdev,
-				  phys_addr_t *aperture_base,
-				  size_t *aperture_size,
-				  size_t *start_offset);
 
 /*
  * IRQS.
@@ -741,10 +752,12 @@ struct radeon_flip_work {
 	struct work_struct		unpin_work;
 	struct radeon_device		*rdev;
 	int				crtc_id;
+	u32				target_vblank;
 	uint64_t			base;
 	struct drm_pending_vblank_event *event;
 	struct radeon_bo		*old_rbo;
-	struct fence			*fence;
+	struct dma_fence		*fence;
+	bool				async;
 };
 
 struct r500_irq_stat_regs {
@@ -763,24 +776,9 @@ struct r600_irq_stat_regs {
 };
 
 struct evergreen_irq_stat_regs {
-	u32 disp_int;
-	u32 disp_int_cont;
-	u32 disp_int_cont2;
-	u32 disp_int_cont3;
-	u32 disp_int_cont4;
-	u32 disp_int_cont5;
-	u32 d1grph_int;
-	u32 d2grph_int;
-	u32 d3grph_int;
-	u32 d4grph_int;
-	u32 d5grph_int;
-	u32 d6grph_int;
-	u32 afmt_status1;
-	u32 afmt_status2;
-	u32 afmt_status3;
-	u32 afmt_status4;
-	u32 afmt_status5;
-	u32 afmt_status6;
+	u32 disp_int[6];
+	u32 grph_int[6];
+	u32 afmt_status[6];
 };
 
 struct cik_irq_stat_regs {
@@ -1666,20 +1664,28 @@ struct radeon_pm {
 	struct radeon_dpm       dpm;
 };
 
+#define RADEON_PCIE_SPEED_25 1
+#define RADEON_PCIE_SPEED_50 2
+#define RADEON_PCIE_SPEED_80 4
+
 int radeon_pm_get_type_index(struct radeon_device *rdev,
 			     enum radeon_pm_state_type ps_type,
 			     int instance);
 /*
  * UVD
  */
-#define RADEON_MAX_UVD_HANDLES	10
-#define RADEON_UVD_STACK_SIZE	(1024*1024)
-#define RADEON_UVD_HEAP_SIZE	(1024*1024)
+#define RADEON_DEFAULT_UVD_HANDLES	10
+#define RADEON_MAX_UVD_HANDLES		30
+#define RADEON_UVD_STACK_SIZE		(200*1024)
+#define RADEON_UVD_HEAP_SIZE		(256*1024)
+#define RADEON_UVD_SESSION_SIZE		(50*1024)
 
 struct radeon_uvd {
+	bool			fw_header_present;
 	struct radeon_bo	*vcpu_bo;
 	void			*cpu_addr;
 	uint64_t		gpu_addr;
+	unsigned		max_handles;
 	atomic_t		handles[RADEON_MAX_UVD_HANDLES];
 	struct drm_file		*filp[RADEON_MAX_UVD_HANDLES];
 	unsigned		img_size[RADEON_MAX_UVD_HANDLES];
@@ -1853,7 +1859,7 @@ struct radeon_asic {
 	int (*resume)(struct radeon_device *rdev);
 	int (*suspend)(struct radeon_device *rdev);
 	void (*vga_set_state)(struct radeon_device *rdev, bool state);
-	int (*asic_reset)(struct radeon_device *rdev);
+	int (*asic_reset)(struct radeon_device *rdev, bool hard);
 	/* Flush the HDP cache via MMIO */
 	void (*mmio_hdp_flush)(struct radeon_device *rdev);
 	/* check if 3D engine is idle */
@@ -1893,7 +1899,7 @@ struct radeon_asic {
 		void (*pad_ib)(struct radeon_ib *ib);
 	} vm;
 	/* ring specific callbacks */
-	struct radeon_asic_ring *ring[RADEON_NUM_RINGS];
+	const struct radeon_asic_ring *ring[RADEON_NUM_RINGS];
 	/* irqs */
 	struct {
 		int (*set)(struct radeon_device *rdev);
@@ -1999,7 +2005,7 @@ struct radeon_asic {
 	} dpm;
 	/* pageflipping */
 	struct {
-		void (*page_flip)(struct radeon_device *rdev, int crtc, u64 crtc_base);
+		void (*page_flip)(struct radeon_device *rdev, int crtc, u64 crtc_base, bool async);
 		bool (*page_flip_pending)(struct radeon_device *rdev, int crtc);
 	} pflip;
 };
@@ -2361,7 +2367,7 @@ struct radeon_device {
 	uint8_t				*bios;
 	bool				is_atom_bios;
 	uint16_t			bios_header_start;
-	struct radeon_bo		*stollen_vga_memory;
+	struct radeon_bo		*stolen_vga_memory;
 	/* Register mmio */
 	resource_size_t			rmmio_base;
 	resource_size_t			rmmio_size;
@@ -2409,7 +2415,7 @@ struct radeon_device {
 	struct radeon_mman		mman;
 	struct radeon_fence_driver	fence_drv[RADEON_NUM_RINGS];
 	wait_queue_head_t		fence_queue;
-	unsigned			fence_context;
+	u64				fence_context;
 	struct rwlock			ring_lock;
 	struct radeon_ring		ring[RADEON_NUM_RINGS];
 	bool				ib_pool_ready;
@@ -2424,8 +2430,8 @@ struct radeon_device {
 	struct radeon_wb		wb;
 	struct radeon_dummy_page	dummy_page;
 	bool				shutdown;
-	bool				suspend;
 	bool				need_dma32;
+	bool				need_swiotlb;
 	bool				accel_working;
 	bool				fastfb_working; /* IGP feature*/
 	bool				needs_reset, in_reset;
@@ -2453,6 +2459,7 @@ struct radeon_device {
 	int num_crtc; /* number of crtcs */
 	struct rwlock dc_hw_i2c_mutex; /* display controller hw i2c mutex */
 	bool has_uvd;
+	bool has_vce;
 	struct r600_audio audio; /* audio stuff */
 	struct notifier_block acpi_nb;
 	/* only one userspace can use Hyperz features or CMASK at a time */
@@ -2476,8 +2483,6 @@ struct radeon_device {
 	struct radeon_atcs		atcs;
 	/* srbm instance registers */
 	struct rwlock			srbm_mutex;
-	/* GRBM index mutex. Protects concurrents access to GRBM index */
-	struct rwlock			grbm_idx_mutex;
 	/* clock, powergating flags */
 	u32 cg_flags;
 	u32 pg_flags;
@@ -2492,9 +2497,6 @@ struct radeon_device {
 	u64 vram_pin_size;
 	u64 gart_pin_size;
 
-	/* amdkfd interface */
-	struct kfd_dev		*kfd;
-
 	struct rwlock	mn_lock;
 	DECLARE_HASHTABLE(mn_hash, 7);
 };
@@ -2538,9 +2540,9 @@ void cik_mm_wdoorbell(struct radeon_device *rdev, u32 index, u32 v);
 /*
  * Cast helper
  */
-extern const struct fence_ops radeon_fence_ops;
+extern const struct dma_fence_ops radeon_fence_ops;
 
-static inline struct radeon_fence *to_radeon_fence(struct fence *f)
+static inline struct radeon_fence *to_radeon_fence(struct dma_fence *f)
 {
 	struct radeon_fence *__f = container_of(f, struct radeon_fence, base);
 
@@ -2563,7 +2565,8 @@ static inline struct radeon_fence *to_radeon_fence(struct fence *f)
 	bus_space_write_2(rdev->memt, rdev->rmmio_bsh, (reg), (v))
 #define RREG32(reg) r100_mm_rreg(rdev, (reg), false)
 #define RREG32_IDX(reg) r100_mm_rreg(rdev, (reg), true)
-#define DREG32(reg) printk(KERN_INFO "REGISTER: " #reg " : 0x%08X\n", r100_mm_rreg(rdev, (reg), false))
+#define DREG32(reg) pr_info("REGISTER: " #reg " : 0x%08X\n",	\
+			    r100_mm_rreg(rdev, (reg), false))
 #define WREG32(reg, v) r100_mm_wreg(rdev, (reg), (v), false)
 #define WREG32_IDX(reg, v) r100_mm_wreg(rdev, (reg), (v), true)
 #define REG_SET(FIELD, v) (((v) << FIELD##_SHIFT) & FIELD##_MASK)
@@ -2753,7 +2756,7 @@ static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 #define radeon_suspend(rdev) (rdev)->asic->suspend((rdev))
 #define radeon_cs_parse(rdev, r, p) (rdev)->asic->ring[(r)]->cs_parse((p))
 #define radeon_vga_set_state(rdev, state) (rdev)->asic->vga_set_state((rdev), (state))
-#define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev))
+#define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev), false)
 #define radeon_gart_tlb_flush(rdev) (rdev)->asic->gart.tlb_flush((rdev))
 #define radeon_gart_get_page_entry(a, f) (rdev)->asic->gart.get_page_entry((a), (f))
 #define radeon_gart_set_page(rdev, i, e) (rdev)->asic->gart.set_page((rdev), (i), (e))
@@ -2811,7 +2814,7 @@ static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v)
 #define radeon_pm_finish(rdev) (rdev)->asic->pm.finish((rdev))
 #define radeon_pm_init_profile(rdev) (rdev)->asic->pm.init_profile((rdev))
 #define radeon_pm_get_dynpm_state(rdev) (rdev)->asic->pm.get_dynpm_state((rdev))
-#define radeon_page_flip(rdev, crtc, base) (rdev)->asic->pflip.page_flip((rdev), (crtc), (base))
+#define radeon_page_flip(rdev, crtc, base, async) (rdev)->asic->pflip.page_flip((rdev), (crtc), (base), (async))
 #define radeon_page_flip_pending(rdev, crtc) (rdev)->asic->pflip.page_flip_pending((rdev), (crtc))
 #define radeon_wait_for_vblank(rdev, crtc) (rdev)->asic->display.wait_for_vblank((rdev), (crtc))
 #define radeon_mc_wait_for_idle(rdev) (rdev)->asic->mc_wait_for_idle((rdev))
@@ -2868,7 +2871,8 @@ extern bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm);
 extern void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64 base);
 extern void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 extern int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
-extern int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon);
+extern int radeon_suspend_kms(struct drm_device *dev, bool suspend,
+			      bool fbcon, bool freeze);
 extern void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size);
 extern void radeon_program_register_sequence(struct radeon_device *rdev,
 					     const u32 *registers,
@@ -3001,6 +3005,12 @@ int r600_cs_common_vline_parse(struct radeon_cs_parser *p,
 			       uint32_t *vline_start_end,
 			       uint32_t *vline_status);
 
+/* interrupt control register helpers */
+void radeon_irq_kms_set_irq_n_enabled(struct radeon_device *rdev,
+				      u32 reg, u32 mask,
+				      bool enable, const char *name,
+				      unsigned n);
+
 #include "radeon_object.h"
 
 #endif
diff --git a/sys/dev/pci/drm/radeon/radeon_acpi.c b/sys/dev/pci/drm/radeon/radeon_acpi.c
index 4b0858c56f1..8d3251a10cd 100644
--- a/sys/dev/pci/drm/radeon/radeon_acpi.c
+++ b/sys/dev/pci/drm/radeon/radeon_acpi.c
@@ -21,13 +21,24 @@
  *
  */
 
+#include <linux/pci.h>
+#include <linux/acpi.h>
+#include <linux/slab.h>
+#include <linux/power_supply.h>
+#include <linux/pm_runtime.h>
 #include <acpi/video.h>
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
 #include "radeon.h"
 #include "radeon_acpi.h"
 #include "atom.h"
 
+#if defined(CONFIG_VGA_SWITCHEROO)
+bool radeon_atpx_dgpu_req_power_for_displays(void);
+#else
+static inline bool radeon_atpx_dgpu_req_power_for_displays(void) { return false; }
+#endif
+
 #define ACPI_AC_CLASS           "ac_adapter"
 
 extern void radeon_pm_acpi_event_handler(struct radeon_device *rdev);
@@ -340,7 +351,7 @@ out:
  * handles it.
  * Returns NOTIFY code
  */
-int radeon_atif_handler(struct radeon_device *rdev,
+static int radeon_atif_handler(struct radeon_device *rdev,
 		struct acpi_bus_event *event)
 {
 	struct radeon_atif *atif = &rdev->atif;
@@ -390,6 +401,16 @@ int radeon_atif_handler(struct radeon_device *rdev,
 #endif
 		}
 	}
+	if (req.pending & ATIF_DGPU_DISPLAY_EVENT) {
+		if ((rdev->flags & RADEON_IS_PX) &&
+		    radeon_atpx_dgpu_req_power_for_displays()) {
+			pm_runtime_get_sync(rdev->ddev->dev);
+			/* Just fire off a uevent and let userspace tell us what to do */
+			drm_helper_hpd_irq_event(rdev->ddev);
+			pm_runtime_mark_last_busy(rdev->ddev->dev);
+			pm_runtime_put_autosuspend(rdev->ddev->dev);
+		}
+	}
 	/* TODO: check other events */
 
 	/* We've handled the event, stop the notifier chain. The ACPI interface
@@ -737,13 +758,6 @@ int radeon_acpi_init(struct radeon_device *rdev)
 		}
 
 		atif->encoder_for_bl = target;
-		if (!target) {
-			/* Brightness change notification is enabled, but we
-			 * didn't find a backlight controller, this should
-			 * never happen.
-			 */
-			DRM_ERROR("Cannot find a backlight controller\n");
-		}
 	}
 
 	if (atif->functions.sbios_requests && !atif->functions.system_params) {
diff --git a/sys/dev/pci/drm/radeon/radeon_acpi.h b/sys/dev/pci/drm/radeon/radeon_acpi.h
index be4af76f213..35202a453e6 100644
--- a/sys/dev/pci/drm/radeon/radeon_acpi.h
+++ b/sys/dev/pci/drm/radeon/radeon_acpi.h
@@ -27,9 +27,6 @@
 struct radeon_device;
 struct acpi_bus_event;
 
-int radeon_atif_handler(struct radeon_device *rdev,
-		struct acpi_bus_event *event);
-
 /* AMD hw uses four ACPI control methods:
  * 1. ATIF
  * ARG0: (ACPI_INTEGER) function code
@@ -97,6 +94,7 @@ int radeon_atif_handler(struct radeon_device *rdev,
 #       define ATIF_SET_PANEL_EXPANSION_MODE_IN_CMOS_SUPPORTED    (1 << 7)
 #       define ATIF_TEMPERATURE_CHANGE_NOTIFICATION_SUPPORTED     (1 << 12)
 #       define ATIF_GET_GRAPHICS_DEVICE_TYPES_SUPPORTED           (1 << 14)
+#       define ATIF_GET_EXTERNAL_GPU_INFORMATION_SUPPORTED        (1 << 20)
 #define ATIF_FUNCTION_GET_SYSTEM_PARAMETERS                        0x1
 /* ARG0: ATIF_FUNCTION_GET_SYSTEM_PARAMETERS
  * ARG1: none
@@ -251,6 +249,17 @@ int radeon_atif_handler(struct radeon_device *rdev,
 #       define ATIF_XGP_PORT                                       (1 << 1)
 #       define ATIF_VGA_ENABLED_GRAPHICS_DEVICE                    (1 << 2)
 #       define ATIF_XGP_PORT_IN_DOCK                               (1 << 3)
+#define ATIF_FUNCTION_GET_EXTERNAL_GPU_INFORMATION                 0x15
+/* ARG0: ATIF_FUNCTION_GET_EXTERNAL_GPU_INFORMATION
+ * ARG1: none
+ * OUTPUT:
+ * WORD  - number of reported external gfx devices
+ * WORD  - device structure size in bytes (excludes device size field)
+ * WORD  - flags         \
+ * WORD  - bus number    / repeated structure
+ */
+/* flags */
+#       define ATIF_EXTERNAL_GRAPHICS_PORT                         (1 << 0)
 
 /* ATPX */
 #define ATPX_FUNCTION_VERIFY_INTERFACE                             0x0
@@ -291,6 +300,8 @@ int radeon_atif_handler(struct radeon_device *rdev,
 #       define ATPX_FIXED_NOT_SUPPORTED                            (1 << 9)
 #       define ATPX_DYNAMIC_DGPU_POWER_OFF_SUPPORTED               (1 << 10)
 #       define ATPX_DGPU_REQ_POWER_FOR_DISPLAYS                    (1 << 11)
+#       define ATPX_DGPU_CAN_DRIVE_DISPLAYS                        (1 << 12)
+#       define ATPX_MS_HYBRID_GFX_SUPPORTED                        (1 << 14)
 #define ATPX_FUNCTION_POWER_CONTROL                                0x2
 /* ARG0: ATPX_FUNCTION_POWER_CONTROL
  * ARG1:
diff --git a/sys/dev/pci/drm/radeon/radeon_agp.c b/sys/dev/pci/drm/radeon/radeon_agp.c
index 71192113c6f..47ef96bec3e 100644
--- a/sys/dev/pci/drm/radeon/radeon_agp.c
+++ b/sys/dev/pci/drm/radeon/radeon_agp.c
@@ -24,11 +24,11 @@
  *    Dave Airlie
  *    Jerome Glisse <glisse@freedesktop.org>
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 
 struct radeon_agpmode_quirk {
 	u32 hostbridge_vendor;
@@ -126,7 +126,7 @@ static struct radeon_agpmode_quirk radeon_agpmode_quirk_list[] = {
 
 int radeon_agp_init(struct radeon_device *rdev)
 {
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 	struct radeon_agpmode_quirk *p = radeon_agpmode_quirk_list;
 	struct drm_agp_mode mode;
 	struct drm_agp_info info;
@@ -267,7 +267,7 @@ int radeon_agp_init(struct radeon_device *rdev)
 
 void radeon_agp_resume(struct radeon_device *rdev)
 {
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 	int r;
 	if (rdev->flags & RADEON_IS_AGP) {
 		r = radeon_agp_init(rdev);
@@ -279,7 +279,7 @@ void radeon_agp_resume(struct radeon_device *rdev)
 
 void radeon_agp_fini(struct radeon_device *rdev)
 {
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 	if (rdev->ddev->agp && rdev->ddev->agp->acquired) {
 		drm_agp_release(rdev->ddev);
 	}
diff --git a/sys/dev/pci/drm/radeon/radeon_asic.c b/sys/dev/pci/drm/radeon/radeon_asic.c
index d2d6a347846..bc5121d1a7b 100644
--- a/sys/dev/pci/drm/radeon/radeon_asic.c
+++ b/sys/dev/pci/drm/radeon/radeon_asic.c
@@ -26,9 +26,11 @@
  *          Jerome Glisse
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/console.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
+#include <linux/vgaarb.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -177,7 +179,7 @@ void radeon_agp_disable(struct radeon_device *rdev)
  * ASIC
  */
 
-static struct radeon_asic_ring r100_gfx_ring = {
+static const struct radeon_asic_ring r100_gfx_ring = {
 	.ib_execute = &r100_ring_ib_execute,
 	.emit_fence = &r100_fence_ring_emit,
 	.emit_semaphore = &r100_semaphore_ring_emit,
@@ -327,7 +329,7 @@ static struct radeon_asic r200_asic = {
 	},
 };
 
-static struct radeon_asic_ring r300_gfx_ring = {
+static const struct radeon_asic_ring r300_gfx_ring = {
 	.ib_execute = &r100_ring_ib_execute,
 	.emit_fence = &r300_fence_ring_emit,
 	.emit_semaphore = &r100_semaphore_ring_emit,
@@ -341,7 +343,7 @@ static struct radeon_asic_ring r300_gfx_ring = {
 	.set_wptr = &r100_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring rv515_gfx_ring = {
+static const struct radeon_asic_ring rv515_gfx_ring = {
 	.ib_execute = &r100_ring_ib_execute,
 	.emit_fence = &r300_fence_ring_emit,
 	.emit_semaphore = &r100_semaphore_ring_emit,
@@ -899,7 +901,7 @@ static struct radeon_asic r520_asic = {
 	},
 };
 
-static struct radeon_asic_ring r600_gfx_ring = {
+static const struct radeon_asic_ring r600_gfx_ring = {
 	.ib_execute = &r600_ring_ib_execute,
 	.emit_fence = &r600_fence_ring_emit,
 	.emit_semaphore = &r600_semaphore_ring_emit,
@@ -912,7 +914,7 @@ static struct radeon_asic_ring r600_gfx_ring = {
 	.set_wptr = &r600_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring r600_dma_ring = {
+static const struct radeon_asic_ring r600_dma_ring = {
 	.ib_execute = &r600_dma_ring_ib_execute,
 	.emit_fence = &r600_dma_fence_ring_emit,
 	.emit_semaphore = &r600_dma_semaphore_ring_emit,
@@ -997,7 +999,7 @@ static struct radeon_asic r600_asic = {
 	},
 };
 
-static struct radeon_asic_ring rv6xx_uvd_ring = {
+static const struct radeon_asic_ring rv6xx_uvd_ring = {
 	.ib_execute = &uvd_v1_0_ib_execute,
 	.emit_fence = &uvd_v1_0_fence_emit,
 	.emit_semaphore = &uvd_v1_0_semaphore_emit,
@@ -1196,7 +1198,7 @@ static struct radeon_asic rs780_asic = {
 	},
 };
 
-static struct radeon_asic_ring rv770_uvd_ring = {
+static const struct radeon_asic_ring rv770_uvd_ring = {
 	.ib_execute = &uvd_v1_0_ib_execute,
 	.emit_fence = &uvd_v2_2_fence_emit,
 	.emit_semaphore = &uvd_v2_2_semaphore_emit,
@@ -1303,7 +1305,7 @@ static struct radeon_asic rv770_asic = {
 	},
 };
 
-static struct radeon_asic_ring evergreen_gfx_ring = {
+static const struct radeon_asic_ring evergreen_gfx_ring = {
 	.ib_execute = &evergreen_ring_ib_execute,
 	.emit_fence = &r600_fence_ring_emit,
 	.emit_semaphore = &r600_semaphore_ring_emit,
@@ -1316,7 +1318,7 @@ static struct radeon_asic_ring evergreen_gfx_ring = {
 	.set_wptr = &r600_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring evergreen_dma_ring = {
+static const struct radeon_asic_ring evergreen_dma_ring = {
 	.ib_execute = &evergreen_dma_ring_ib_execute,
 	.emit_fence = &evergreen_dma_fence_ring_emit,
 	.emit_semaphore = &r600_dma_semaphore_ring_emit,
@@ -1610,7 +1612,7 @@ static struct radeon_asic btc_asic = {
 	},
 };
 
-static struct radeon_asic_ring cayman_gfx_ring = {
+static const struct radeon_asic_ring cayman_gfx_ring = {
 	.ib_execute = &cayman_ring_ib_execute,
 	.ib_parse = &evergreen_ib_parse,
 	.emit_fence = &cayman_fence_ring_emit,
@@ -1625,7 +1627,7 @@ static struct radeon_asic_ring cayman_gfx_ring = {
 	.set_wptr = &cayman_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring cayman_dma_ring = {
+static const struct radeon_asic_ring cayman_dma_ring = {
 	.ib_execute = &cayman_dma_ring_ib_execute,
 	.ib_parse = &evergreen_dma_ib_parse,
 	.emit_fence = &evergreen_dma_fence_ring_emit,
@@ -1640,7 +1642,7 @@ static struct radeon_asic_ring cayman_dma_ring = {
 	.set_wptr = &cayman_dma_set_wptr
 };
 
-static struct radeon_asic_ring cayman_uvd_ring = {
+static const struct radeon_asic_ring cayman_uvd_ring = {
 	.ib_execute = &uvd_v1_0_ib_execute,
 	.emit_fence = &uvd_v2_2_fence_emit,
 	.emit_semaphore = &uvd_v3_1_semaphore_emit,
@@ -1758,7 +1760,7 @@ static struct radeon_asic cayman_asic = {
 	},
 };
 
-static struct radeon_asic_ring trinity_vce_ring = {
+static const struct radeon_asic_ring trinity_vce_ring = {
 	.ib_execute = &radeon_vce_ib_execute,
 	.emit_fence = &radeon_vce_fence_emit,
 	.emit_semaphore = &radeon_vce_semaphore_emit,
@@ -1879,7 +1881,7 @@ static struct radeon_asic trinity_asic = {
 	},
 };
 
-static struct radeon_asic_ring si_gfx_ring = {
+static const struct radeon_asic_ring si_gfx_ring = {
 	.ib_execute = &si_ring_ib_execute,
 	.ib_parse = &si_ib_parse,
 	.emit_fence = &si_fence_ring_emit,
@@ -1894,7 +1896,7 @@ static struct radeon_asic_ring si_gfx_ring = {
 	.set_wptr = &cayman_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring si_dma_ring = {
+static const struct radeon_asic_ring si_dma_ring = {
 	.ib_execute = &cayman_dma_ring_ib_execute,
 	.ib_parse = &evergreen_dma_ib_parse,
 	.emit_fence = &evergreen_dma_fence_ring_emit,
@@ -2021,7 +2023,7 @@ static struct radeon_asic si_asic = {
 	},
 };
 
-static struct radeon_asic_ring ci_gfx_ring = {
+static const struct radeon_asic_ring ci_gfx_ring = {
 	.ib_execute = &cik_ring_ib_execute,
 	.ib_parse = &cik_ib_parse,
 	.emit_fence = &cik_fence_gfx_ring_emit,
@@ -2036,7 +2038,7 @@ static struct radeon_asic_ring ci_gfx_ring = {
 	.set_wptr = &cik_gfx_set_wptr,
 };
 
-static struct radeon_asic_ring ci_cp_ring = {
+static const struct radeon_asic_ring ci_cp_ring = {
 	.ib_execute = &cik_ring_ib_execute,
 	.ib_parse = &cik_ib_parse,
 	.emit_fence = &cik_fence_compute_ring_emit,
@@ -2051,7 +2053,7 @@ static struct radeon_asic_ring ci_cp_ring = {
 	.set_wptr = &cik_compute_set_wptr,
 };
 
-static struct radeon_asic_ring ci_dma_ring = {
+static const struct radeon_asic_ring ci_dma_ring = {
 	.ib_execute = &cik_sdma_ring_ib_execute,
 	.ib_parse = &cik_ib_parse,
 	.emit_fence = &cik_sdma_fence_ring_emit,
@@ -2066,7 +2068,7 @@ static struct radeon_asic_ring ci_dma_ring = {
 	.set_wptr = &cik_sdma_set_wptr,
 };
 
-static struct radeon_asic_ring ci_vce_ring = {
+static const struct radeon_asic_ring ci_vce_ring = {
 	.ib_execute = &radeon_vce_ib_execute,
 	.emit_fence = &radeon_vce_fence_emit,
 	.emit_semaphore = &radeon_vce_semaphore_emit,
@@ -2322,6 +2324,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 		rdev->num_crtc = 2;
 
 	rdev->has_uvd = false;
+	rdev->has_vce = false;
 
 	switch (rdev->family) {
 	case CHIP_R100:
@@ -2452,6 +2455,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 		/* set num crtcs */
 		rdev->num_crtc = 4;
 		rdev->has_uvd = true;
+		rdev->has_vce = true;
 		rdev->cg_flags =
 			RADEON_CG_SUPPORT_VCE_MGCG;
 		break;
@@ -2468,10 +2472,13 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->num_crtc = 2;
 		else
 			rdev->num_crtc = 6;
-		if (rdev->family == CHIP_HAINAN)
+		if (rdev->family == CHIP_HAINAN) {
 			rdev->has_uvd = false;
-		else
+			rdev->has_vce = false;
+		} else {
 			rdev->has_uvd = true;
+			rdev->has_vce = true;
+		}
 		switch (rdev->family) {
 		case CHIP_TAHITI:
 			rdev->cg_flags =
@@ -2576,6 +2583,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 		rdev->asic = &ci_asic;
 		rdev->num_crtc = 6;
 		rdev->has_uvd = true;
+		rdev->has_vce = true;
 		if (rdev->family == CHIP_BONAIRE) {
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
@@ -2676,6 +2684,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 				RADEON_PG_SUPPORT_SAMU;*/
 		}
 		rdev->has_uvd = true;
+		rdev->has_vce = true;
 		break;
 	default:
 		/* FIXME: not supported yet */
@@ -2687,6 +2696,11 @@ int radeon_asic_init(struct radeon_device *rdev)
 		rdev->asic->pm.set_memory_clock = NULL;
 	}
 
+	if (!radeon_uvd)
+		rdev->has_uvd = false;
+	if (!radeon_vce)
+		rdev->has_vce = false;
+
 	return 0;
 }
 
diff --git a/sys/dev/pci/drm/radeon/radeon_asic.h b/sys/dev/pci/drm/radeon/radeon_asic.h
index e0aa33262ea..e3f036c20d6 100644
--- a/sys/dev/pci/drm/radeon/radeon_asic.h
+++ b/sys/dev/pci/drm/radeon/radeon_asic.h
@@ -64,7 +64,7 @@ int r100_suspend(struct radeon_device *rdev);
 int r100_resume(struct radeon_device *rdev);
 void r100_vga_set_state(struct radeon_device *rdev, bool state);
 bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
-int r100_asic_reset(struct radeon_device *rdev);
+int r100_asic_reset(struct radeon_device *rdev, bool hard);
 u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc);
 void r100_pci_gart_tlb_flush(struct radeon_device *rdev);
 uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags);
@@ -138,7 +138,7 @@ extern void r100_pm_finish(struct radeon_device *rdev);
 extern void r100_pm_init_profile(struct radeon_device *rdev);
 extern void r100_pm_get_dynpm_state(struct radeon_device *rdev);
 extern void r100_page_flip(struct radeon_device *rdev, int crtc,
-			   u64 crtc_base);
+			   u64 crtc_base, bool async);
 extern bool r100_page_flip_pending(struct radeon_device *rdev, int crtc);
 extern void r100_wait_for_vblank(struct radeon_device *rdev, int crtc);
 extern int r100_mc_wait_for_idle(struct radeon_device *rdev);
@@ -167,7 +167,7 @@ extern int r300_init(struct radeon_device *rdev);
 extern void r300_fini(struct radeon_device *rdev);
 extern int r300_suspend(struct radeon_device *rdev);
 extern int r300_resume(struct radeon_device *rdev);
-extern int r300_asic_reset(struct radeon_device *rdev);
+extern int r300_asic_reset(struct radeon_device *rdev, bool hard);
 extern void r300_ring_start(struct radeon_device *rdev, struct radeon_ring *ring);
 extern void r300_fence_ring_emit(struct radeon_device *rdev,
 				struct radeon_fence *fence);
@@ -225,7 +225,7 @@ extern int rs400_mc_wait_for_idle(struct radeon_device *rdev);
 /*
  * rs600.
  */
-extern int rs600_asic_reset(struct radeon_device *rdev);
+extern int rs600_asic_reset(struct radeon_device *rdev, bool hard);
 extern int rs600_init(struct radeon_device *rdev);
 extern void rs600_fini(struct radeon_device *rdev);
 extern int rs600_suspend(struct radeon_device *rdev);
@@ -250,7 +250,7 @@ extern void rs600_pm_misc(struct radeon_device *rdev);
 extern void rs600_pm_prepare(struct radeon_device *rdev);
 extern void rs600_pm_finish(struct radeon_device *rdev);
 extern void rs600_page_flip(struct radeon_device *rdev, int crtc,
-			    u64 crtc_base);
+			    u64 crtc_base, bool async);
 extern bool rs600_page_flip_pending(struct radeon_device *rdev, int crtc);
 void rs600_set_safe_registers(struct radeon_device *rdev);
 extern void avivo_wait_for_vblank(struct radeon_device *rdev, int crtc);
@@ -334,7 +334,7 @@ bool r600_dma_semaphore_ring_emit(struct radeon_device *rdev,
 void r600_dma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 bool r600_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring);
 bool r600_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
-int r600_asic_reset(struct radeon_device *rdev);
+int r600_asic_reset(struct radeon_device *rdev, bool hard);
 int r600_set_surface_reg(struct radeon_device *rdev, int reg,
 			 uint32_t tiling_flags, uint32_t pitch,
 			 uint32_t offset, uint32_t obj_size);
@@ -464,7 +464,8 @@ void rv770_fini(struct radeon_device *rdev);
 int rv770_suspend(struct radeon_device *rdev);
 int rv770_resume(struct radeon_device *rdev);
 void rv770_pm_misc(struct radeon_device *rdev);
-void rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
+void rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base,
+		     bool async);
 bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc);
 void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 void r700_cp_stop(struct radeon_device *rdev);
@@ -513,7 +514,7 @@ int evergreen_suspend(struct radeon_device *rdev);
 int evergreen_resume(struct radeon_device *rdev);
 bool evergreen_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
 bool evergreen_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
-int evergreen_asic_reset(struct radeon_device *rdev);
+int evergreen_asic_reset(struct radeon_device *rdev, bool hard);
 void evergreen_bandwidth_update(struct radeon_device *rdev);
 void evergreen_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 void evergreen_hpd_init(struct radeon_device *rdev);
@@ -534,7 +535,7 @@ extern void btc_pm_init_profile(struct radeon_device *rdev);
 int sumo_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int evergreen_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 extern void evergreen_page_flip(struct radeon_device *rdev, int crtc,
-				u64 crtc_base);
+				u64 crtc_base, bool async);
 extern bool evergreen_page_flip_pending(struct radeon_device *rdev, int crtc);
 extern void dce4_wait_for_vblank(struct radeon_device *rdev, int crtc);
 void evergreen_disable_interrupt_state(struct radeon_device *rdev);
@@ -606,7 +607,7 @@ int cayman_init(struct radeon_device *rdev);
 void cayman_fini(struct radeon_device *rdev);
 int cayman_suspend(struct radeon_device *rdev);
 int cayman_resume(struct radeon_device *rdev);
-int cayman_asic_reset(struct radeon_device *rdev);
+int cayman_asic_reset(struct radeon_device *rdev, bool hard);
 void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int cayman_vm_init(struct radeon_device *rdev);
 void cayman_vm_fini(struct radeon_device *rdev);
@@ -712,7 +713,7 @@ int si_suspend(struct radeon_device *rdev);
 int si_resume(struct radeon_device *rdev);
 bool si_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
 bool si_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
-int si_asic_reset(struct radeon_device *rdev);
+int si_asic_reset(struct radeon_device *rdev, bool hard);
 void si_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int si_irq_set(struct radeon_device *rdev);
 int si_irq_process(struct radeon_device *rdev);
@@ -817,7 +818,7 @@ void cik_fini(struct radeon_device *rdev);
 int cik_suspend(struct radeon_device *rdev);
 int cik_resume(struct radeon_device *rdev);
 bool cik_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
-int cik_asic_reset(struct radeon_device *rdev);
+int cik_asic_reset(struct radeon_device *rdev, bool hard);
 void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
 int cik_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
 int cik_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
diff --git a/sys/dev/pci/drm/radeon/radeon_atombios.c b/sys/dev/pci/drm/radeon/radeon_atombios.c
index e817bcb2daa..3163a6c09dc 100644
--- a/sys/dev/pci/drm/radeon/radeon_atombios.c
+++ b/sys/dev/pci/drm/radeon/radeon_atombios.c
@@ -23,12 +23,13 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 #include "atom.h"
 #include "atom-bits.h"
+#include "radeon_asic.h"
 
 extern void
 radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum,
@@ -2095,7 +2096,7 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 	struct radeon_i2c_bus_rec i2c_bus;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 
 	if (!atom_parse_data_header(mode_info->atom_context, index, NULL,
@@ -2127,13 +2128,16 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 		num_modes = ATOM_MAX_NUMBEROF_POWER_BLOCK;
 	if (num_modes == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) * num_modes, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(num_modes,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	/* last mode is usually default, array is low to high */
 	for (i = 0; i < num_modes; i++) {
 		rdev->pm.power_state[state_index].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[state_index].clock_info)
 			return state_index;
 		rdev->pm.power_state[state_index].num_clock_modes = 1;
@@ -2579,7 +2583,7 @@ static int radeon_atombios_parse_power_table_4_5(struct radeon_device *rdev)
 	bool valid;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 
 	if (!atom_parse_data_header(mode_info->atom_context, index, NULL,
@@ -2590,8 +2594,9 @@ static int radeon_atombios_parse_power_table_4_5(struct radeon_device *rdev)
 	radeon_atombios_add_pplib_thermal_controller(rdev, &power_info->pplib.sThermalController);
 	if (power_info->pplib.ucNumStates == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) *
-				       power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(power_info->pplib.ucNumStates,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	/* first mode is usually default, followed by low to high */
@@ -2606,10 +2611,11 @@ static int radeon_atombios_parse_power_table_4_5(struct radeon_device *rdev)
 			 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset) +
 			 (power_state->v1.ucNonClockStateIndex *
 			  power_info->pplib.ucNonClockSize));
-		rdev->pm.power_state[i].clock_info = kzalloc(sizeof(struct radeon_pm_clock_info) *
-							     ((power_info->pplib.ucStateEntrySize - 1) ?
-							      (power_info->pplib.ucStateEntrySize - 1) : 1),
-							     GFP_KERNEL);
+		rdev->pm.power_state[i].clock_info =
+			kcalloc((power_info->pplib.ucStateEntrySize - 1) ?
+				(power_info->pplib.ucStateEntrySize - 1) : 1,
+				sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[i].clock_info)
 			return state_index;
 		if (power_info->pplib.ucStateEntrySize - 1) {
@@ -2670,7 +2676,7 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev)
 	bool valid;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 
@@ -2691,8 +2697,9 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev)
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 	if (state_array->ucNumEntries == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) *
-				       state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(state_array->ucNumEntries,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	power_state_offset = (u8 *)state_array->states;
@@ -2702,10 +2709,11 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev)
 		non_clock_array_index = power_state->v2.nonClockInfoIndex;
 		non_clock_info = (struct _ATOM_PPLIB_NONCLOCK_INFO *)
 			&non_clock_info_array->nonClockInfo[non_clock_array_index];
-		rdev->pm.power_state[i].clock_info = kzalloc(sizeof(struct radeon_pm_clock_info) *
-							     (power_state->v2.ucNumDPMLevels ?
-							      power_state->v2.ucNumDPMLevels : 1),
-							     GFP_KERNEL);
+		rdev->pm.power_state[i].clock_info =
+			kcalloc(power_state->v2.ucNumDPMLevels ?
+				power_state->v2.ucNumDPMLevels : 1,
+				sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[i].clock_info)
 			return state_index;
 		if (power_state->v2.ucNumDPMLevels) {
@@ -2785,7 +2793,9 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev)
 		rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state), GFP_KERNEL);
 		if (rdev->pm.power_state) {
 			rdev->pm.power_state[0].clock_info =
-				kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+				kcalloc(1,
+				        sizeof(struct radeon_pm_clock_info),
+				        GFP_KERNEL);
 			if (rdev->pm.power_state[0].clock_info) {
 				/* add the default mode */
 				rdev->pm.power_state[state_index].type =
diff --git a/sys/dev/pci/drm/radeon/radeon_audio.c b/sys/dev/pci/drm/radeon/radeon_audio.c
index 6c9d95c1990..770e31f5fd1 100644
--- a/sys/dev/pci/drm/radeon/radeon_audio.c
+++ b/sys/dev/pci/drm/radeon/radeon_audio.c
@@ -22,8 +22,9 @@
  * Authors: Slava Grigorev <slava.grigorev@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc.h>
+#include <linux/gcd.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
 #include "radeon.h"
 #include "atom.h"
 #include "radeon_audio.h"
@@ -515,7 +516,7 @@ static int radeon_audio_set_avi_packet(struct drm_encoder *encoder,
 	if (!connector)
 		return -EINVAL;
 
-	err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode);
+	err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode, false);
 	if (err < 0) {
 		DRM_ERROR("failed to setup AVI infoframe: %d\n", err);
 		return err;
@@ -575,9 +576,9 @@ static void radeon_audio_calc_cts(unsigned int clock, int *CTS, int *N, int freq
 
 	/* Check that we are in spec (not always possible) */
 	if (n < (128*freq/1500))
-		printk(KERN_WARNING "Calculated ACR N value is too small. You may experience audio problems.\n");
+		pr_warn("Calculated ACR N value is too small. You may experience audio problems.\n");
 	if (n > (128*freq/300))
-		printk(KERN_WARNING "Calculated ACR N value is too large. You may experience audio problems.\n");
+		pr_warn("Calculated ACR N value is too large. You may experience audio problems.\n");
 
 	*N = n;
 	*CTS = cts;
diff --git a/sys/dev/pci/drm/radeon/radeon_audio.h b/sys/dev/pci/drm/radeon/radeon_audio.h
index 7379bd3f13f..5c70cceaa4a 100644
--- a/sys/dev/pci/drm/radeon/radeon_audio.h
+++ b/sys/dev/pci/drm/radeon/radeon_audio.h
@@ -25,7 +25,7 @@
 #ifndef __RADEON_AUDIO_H__
 #define __RADEON_AUDIO_H__
 
-#include <sys/types.h>
+#include <linux/types.h>
 
 #define RREG32_ENDPOINT(block, reg)		\
 	radeon_audio_endpoint_rreg(rdev, (block), (reg))
diff --git a/sys/dev/pci/drm/radeon/radeon_benchmark.c b/sys/dev/pci/drm/radeon/radeon_benchmark.c
index 33a574967b3..ec3111d6872 100644
--- a/sys/dev/pci/drm/radeon/radeon_benchmark.c
+++ b/sys/dev/pci/drm/radeon/radeon_benchmark.c
@@ -21,8 +21,8 @@
  *
  * Authors: Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 
diff --git a/sys/dev/pci/drm/radeon/radeon_bios.c b/sys/dev/pci/drm/radeon/radeon_bios.c
index c35485b6b8b..182c4d87285 100644
--- a/sys/dev/pci/drm/radeon/radeon_bios.c
+++ b/sys/dev/pci/drm/radeon/radeon_bios.c
@@ -25,11 +25,14 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "atom.h"
 
+#include <linux/slab.h>
+#include <linux/acpi.h>
+
 #if defined(__amd64__) || defined(__i386__)
 #include <dev/isa/isareg.h>
 #include <dev/isa/isavar.h>
@@ -742,51 +745,56 @@ static bool radeon_read_disabled_bios(struct radeon_device *rdev)
 #ifdef CONFIG_ACPI
 static bool radeon_acpi_vfct_bios(struct radeon_device *rdev)
 {
-	bool ret = false;
 	struct acpi_table_header *hdr;
 	acpi_size tbl_size;
 	UEFI_ACPI_VFCT *vfct;
-	GOP_VBIOS_CONTENT *vbios;
-	VFCT_IMAGE_HEADER *vhdr;
+	unsigned offset;
 
-	if (!ACPI_SUCCESS(acpi_get_table_with_size("VFCT", 1, &hdr, &tbl_size)))
+	if (!ACPI_SUCCESS(acpi_get_table("VFCT", 1, &hdr)))
 		return false;
+	tbl_size = hdr->length;
 	if (tbl_size < sizeof(UEFI_ACPI_VFCT)) {
 		DRM_ERROR("ACPI VFCT table present but broken (too short #1)\n");
-		goto out_unmap;
+		return false;
 	}
 
 	vfct = (UEFI_ACPI_VFCT *)hdr;
-	if (vfct->VBIOSImageOffset + sizeof(VFCT_IMAGE_HEADER) > tbl_size) {
-		DRM_ERROR("ACPI VFCT table present but broken (too short #2)\n");
-		goto out_unmap;
-	}
+	offset = vfct->VBIOSImageOffset;
 
-	vbios = (GOP_VBIOS_CONTENT *)((char *)hdr + vfct->VBIOSImageOffset);
-	vhdr = &vbios->VbiosHeader;
-	DRM_INFO("ACPI VFCT contains a BIOS for %02x:%02x.%d %04x:%04x, size %d\n",
-			vhdr->PCIBus, vhdr->PCIDevice, vhdr->PCIFunction,
-			vhdr->VendorID, vhdr->DeviceID, vhdr->ImageLength);
+	while (offset < tbl_size) {
+		GOP_VBIOS_CONTENT *vbios = (GOP_VBIOS_CONTENT *)((char *)hdr + offset);
+		VFCT_IMAGE_HEADER *vhdr = &vbios->VbiosHeader;
 
-	if (vhdr->PCIBus != rdev->pdev->bus->number ||
-	    vhdr->PCIDevice != PCI_SLOT(rdev->pdev->devfn) ||
-	    vhdr->PCIFunction != PCI_FUNC(rdev->pdev->devfn) ||
-	    vhdr->VendorID != rdev->pdev->vendor ||
-	    vhdr->DeviceID != rdev->pdev->device) {
-		DRM_INFO("ACPI VFCT table is not for this card\n");
-		goto out_unmap;
-	}
+		offset += sizeof(VFCT_IMAGE_HEADER);
+		if (offset > tbl_size) {
+			DRM_ERROR("ACPI VFCT image header truncated\n");
+			return false;
+		}
 
-	if (vfct->VBIOSImageOffset + sizeof(VFCT_IMAGE_HEADER) + vhdr->ImageLength > tbl_size) {
-		DRM_ERROR("ACPI VFCT image truncated\n");
-		goto out_unmap;
-	}
+		offset += vhdr->ImageLength;
+		if (offset > tbl_size) {
+			DRM_ERROR("ACPI VFCT image truncated\n");
+			return false;
+		}
 
-	rdev->bios = kmemdup(&vbios->VbiosContent, vhdr->ImageLength, GFP_KERNEL);
-	ret = !!rdev->bios;
+		if (vhdr->ImageLength &&
+		    vhdr->PCIBus == rdev->pdev->bus->number &&
+		    vhdr->PCIDevice == PCI_SLOT(rdev->pdev->devfn) &&
+		    vhdr->PCIFunction == PCI_FUNC(rdev->pdev->devfn) &&
+		    vhdr->VendorID == rdev->pdev->vendor &&
+		    vhdr->DeviceID == rdev->pdev->device) {
+			rdev->bios = kmemdup(&vbios->VbiosContent,
+					     vhdr->ImageLength,
+					     GFP_KERNEL);
+
+			if (!rdev->bios)
+				return false;
+			return true;
+		}
+	}
 
-out_unmap:
-	return ret;
+	DRM_ERROR("ACPI VFCT table present but broken (too short #2)\n");
+	return false;
 }
 #else
 static inline bool radeon_acpi_vfct_bios(struct radeon_device *rdev)
diff --git a/sys/dev/pci/drm/radeon/radeon_clocks.c b/sys/dev/pci/drm/radeon/radeon_clocks.c
index a2d1bd915ac..bfd0bb145c7 100644
--- a/sys/dev/pci/drm/radeon/radeon_clocks.c
+++ b/sys/dev/pci/drm/radeon/radeon_clocks.c
@@ -25,10 +25,11 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "atom.h"
 
 #if defined(__macppc__) || defined(__sparc64__)
@@ -116,7 +117,7 @@ static bool radeon_read_clocks_OF(struct drm_device *dev)
 		return false;
 	val = of_get_property(dp, "ATY,RefCLK", NULL);
 	if (!val || !*val) {
-		printk(KERN_WARNING "radeonfb: No ATY,RefCLK property !\n");
+		pr_warn("radeonfb: No ATY,RefCLK property !\n");
 		return false;
 	}
 	p1pll->reference_freq = p2pll->reference_freq = (*val) / 10;
diff --git a/sys/dev/pci/drm/radeon/radeon_combios.c b/sys/dev/pci/drm/radeon/radeon_combios.c
index 749d9ccfe99..ac5426487d8 100644
--- a/sys/dev/pci/drm/radeon/radeon_combios.c
+++ b/sys/dev/pci/drm/radeon/radeon_combios.c
@@ -24,8 +24,8 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "atom.h"
 
@@ -34,7 +34,6 @@
 #include <asm/machdep.h>
 #include <asm/pmac_feature.h>
 #include <asm/prom.h>
-#include <asm/pci-bridge.h>
 #endif /* CONFIG_PPC_PMAC */
 
 /* from radeon_legacy_encoder.c */
@@ -2652,13 +2651,16 @@ void radeon_combios_get_power_modes(struct radeon_device *rdev)
 	rdev->pm.default_power_state_index = -1;
 
 	/* allocate 2 power states */
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) * 2, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(2, sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (rdev->pm.power_state) {
 		/* allocate 1 clock mode per state */
 		rdev->pm.power_state[0].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		rdev->pm.power_state[1].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[0].clock_info ||
 		    !rdev->pm.power_state[1].clock_info)
 			goto pm_failed;
diff --git a/sys/dev/pci/drm/radeon/radeon_connectors.c b/sys/dev/pci/drm/radeon/radeon_connectors.c
index ad69ad9595e..f2ccb598bd9 100644
--- a/sys/dev/pci/drm/radeon/radeon_connectors.c
+++ b/sys/dev/pci/drm/radeon/radeon_connectors.c
@@ -23,16 +23,18 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_edid.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/drm_fb_helper.h>
-#include <dev/pci/drm/drm_dp_mst_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_dp_mst_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_audio.h"
 #include "atom.h"
 
+#include <linux/pm_runtime.h>
+#include <linux/vga_switcheroo.h>
 
 static int radeon_dp_handle_hpd(struct drm_connector *connector)
 {
@@ -189,12 +191,12 @@ int radeon_get_monitor_bpc(struct drm_connector *connector)
 		}
 
 		/* Any defined maximum tmds clock limit we must not exceed? */
-		if (connector->max_tmds_clock > 0) {
+		if (connector->display_info.max_tmds_clock > 0) {
 			/* mode_clock is clock in kHz for mode to be modeset on this connector */
 			mode_clock = radeon_connector->pixelclock_for_modeset;
 
 			/* Maximum allowable input clock in kHz */
-			max_tmds_clock = connector->max_tmds_clock * 1000;
+			max_tmds_clock = connector->display_info.max_tmds_clock;
 
 			DRM_DEBUG("%s: hdmi mode dotclock %d kHz, max tmds input clock %d kHz.\n",
 					  connector->name, mode_clock, max_tmds_clock);
@@ -242,23 +244,15 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 {
 	struct drm_device *dev = connector->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct drm_encoder *best_encoder = NULL;
-	struct drm_encoder *encoder = NULL;
+	struct drm_encoder *best_encoder;
+	struct drm_encoder *encoder;
 	const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private;
 	bool connected;
 	int i;
 
 	best_encoder = connector_funcs->best_encoder(connector);
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev,
-					   connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if ((encoder == best_encoder) && (status == connector_status_connected))
 			connected = true;
 		else
@@ -268,7 +262,6 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 			radeon_atombios_connected_scratch_regs(connector, encoder, connected);
 		else
 			radeon_combios_connected_scratch_regs(connector, encoder, connected);
-
 	}
 }
 
@@ -277,17 +270,11 @@ static struct drm_encoder *radeon_find_encoder(struct drm_connector *connector,
 	struct drm_encoder *encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (encoder->encoder_type == encoder_type)
 			return encoder;
 	}
+
 	return NULL;
 }
 
@@ -336,6 +323,11 @@ static void radeon_connector_get_edid(struct drm_connector *connector)
 		else if (radeon_connector->ddc_bus)
 			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
 							      &radeon_connector->ddc_bus->adapter);
+	} else if (vga_switcheroo_handler_flags() & VGA_SWITCHEROO_CAN_SWITCH_DDC &&
+		   connector->connector_type == DRM_MODE_CONNECTOR_LVDS &&
+		   radeon_connector->ddc_bus) {
+		radeon_connector->edid = drm_get_edid_switcheroo(&radeon_connector->base,
+								 &radeon_connector->ddc_bus->adapter);
 	} else if (radeon_connector->ddc_bus) {
 		radeon_connector->edid = drm_get_edid(&radeon_connector->base,
 						      &radeon_connector->ddc_bus->adapter);
@@ -376,21 +368,23 @@ static int radeon_ddc_get_modes(struct drm_connector *connector)
 	int ret;
 
 	if (radeon_connector->edid) {
-		drm_mode_connector_update_edid_property(connector, radeon_connector->edid);
+		drm_connector_update_edid_property(connector, radeon_connector->edid);
 		ret = drm_add_edid_modes(connector, radeon_connector->edid);
-		drm_edid_to_eld(connector, radeon_connector->edid);
 		return ret;
 	}
-	drm_mode_connector_update_edid_property(connector, NULL);
+	drm_connector_update_edid_property(connector, NULL);
 	return 0;
 }
 
 static struct drm_encoder *radeon_best_single_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
-	/* pick the encoder ids */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+	struct drm_encoder *encoder;
+	int i;
+
+	/* pick the first one */
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -430,19 +424,19 @@ radeon_connector_analog_encoder_conflict_solve(struct drm_connector *connector,
 	struct drm_device *dev = connector->dev;
 	struct drm_connector *conflict;
 	struct radeon_connector *radeon_conflict;
-	int i;
 
 	list_for_each_entry(conflict, &dev->mode_config.connector_list, head) {
+		struct drm_encoder *enc;
+		int i;
+
 		if (conflict == connector)
 			continue;
 
 		radeon_conflict = to_radeon_connector(conflict);
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (conflict->encoder_ids[i] == 0)
-				break;
 
+		drm_connector_for_each_possible_encoder(conflict, enc, i) {
 			/* if the IDs match */
-			if (conflict->encoder_ids[i] == encoder->base.id) {
+			if (enc == encoder) {
 				if (conflict->status != connector_status_connected)
 					continue;
 
@@ -759,12 +753,15 @@ static int radeon_connector_set_property(struct drm_connector *connector, struct
 
 		if (connector->encoder->crtc) {
 			struct drm_crtc *crtc  = connector->encoder->crtc;
-			const struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private;
 			struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 
 			radeon_crtc->output_csc = radeon_encoder->output_csc;
 
-			(*crtc_funcs->load_lut)(crtc);
+			/*
+			 * Our .gamma_set assumes the .gamma_store has been
+			 * prefilled and don't care about its arguments.
+			 */
+			crtc->funcs->gamma_set(crtc, NULL, NULL, NULL, 0, NULL);
 		}
 	}
 
@@ -913,6 +910,16 @@ radeon_lvds_detect(struct drm_connector *connector, bool force)
 	return ret;
 }
 
+static void radeon_connector_unregister(struct drm_connector *connector)
+{
+	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+
+	if (radeon_connector->ddc_bus && radeon_connector->ddc_bus->has_aux) {
+		drm_dp_aux_unregister(&radeon_connector->ddc_bus->aux);
+		radeon_connector->ddc_bus->has_aux = false;
+	}
+}
+
 static void radeon_connector_destroy(struct drm_connector *connector)
 {
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
@@ -970,6 +977,7 @@ static const struct drm_connector_funcs radeon_lvds_connector_funcs = {
 	.dpms = drm_helper_connector_dpms,
 	.detect = radeon_lvds_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.set_property = radeon_lvds_set_property,
 };
@@ -1097,6 +1105,7 @@ static const struct drm_connector_funcs radeon_vga_connector_funcs = {
 	.dpms = drm_helper_connector_dpms,
 	.detect = radeon_vga_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.set_property = radeon_connector_set_property,
 };
@@ -1174,6 +1183,7 @@ static const struct drm_connector_funcs radeon_tv_connector_funcs = {
 	.dpms = drm_helper_connector_dpms,
 	.detect = radeon_tv_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.set_property = radeon_connector_set_property,
 };
@@ -1218,7 +1228,7 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder = NULL;
 	const struct drm_encoder_helper_funcs *encoder_funcs;
-	int i, r;
+	int r;
 	enum drm_connector_status ret = connector_status_disconnected;
 	bool dret = false, broken_edid = false;
 
@@ -1334,15 +1344,9 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 
 	/* find analog encoder */
 	if (radeon_connector->dac_load_detect) {
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (connector->encoder_ids[i] == 0)
-				break;
-
-			encoder = drm_encoder_find(connector->dev,
-						   connector->encoder_ids[i]);
-			if (!encoder)
-				continue;
+		int i;
 
+		drm_connector_for_each_possible_encoder(connector, encoder, i) {
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC &&
 			    encoder->encoder_type != DRM_MODE_ENCODER_TVDAC)
 				continue;
@@ -1416,18 +1420,11 @@ exit:
 /* okay need to be smart in here about which encoder to pick */
 static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder;
 	int i;
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
 
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (radeon_connector->use_digital == true) {
 			if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS)
 				return encoder;
@@ -1442,8 +1439,9 @@ static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 
 	/* then check use digitial */
 	/* pick the first one */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, enc_id);
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -1505,6 +1503,7 @@ static const struct drm_connector_funcs radeon_dvi_connector_funcs = {
 	.detect = radeon_dvi_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_connector_set_property,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
@@ -1585,14 +1584,7 @@ u16 radeon_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 	struct radeon_encoder *radeon_encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		radeon_encoder = to_radeon_encoder(encoder);
 
 		switch (radeon_encoder->encoder_id) {
@@ -1614,14 +1606,7 @@ static bool radeon_connector_encoder_is_hbr2(struct drm_connector *connector)
 	int i;
 	bool found = false;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		radeon_encoder = to_radeon_encoder(encoder);
 		if (radeon_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2)
 			found = true;
@@ -1818,6 +1803,7 @@ static const struct drm_connector_funcs radeon_dp_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_connector_set_property,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
@@ -1827,6 +1813,7 @@ static const struct drm_connector_funcs radeon_edp_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_lvds_set_property,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
@@ -1836,6 +1823,7 @@ static const struct drm_connector_funcs radeon_lvds_bridge_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_lvds_set_property,
+	.early_unregister = radeon_connector_unregister,
 	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
diff --git a/sys/dev/pci/drm/radeon/radeon_cs.c b/sys/dev/pci/drm/radeon/radeon_cs.c
index 6f35c794d11..80fcd5a4939 100644
--- a/sys/dev/pci/drm/radeon/radeon_cs.c
+++ b/sys/dev/pci/drm/radeon/radeon_cs.c
@@ -24,8 +24,9 @@
  * Authors:
  *    Jerome Glisse <glisse@freedesktop.org>
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/list_sort.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_trace.h"
@@ -73,7 +74,6 @@ static void radeon_cs_buckets_get_list(struct radeon_cs_buckets *b,
 
 static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 {
-	struct drm_device *ddev = p->rdev->ddev;
 	struct radeon_cs_chunk *chunk;
 	struct radeon_cs_buckets buckets;
 	unsigned i;
@@ -87,7 +87,8 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	p->dma_reloc_idx = 0;
 	/* FIXME: we assume that each relocs use 4 dwords */
 	p->nrelocs = chunk->length_dw / 4;
-	p->relocs = drm_calloc_large(p->nrelocs, sizeof(struct radeon_bo_list));
+	p->relocs = kvmalloc_array(p->nrelocs, sizeof(struct radeon_bo_list),
+			GFP_KERNEL | __GFP_ZERO);
 	if (p->relocs == NULL) {
 		return -ENOMEM;
 	}
@@ -100,7 +101,7 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 		unsigned priority;
 
 		r = (struct drm_radeon_cs_reloc *)&chunk->kdata[i*4];
-		gobj = drm_gem_object_lookup(ddev, p->filp, r->handle);
+		gobj = drm_gem_object_lookup(p->filp, r->handle);
 		if (gobj == NULL) {
 			DRM_ERROR("gem object lookup failed 0x%x\n",
 				  r->handle);
@@ -117,16 +118,18 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 		priority = (r->flags & RADEON_RELOC_PRIO_MASK) * 2
 			   + !!r->write_domain;
 
-		/* the first reloc of an UVD job is the msg and that must be in
-		   VRAM, also but everything into VRAM on AGP cards and older
-		   IGP chips to avoid image corruptions */
+		/* The first reloc of an UVD job is the msg and that must be in
+		 * VRAM, the second reloc is the DPB and for WMV that must be in
+		 * VRAM as well. Also put everything into VRAM on AGP cards and older
+		 * IGP chips to avoid image corruptions
+		 */
 		if (p->ring == R600_RING_TYPE_UVD_INDEX &&
-		    (i == 0 || (p->rdev->flags & RADEON_IS_AGP) ||
+		    (i <= 0 || (p->rdev->flags & RADEON_IS_AGP) ||
 		     p->rdev->family == CHIP_RS780 ||
 		     p->rdev->family == CHIP_RS880)) {
 
 			/* TODO: is this still needed for NI+ ? */
-			p->relocs[i].prefered_domains =
+			p->relocs[i].preferred_domains =
 				RADEON_GEM_DOMAIN_VRAM;
 
 			p->relocs[i].allowed_domains =
@@ -144,14 +147,14 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 				return -EINVAL;
 			}
 
-			p->relocs[i].prefered_domains = domain;
+			p->relocs[i].preferred_domains = domain;
 			if (domain == RADEON_GEM_DOMAIN_VRAM)
 				domain |= RADEON_GEM_DOMAIN_GTT;
 			p->relocs[i].allowed_domains = domain;
 		}
 
 		if (radeon_ttm_tt_has_userptr(p->relocs[i].robj->tbo.ttm)) {
-			uint32_t domain = p->relocs[i].prefered_domains;
+			uint32_t domain = p->relocs[i].preferred_domains;
 			if (!(domain & RADEON_GEM_DOMAIN_GTT)) {
 				DRM_ERROR("Only RADEON_GEM_DOMAIN_GTT is "
 					  "allowed for userptr BOs\n");
@@ -159,10 +162,20 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			}
 			need_mmap_lock = true;
 			domain = RADEON_GEM_DOMAIN_GTT;
-			p->relocs[i].prefered_domains = domain;
+			p->relocs[i].preferred_domains = domain;
 			p->relocs[i].allowed_domains = domain;
 		}
 
+		/* Objects shared as dma-bufs cannot be moved to VRAM */
+		if (p->relocs[i].robj->prime_shared_count) {
+			p->relocs[i].allowed_domains &= ~RADEON_GEM_DOMAIN_VRAM;
+			if (!p->relocs[i].allowed_domains) {
+				DRM_ERROR("BO associated with dma-buf cannot "
+					  "be moved to VRAM\n");
+				return -EINVAL;
+			}
+		}
+
 		p->relocs[i].tv.bo = &p->relocs[i].robj->tbo;
 		p->relocs[i].tv.shared = !r->write_domain;
 
@@ -332,7 +345,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 				continue;
 		}
 
-		p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t));
+		p->chunks[i].kdata = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
 		size *= sizeof(uint32_t);
 		if (p->chunks[i].kdata == NULL) {
 			return -ENOMEM;
@@ -427,14 +440,14 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 			if (bo == NULL)
 				continue;
 
-			drm_gem_object_unreference_unlocked(&bo->gem_base);
+			drm_gem_object_put_unlocked(&bo->gem_base);
 		}
 	}
 	kfree(parser->track);
-	drm_free_large(parser->relocs);
-	drm_free_large(parser->vm_bos);
+	kvfree(parser->relocs);
+	kvfree(parser->vm_bos);
 	for (i = 0; i < parser->nchunks; i++)
-		drm_free_large(parser->chunks[i].kdata);
+		kvfree(parser->chunks[i].kdata);
 	kfree(parser->chunks);
 	kfree(parser->chunks_array);
 	radeon_ib_free(parser->rdev, &parser->ib);
diff --git a/sys/dev/pci/drm/radeon/radeon_cursor.c b/sys/dev/pci/drm/radeon/radeon_cursor.c
index 2f4e73496d0..91952277557 100644
--- a/sys/dev/pci/drm/radeon/radeon_cursor.c
+++ b/sys/dev/pci/drm/radeon/radeon_cursor.c
@@ -23,8 +23,8 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 static void radeon_lock_cursor(struct drm_crtc *crtc, bool lock)
@@ -298,7 +298,7 @@ int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
 		return -EINVAL;
 	}
 
-	obj = drm_gem_object_lookup(crtc->dev, file_priv, handle);
+	obj = drm_gem_object_lookup(file_priv, handle);
 	if (!obj) {
 		DRM_ERROR("Cannot find cursor object %x for crtc %d\n", handle, radeon_crtc->crtc_id);
 		return -ENOENT;
@@ -307,7 +307,7 @@ int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
 	robj = gem_to_radeon_bo(obj);
 	ret = radeon_bo_reserve(robj, false);
 	if (ret != 0) {
-		drm_gem_object_unreference_unlocked(obj);
+		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
 	/* Only 27 bit offset for legacy cursor */
@@ -317,7 +317,7 @@ int radeon_crtc_cursor_set2(struct drm_crtc *crtc,
 	radeon_bo_unreserve(robj);
 	if (ret) {
 		DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
-		drm_gem_object_unreference_unlocked(obj);
+		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
 
@@ -352,7 +352,7 @@ unpin:
 			radeon_bo_unpin(robj);
 			radeon_bo_unreserve(robj);
 		}
-		drm_gem_object_unreference_unlocked(radeon_crtc->cursor_bo);
+		drm_gem_object_put_unlocked(radeon_crtc->cursor_bo);
 	}
 
 	radeon_crtc->cursor_bo = obj;
diff --git a/sys/dev/pci/drm/radeon/radeon_device.c b/sys/dev/pci/drm/radeon/radeon_device.c
index 33d4ea7d8af..67cd6d06352 100644
--- a/sys/dev/pci/drm/radeon/radeon_device.c
+++ b/sys/dev/pci/drm/radeon/radeon_device.c
@@ -25,9 +25,16 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/console.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_cache.h>
+#include <drm/radeon_drm.h>
+#include <linux/pm_runtime.h>
+#include <linux/vgaarb.h>
+#include <linux/vga_switcheroo.h>
+#include <linux/efi.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "atom.h"
@@ -98,8 +105,15 @@ static const char radeon_family_name[][16] = {
 	"LAST",
 };
 
+#if defined(CONFIG_VGA_SWITCHEROO)
+bool radeon_has_atpx_dgpu_power_cntl(void);
+bool radeon_is_atpx_hybrid(void);
+#else
+static inline bool radeon_has_atpx_dgpu_power_cntl(void) { return false; }
+static inline bool radeon_is_atpx_hybrid(void) { return false; }
+#endif
+
 #define RADEON_PX_QUIRK_DISABLE_PX  (1 << 0)
-#define RADEON_PX_QUIRK_LONG_WAKEUP (1 << 1)
 
 struct radeon_px_quirk {
 	u32 chip_vendor;
@@ -126,8 +140,10 @@ static struct radeon_px_quirk radeon_px_quirk_list[] = {
 	 * https://bugs.freedesktop.org/show_bug.cgi?id=101491
 	 */
 	{ PCI_VENDOR_ID_ATI, 0x6741, 0x1043, 0x2122, RADEON_PX_QUIRK_DISABLE_PX },
-	/* macbook pro 8.2 */
-	{ PCI_VENDOR_ID_ATI, 0x6741, PCI_VENDOR_ID_APPLE, 0x00e2, RADEON_PX_QUIRK_LONG_WAKEUP },
+	/* Asus K73TK laptop with AMD A6-3420M APU and Radeon 7670m GPU
+	 * https://bugzilla.kernel.org/show_bug.cgi?id=51381#c52
+	 */
+	{ PCI_VENDOR_ID_ATI, 0x6840, 0x1043, 0x2123, RADEON_PX_QUIRK_DISABLE_PX },
 	{ 0, 0, 0, 0, 0 },
 };
 
@@ -158,6 +174,11 @@ static void radeon_device_handle_px_quirks(struct radeon_device *rdev)
 
 	if (rdev->px_quirk_flags & RADEON_PX_QUIRK_DISABLE_PX)
 		rdev->flags &= ~RADEON_IS_PX;
+
+	/* disable PX is the system doesn't support dGPU power control or hybrid gfx */
+	if (!radeon_is_atpx_hybrid() &&
+	    !radeon_has_atpx_dgpu_power_cntl())
+		rdev->flags &= ~RADEON_IS_PX;
 }
 
 /**
@@ -387,37 +408,6 @@ void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell)
 		__clear_bit(doorbell, rdev->doorbell.used);
 }
 
-/**
- * radeon_doorbell_get_kfd_info - Report doorbell configuration required to
- *                                setup KFD
- *
- * @rdev: radeon_device pointer
- * @aperture_base: output returning doorbell aperture base physical address
- * @aperture_size: output returning doorbell aperture size in bytes
- * @start_offset: output returning # of doorbell bytes reserved for radeon.
- *
- * Radeon and the KFD share the doorbell aperture. Radeon sets it up,
- * takes doorbells required for its own rings and reports the setup to KFD.
- * Radeon reserved doorbells are at the start of the doorbell aperture.
- */
-void radeon_doorbell_get_kfd_info(struct radeon_device *rdev,
-				  phys_addr_t *aperture_base,
-				  size_t *aperture_size,
-				  size_t *start_offset)
-{
-	/* The first num_doorbells are used by radeon.
-	 * KFD takes whatever's left in the aperture. */
-	if (rdev->doorbell.size > rdev->doorbell.num_doorbells * sizeof(u32)) {
-		*aperture_base = rdev->doorbell.base;
-		*aperture_size = rdev->doorbell.size;
-		*start_offset = rdev->doorbell.num_doorbells * sizeof(u32);
-	} else {
-		*aperture_base = 0;
-		*aperture_size = 0;
-		*start_offset = 0;
-	}
-}
-
 /*
  * radeon_wb_*()
  * Writeback is the the method by which the the GPU updates special pages
@@ -648,7 +638,7 @@ void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc)
  * Used at driver startup.
  * Returns true if virtual or false if not.
  */
-static bool radeon_device_is_virtual(void)
+bool radeon_device_is_virtual(void)
 {
 #if defined(__amd64__) || defined(__i386__)
 	return (cpu_ecxfeature & CPUIDECX_HV);
@@ -1180,14 +1170,14 @@ static void radeon_check_arguments(struct radeon_device *rdev)
 	}
 
 	if (radeon_vm_size < 1) {
-		dev_warn(rdev->dev, "VM size (%d) to small, min is 1GB\n",
+		dev_warn(rdev->dev, "VM size (%d) too small, min is 1GB\n",
 			 radeon_vm_size);
 		radeon_vm_size = 4;
 	}
 
-       /*
-        * Max GPUVM size for Cayman, SI and CI are 40 bits.
-        */
+	/*
+	 * Max GPUVM size for Cayman, SI and CI are 40 bits.
+	 */
 	if (radeon_vm_size > 1024) {
 		dev_warn(rdev->dev, "VM size (%d) too large, max is 1TB\n",
 			 radeon_vm_size);
@@ -1236,32 +1226,24 @@ static void radeon_check_arguments(struct radeon_device *rdev)
 static void radeon_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_state state)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
-	struct radeon_device *rdev = dev->dev_private;
 
 	if (radeon_is_px(dev) && state == VGA_SWITCHEROO_OFF)
 		return;
 
 	if (state == VGA_SWITCHEROO_ON) {
-		unsigned d3_delay = dev->pdev->d3_delay;
-
-		printk(KERN_INFO "radeon: switched on\n");
+		pr_info("radeon: switched on\n");
 		/* don't suspend or resume card normally */
 		dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 
-		if (d3_delay < 20 && (rdev->px_quirk_flags & RADEON_PX_QUIRK_LONG_WAKEUP))
-			dev->pdev->d3_delay = 20;
-
 		radeon_resume_kms(dev, true, true);
 
-		dev->pdev->d3_delay = d3_delay;
-
 		dev->switch_power_state = DRM_SWITCH_POWER_ON;
 		drm_kms_helper_poll_enable(dev);
 	} else {
-		printk(KERN_INFO "radeon: switched off\n");
+		pr_info("radeon: switched off\n");
 		drm_kms_helper_poll_disable(dev);
 		dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
-		radeon_suspend_kms(dev, true, true);
+		radeon_suspend_kms(dev, true, true, false);
 		dev->switch_power_state = DRM_SWITCH_POWER_OFF;
 	}
 }
@@ -1328,11 +1310,11 @@ int radeon_device_init(struct radeon_device *rdev,
 	for (i = 0; i < RADEON_NUM_RINGS; i++) {
 		rdev->ring[i].idx = i;
 	}
-	rdev->fence_context = fence_context_alloc(RADEON_NUM_RINGS);
+	rdev->fence_context = dma_fence_context_alloc(RADEON_NUM_RINGS);
 
-	printf("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X).\n",
-		radeon_family_name[rdev->family], pdev->vendor, pdev->device,
-		pdev->subsystem_vendor, pdev->subsystem_device);
+	printf("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X 0x%02X).\n",
+		 radeon_family_name[rdev->family], pdev->vendor, pdev->device,
+		 pdev->subsystem_vendor, pdev->subsystem_device, pdev->revision);
 
 	/* mutex initialization are all done here so we
 	 * can recall function without having locking issues */
@@ -1343,7 +1325,6 @@ int radeon_device_init(struct radeon_device *rdev,
 	rw_init(&rdev->pm.mutex, "pm");
 	rw_init(&rdev->gpu_clock_mutex, "gpuclk");
 	rw_init(&rdev->srbm_mutex, "srbm");
-	rw_init(&rdev->grbm_idx_mutex, "grbm");
 	rw_init(&rdev->pm.mclk_lock, "mclk");
 	rw_init(&rdev->exclusive_lock, "rdnexc");
 	init_waitqueue_head(&rdev->irq.vblank_queue);
@@ -1399,6 +1380,10 @@ int radeon_device_init(struct radeon_device *rdev,
 	if ((rdev->flags & RADEON_IS_PCI) &&
 	    (rdev->family <= CHIP_RS740))
 		rdev->need_dma32 = true;
+#ifdef CONFIG_PPC64
+	if (rdev->family == CHIP_CEDAR)
+		rdev->need_dma32 = true;
+#endif
 
 	dma_bits = rdev->need_dma32 ? 32 : 40;
 #ifdef notyet
@@ -1406,13 +1391,14 @@ int radeon_device_init(struct radeon_device *rdev,
 	if (r) {
 		rdev->need_dma32 = true;
 		dma_bits = 32;
-		printk(KERN_WARNING "radeon: No suitable DMA available.\n");
+		pr_warn("radeon: No suitable DMA available\n");
 	}
 	r = pci_set_consistent_dma_mask(rdev->pdev, DMA_BIT_MASK(dma_bits));
 	if (r) {
 		pci_set_consistent_dma_mask(rdev->pdev, DMA_BIT_MASK(32));
-		printk(KERN_WARNING "radeon: No coherent DMA available.\n");
+		pr_warn("radeon: No coherent DMA available\n");
 	}
+	rdev->need_swiotlb = drm_get_max_iomem() > ((u64)1 << dma_bits);
 #endif
 
 	/* Registers mapping */
@@ -1438,12 +1424,9 @@ int radeon_device_init(struct radeon_device *rdev,
 		rdev->rmmio_size = pci_resource_len(rdev->pdev, 2);
 	}
 	rdev->rmmio = ioremap(rdev->rmmio_base, rdev->rmmio_size);
-	if (rdev->rmmio == NULL) {
+	if (rdev->rmmio == NULL)
 		return -ENOMEM;
-	}
 #endif
-	DRM_INFO("register mmio base: 0x%08X\n", (uint32_t)rdev->rmmio_base);
-	DRM_INFO("register mmio size: %u\n", (unsigned)rdev->rmmio_size);
 
 	/* doorbell bar mapping */
 	if (rdev->family >= CHIP_BONAIRE)
@@ -1475,7 +1458,9 @@ int radeon_device_init(struct radeon_device *rdev,
 	if (rdev->flags & RADEON_IS_PX)
 		runtime = true;
 #ifdef notyet
-	vga_switcheroo_register_client(rdev->pdev, &radeon_switcheroo_ops, runtime);
+	if (!pci_is_thunderbolt_attached(rdev->pdev))
+		vga_switcheroo_register_client(rdev->pdev,
+					       &radeon_switcheroo_ops, runtime);
 	if (runtime)
 		vga_switcheroo_init_domain_pm_ops(rdev->dev, &rdev->vga_pm_domain);
 #endif
@@ -1546,6 +1531,9 @@ int radeon_device_init(struct radeon_device *rdev,
 	return 0;
 
 failed:
+	/* balance pm_runtime_get_sync() in radeon_driver_unload_kms() */
+	if (radeon_is_px(ddev))
+		pm_runtime_put_noidle(ddev->dev);
 #ifdef notyet
 	if (runtime)
 		vga_switcheroo_fini_domain_pm_ops(rdev->dev);
@@ -1553,8 +1541,6 @@ failed:
 	return r;
 }
 
-static void radeon_debugfs_remove_files(struct radeon_device *rdev);
-
 /**
  * radeon_device_fini - tear down the driver
  *
@@ -1570,7 +1556,8 @@ void radeon_device_fini(struct radeon_device *rdev)
 	/* evict vram memory */
 	radeon_bo_evict_vram(rdev);
 	radeon_fini(rdev);
-	vga_switcheroo_unregister_client(rdev->pdev);
+	if (!pci_is_thunderbolt_attached(rdev->pdev))
+		vga_switcheroo_unregister_client(rdev->pdev);
 	if (rdev->flags & RADEON_IS_PX)
 		vga_switcheroo_fini_domain_pm_ops(rdev->dev);
 	vga_client_register(rdev->pdev, NULL, NULL, NULL);
@@ -1591,7 +1578,6 @@ void radeon_device_fini(struct radeon_device *rdev)
 #endif
 	if (rdev->family >= CHIP_BONAIRE)
 		radeon_doorbell_fini(rdev);
-	radeon_debugfs_remove_files(rdev);
 }
 
 
@@ -1608,7 +1594,8 @@ void radeon_device_fini(struct radeon_device *rdev)
  * Returns 0 for success or an error on failure.
  * Called at driver suspend.
  */
-int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
+int radeon_suspend_kms(struct drm_device *dev, bool suspend,
+		       bool fbcon, bool freeze)
 {
 	struct radeon_device *rdev;
 	struct drm_crtc *crtc;
@@ -1640,7 +1627,7 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 	/* unpin the front buffers and cursors */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-		struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->primary->fb);
+		struct drm_framebuffer *fb = crtc->primary->fb;
 		struct radeon_bo *robj;
 
 		if (radeon_crtc->cursor_bo) {
@@ -1652,10 +1639,10 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 			}
 		}
 
-		if (rfb == NULL || rfb->obj == NULL) {
+		if (fb == NULL || fb->obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_radeon_bo(rfb->obj);
+		robj = gem_to_radeon_bo(fb->obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!radeon_fbdev_robj_is_fb(rdev, robj)) {
 			r = radeon_bo_reserve(robj, false);
@@ -1681,13 +1668,19 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 
 	radeon_suspend(rdev);
 	radeon_hpd_fini(rdev);
-	/* evict remaining vram memory */
+	/* evict remaining vram memory
+	 * This second call to evict vram is to evict the gart page table
+	 * using the CPU.
+	 */
 	radeon_bo_evict_vram(rdev);
 
 	radeon_agp_suspend(rdev);
 
 	pci_save_state(dev->pdev);
-	if (suspend) {
+	if (freeze && rdev->family >= CHIP_CEDAR && !(rdev->flags & RADEON_IS_IGP)) {
+		rdev->asic->asic_reset(rdev, true);
+		pci_restore_state(dev->pdev);
+	} else if (suspend) {
 		/* Shut down the device */
 		pci_disable_device(dev->pdev);
 		pci_set_power_state(dev->pdev, PCI_D3hot);
@@ -1909,9 +1902,7 @@ int radeon_gpu_reset(struct radeon_device *rdev)
 	rdev->in_reset = true;
 	rdev->needs_reset = false;
 
-#ifdef notyet
 	downgrade_write(&rdev->exclusive_lock);
-#endif
 
 	drm_helper_resume_force_mode(rdev->ddev);
 
@@ -1956,7 +1947,7 @@ int radeon_debugfs_add_files(struct radeon_device *rdev,
 	if (i > RADEON_DEBUGFS_MAX_COMPONENTS) {
 		DRM_ERROR("Reached maximum number of debugfs components.\n");
 		DRM_ERROR("Report so we increase "
-		          "RADEON_DEBUGFS_MAX_COMPONENTS.\n");
+			  "RADEON_DEBUGFS_MAX_COMPONENTS.\n");
 		return -EINVAL;
 	}
 	rdev->debugfs[rdev->debugfs_count].files = files;
@@ -1964,38 +1955,8 @@ int radeon_debugfs_add_files(struct radeon_device *rdev,
 	rdev->debugfs_count = i;
 #if defined(CONFIG_DEBUG_FS)
 	drm_debugfs_create_files(files, nfiles,
-				 rdev->ddev->control->debugfs_root,
-				 rdev->ddev->control);
-	drm_debugfs_create_files(files, nfiles,
 				 rdev->ddev->primary->debugfs_root,
 				 rdev->ddev->primary);
 #endif
 	return 0;
 }
-
-static void radeon_debugfs_remove_files(struct radeon_device *rdev)
-{
-#if defined(CONFIG_DEBUG_FS)
-	unsigned i;
-
-	for (i = 0; i < rdev->debugfs_count; i++) {
-		drm_debugfs_remove_files(rdev->debugfs[i].files,
-					 rdev->debugfs[i].num_files,
-					 rdev->ddev->control);
-		drm_debugfs_remove_files(rdev->debugfs[i].files,
-					 rdev->debugfs[i].num_files,
-					 rdev->ddev->primary);
-	}
-#endif
-}
-
-#if defined(CONFIG_DEBUG_FS)
-int radeon_debugfs_init(struct drm_minor *minor)
-{
-	return 0;
-}
-
-void radeon_debugfs_cleanup(struct drm_minor *minor)
-{
-}
-#endif
diff --git a/sys/dev/pci/drm/radeon/radeon_display.c b/sys/dev/pci/drm/radeon/radeon_display.c
index 821c42c85b1..62fe4fae18f 100644
--- a/sys/dev/pci/drm/radeon/radeon_display.c
+++ b/sys/dev/pci/drm/radeon/radeon_display.c
@@ -23,22 +23,28 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 #include "atom.h"
+#include <asm/div64.h>
 
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/drm_plane_helper.h>
-#include <dev/pci/drm/drm_edid.h>
+#include <linux/pm_runtime.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_edid.h>
 
+#include <linux/gcd.h>
 
 static void avivo_crtc_load_lut(struct drm_crtc *crtc)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	u16 *r, *g, *b;
 	int i;
 
 	DRM_DEBUG_KMS("%d\n", radeon_crtc->crtc_id);
@@ -57,11 +63,14 @@ static void avivo_crtc_load_lut(struct drm_crtc *crtc)
 	WREG32(AVIVO_DC_LUT_WRITE_EN_MASK, 0x0000003f);
 
 	WREG8(AVIVO_DC_LUT_RW_INDEX, 0);
+	r = crtc->gamma_store;
+	g = r + crtc->gamma_size;
+	b = g + crtc->gamma_size;
 	for (i = 0; i < 256; i++) {
 		WREG32(AVIVO_DC_LUT_30_COLOR,
-			     (radeon_crtc->lut_r[i] << 20) |
-			     (radeon_crtc->lut_g[i] << 10) |
-			     (radeon_crtc->lut_b[i] << 0));
+		       ((*r++ & 0xffc0) << 14) |
+		       ((*g++ & 0xffc0) << 4) |
+		       (*b++ >> 6));
 	}
 
 	/* Only change bit 0 of LUT_SEL, other bits are set elsewhere */
@@ -73,6 +82,7 @@ static void dce4_crtc_load_lut(struct drm_crtc *crtc)
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	u16 *r, *g, *b;
 	int i;
 
 	DRM_DEBUG_KMS("%d\n", radeon_crtc->crtc_id);
@@ -90,11 +100,14 @@ static void dce4_crtc_load_lut(struct drm_crtc *crtc)
 	WREG32(EVERGREEN_DC_LUT_WRITE_EN_MASK + radeon_crtc->crtc_offset, 0x00000007);
 
 	WREG32(EVERGREEN_DC_LUT_RW_INDEX + radeon_crtc->crtc_offset, 0);
+	r = crtc->gamma_store;
+	g = r + crtc->gamma_size;
+	b = g + crtc->gamma_size;
 	for (i = 0; i < 256; i++) {
 		WREG32(EVERGREEN_DC_LUT_30_COLOR + radeon_crtc->crtc_offset,
-		       (radeon_crtc->lut_r[i] << 20) |
-		       (radeon_crtc->lut_g[i] << 10) |
-		       (radeon_crtc->lut_b[i] << 0));
+		       ((*r++ & 0xffc0) << 14) |
+		       ((*g++ & 0xffc0) << 4) |
+		       (*b++ >> 6));
 	}
 }
 
@@ -103,6 +116,7 @@ static void dce5_crtc_load_lut(struct drm_crtc *crtc)
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	u16 *r, *g, *b;
 	int i;
 
 	DRM_DEBUG_KMS("%d\n", radeon_crtc->crtc_id);
@@ -132,11 +146,14 @@ static void dce5_crtc_load_lut(struct drm_crtc *crtc)
 	WREG32(EVERGREEN_DC_LUT_WRITE_EN_MASK + radeon_crtc->crtc_offset, 0x00000007);
 
 	WREG32(EVERGREEN_DC_LUT_RW_INDEX + radeon_crtc->crtc_offset, 0);
+	r = crtc->gamma_store;
+	g = r + crtc->gamma_size;
+	b = g + crtc->gamma_size;
 	for (i = 0; i < 256; i++) {
 		WREG32(EVERGREEN_DC_LUT_30_COLOR + radeon_crtc->crtc_offset,
-		       (radeon_crtc->lut_r[i] << 20) |
-		       (radeon_crtc->lut_g[i] << 10) |
-		       (radeon_crtc->lut_b[i] << 0));
+		       ((*r++ & 0xffc0) << 14) |
+		       ((*g++ & 0xffc0) << 4) |
+		       (*b++ >> 6));
 	}
 
 	WREG32(NI_DEGAMMA_CONTROL + radeon_crtc->crtc_offset,
@@ -169,6 +186,7 @@ static void legacy_crtc_load_lut(struct drm_crtc *crtc)
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	u16 *r, *g, *b;
 	int i;
 	uint32_t dac2_cntl;
 
@@ -180,11 +198,14 @@ static void legacy_crtc_load_lut(struct drm_crtc *crtc)
 	WREG32(RADEON_DAC_CNTL2, dac2_cntl);
 
 	WREG8(RADEON_PALETTE_INDEX, 0);
+	r = crtc->gamma_store;
+	g = r + crtc->gamma_size;
+	b = g + crtc->gamma_size;
 	for (i = 0; i < 256; i++) {
 		WREG32(RADEON_PALETTE_30_DATA,
-			     (radeon_crtc->lut_r[i] << 20) |
-			     (radeon_crtc->lut_g[i] << 10) |
-			     (radeon_crtc->lut_b[i] << 0));
+		       ((*r++ & 0xffc0) << 14) |
+		       ((*g++ & 0xffc0) << 4) |
+		       (*b++ >> 6));
 	}
 }
 
@@ -206,41 +227,13 @@ void radeon_crtc_load_lut(struct drm_crtc *crtc)
 		legacy_crtc_load_lut(crtc);
 }
 
-/** Sets the color ramps on behalf of fbcon */
-void radeon_crtc_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green,
-			      u16 blue, int regno)
+static int radeon_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
+				 u16 *blue, uint32_t size,
+				 struct drm_modeset_acquire_ctx *ctx)
 {
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-
-	radeon_crtc->lut_r[regno] = red >> 6;
-	radeon_crtc->lut_g[regno] = green >> 6;
-	radeon_crtc->lut_b[regno] = blue >> 6;
-}
-
-/** Gets the color ramps on behalf of fbcon */
-void radeon_crtc_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green,
-			      u16 *blue, int regno)
-{
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-
-	*red = radeon_crtc->lut_r[regno] << 6;
-	*green = radeon_crtc->lut_g[regno] << 6;
-	*blue = radeon_crtc->lut_b[regno] << 6;
-}
-
-static void radeon_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
-				  u16 *blue, uint32_t start, uint32_t size)
-{
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	int end = (start + size > 256) ? 256 : start + size, i;
-
-	/* userspace palettes are always correct as is */
-	for (i = start; i < end; i++) {
-		radeon_crtc->lut_r[i] = red[i] >> 6;
-		radeon_crtc->lut_g[i] = green[i] >> 6;
-		radeon_crtc->lut_b[i] = blue[i] >> 6;
-	}
 	radeon_crtc_load_lut(crtc);
+
+	return 0;
 }
 
 static void radeon_crtc_destroy(struct drm_crtc *crtc)
@@ -276,7 +269,7 @@ static void radeon_unpin_work_func(struct work_struct *__work)
 	} else
 		DRM_ERROR("failed to reserve buffer after flip\n");
 
-	drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base);
+	drm_gem_object_put_unlocked(&work->old_rbo->gem_base);
 	kfree(work);
 }
 
@@ -316,16 +309,30 @@ void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id)
 	update_pending = radeon_page_flip_pending(rdev, crtc_id);
 
 	/* Has the pageflip already completed in crtc, or is it certain
-	 * to complete in this vblank?
+	 * to complete in this vblank? GET_DISTANCE_TO_VBLANKSTART provides
+	 * distance to start of "fudged earlier" vblank in vpos, distance to
+	 * start of real vblank in hpos. vpos >= 0 && hpos < 0 means we are in
+	 * the last few scanlines before start of real vblank, where the vblank
+	 * irq can fire, so we have sampled update_pending a bit too early and
+	 * know the flip will complete at leading edge of the upcoming real
+	 * vblank. On pre-AVIVO hardware, flips also complete inside the real
+	 * vblank, not only at leading edge, so if update_pending for hpos >= 0
+	 *  == inside real vblank, the flip will complete almost immediately.
+	 * Note that this method of completion handling is still not 100% race
+	 * free, as we could execute before the radeon_flip_work_func managed
+	 * to run and set the RADEON_FLIP_SUBMITTED status, thereby we no-op,
+	 * but the flip still gets programmed into hw and completed during
+	 * vblank, leading to a delayed emission of the flip completion event.
+	 * This applies at least to pre-AVIVO hardware, where flips are always
+	 * completing inside vblank, not only at leading edge of vblank.
 	 */
 	if (update_pending &&
-	    (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev,
-							       crtc_id,
-							       USE_REAL_VBLANKSTART,
-							       &vpos, &hpos, NULL, NULL,
-							       &rdev->mode_info.crtcs[crtc_id]->base.hwmode)) &&
-	    ((vpos >= (99 * rdev->mode_info.crtcs[crtc_id]->base.hwmode.crtc_vdisplay)/100) ||
-	     (vpos < 0 && !ASIC_IS_AVIVO(rdev)))) {
+	    (DRM_SCANOUTPOS_VALID &
+	     radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id,
+					GET_DISTANCE_TO_VBLANKSTART,
+					&vpos, &hpos, NULL, NULL,
+					&rdev->mode_info.crtcs[crtc_id]->base.hwmode)) &&
+	    ((vpos >= 0 && hpos < 0) || (hpos >= 0 && !ASIC_IS_AVIVO(rdev)))) {
 		/* crtc didn't flip in this target vblank interval,
 		 * but flip is pending in crtc. Based on the current
 		 * scanout position we know that the current frame is
@@ -374,11 +381,11 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
 
 	/* wakeup userspace */
 	if (work->event)
-		drm_send_vblank_event(rdev->ddev, crtc_id, work->event);
+		drm_crtc_send_vblank_event(&radeon_crtc->base, work->event);
 
 	spin_unlock_irqrestore(&rdev->ddev->event_lock, flags);
 
-	drm_vblank_put(rdev->ddev, radeon_crtc->crtc_id);
+	drm_crtc_vblank_put(&radeon_crtc->base);
 	radeon_irq_kms_pflip_irq_put(rdev, work->crtc_id);
 	queue_work(radeon_crtc->flip_queue, &work->unpin_work);
 }
@@ -395,16 +402,15 @@ static void radeon_flip_work_func(struct work_struct *__work)
 	struct radeon_flip_work *work =
 		container_of(__work, struct radeon_flip_work, flip_work);
 	struct radeon_device *rdev = work->rdev;
+	struct drm_device *dev = rdev->ddev;
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[work->crtc_id];
 
 	struct drm_crtc *crtc = &radeon_crtc->base;
 	unsigned long flags;
 	int r;
-	int vpos, hpos, stat, min_udelay = 0;
-	unsigned repcnt = 4;
-	struct drm_vblank_crtc *vblank = &crtc->dev->vblank[work->crtc_id];
+	int vpos, hpos;
 
-        down_read(&rdev->exclusive_lock);
+	down_read(&rdev->exclusive_lock);
 	if (work->fence) {
 		struct radeon_fence *fence;
 
@@ -419,7 +425,7 @@ static void radeon_flip_work_func(struct work_struct *__work)
 				down_read(&rdev->exclusive_lock);
 			}
 		} else
-			r = fence_wait(work->fence, false);
+			r = dma_fence_wait(work->fence, false);
 
 		if (r)
 			DRM_ERROR("failed to wait on page flip fence (%d)!\n", r);
@@ -429,81 +435,50 @@ static void radeon_flip_work_func(struct work_struct *__work)
 		 * confused about which BO the CRTC is scanning out
 		 */
 
-		fence_put(work->fence);
+		dma_fence_put(work->fence);
 		work->fence = NULL;
 	}
 
+	/* Wait until we're out of the vertical blank period before the one
+	 * targeted by the flip. Always wait on pre DCE4 to avoid races with
+	 * flip completion handling from vblank irq, as these old asics don't
+	 * have reliable pageflip completion interrupts.
+	 */
+	while (radeon_crtc->enabled &&
+		(radeon_get_crtc_scanoutpos(dev, work->crtc_id, 0,
+					    &vpos, &hpos, NULL, NULL,
+					    &crtc->hwmode)
+		& (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) ==
+		(DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) &&
+		(!ASIC_IS_AVIVO(rdev) ||
+		((int) (work->target_vblank -
+		dev->driver->get_vblank_counter(dev, work->crtc_id)) > 0)))
+		usleep_range(1000, 2000);
+
 	/* We borrow the event spin lock for protecting flip_status */
 	spin_lock_irqsave(&crtc->dev->event_lock, flags);
 
 	/* set the proper interrupt */
 	radeon_irq_kms_pflip_irq_get(rdev, radeon_crtc->crtc_id);
 
-	/* If this happens to execute within the "virtually extended" vblank
-	 * interval before the start of the real vblank interval then it needs
-	 * to delay programming the mmio flip until the real vblank is entered.
-	 * This prevents completing a flip too early due to the way we fudge
-	 * our vblank counter and vblank timestamps in order to work around the
-	 * problem that the hw fires vblank interrupts before actual start of
-	 * vblank (when line buffer refilling is done for a frame). It
-	 * complements the fudging logic in radeon_get_crtc_scanoutpos() for
-	 * timestamping and radeon_get_vblank_counter_kms() for vblank counts.
-	 *
-	 * In practice this won't execute very often unless on very fast
-	 * machines because the time window for this to happen is very small.
-	 */
-	while (radeon_crtc->enabled && --repcnt) {
-		/* GET_DISTANCE_TO_VBLANKSTART returns distance to real vblank
-		 * start in hpos, and to the "fudged earlier" vblank start in
-		 * vpos.
-		 */
-		stat = radeon_get_crtc_scanoutpos(rdev->ddev, work->crtc_id,
-						  GET_DISTANCE_TO_VBLANKSTART,
-						  &vpos, &hpos, NULL, NULL,
-						  &crtc->hwmode);
-
-		if ((stat & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE)) !=
-		    (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE) ||
-		    !(vpos >= 0 && hpos <= 0))
-			break;
-
-		/* Sleep at least until estimated real start of hw vblank */
-		min_udelay = (-hpos + 1) * max(vblank->linedur_ns / 1000, 5);
-		if (min_udelay > vblank->framedur_ns / 2000) {
-			/* Don't wait ridiculously long - something is wrong */
-			repcnt = 0;
-			break;
-		}
-		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
-		usleep_range(min_udelay, 2 * min_udelay);
-		spin_lock_irqsave(&crtc->dev->event_lock, flags);
-	};
-
-	if (!repcnt)
-		DRM_DEBUG_DRIVER("Delay problem on crtc %d: min_udelay %d, "
-				 "framedur %d, linedur %d, stat %d, vpos %d, "
-				 "hpos %d\n", work->crtc_id, min_udelay,
-				 vblank->framedur_ns / 1000,
-				 vblank->linedur_ns / 1000, stat, vpos, hpos);
-
 	/* do the flip (mmio) */
-	radeon_page_flip(rdev, radeon_crtc->crtc_id, work->base);
+	radeon_page_flip(rdev, radeon_crtc->crtc_id, work->base, work->async);
 
 	radeon_crtc->flip_status = RADEON_FLIP_SUBMITTED;
 	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 	up_read(&rdev->exclusive_lock);
 }
 
-static int radeon_crtc_page_flip(struct drm_crtc *crtc,
-				 struct drm_framebuffer *fb,
-				 struct drm_pending_vblank_event *event,
-				 uint32_t page_flip_flags)
+static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
+					struct drm_framebuffer *fb,
+					struct drm_pending_vblank_event *event,
+					uint32_t page_flip_flags,
+					uint32_t target,
+					struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *old_radeon_fb;
-	struct radeon_framebuffer *new_radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_flip_work *work;
 	struct radeon_bo *new_rbo;
@@ -522,17 +497,16 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 	work->rdev = rdev;
 	work->crtc_id = radeon_crtc->crtc_id;
 	work->event = event;
+	work->async = (page_flip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 
 	/* schedule unpin of the old buffer */
-	old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-	obj = old_radeon_fb->obj;
+	obj = crtc->primary->fb->obj[0];
 
 	/* take a reference to the old object */
-	drm_gem_object_reference(obj);
+	drm_gem_object_get(obj);
 	work->old_rbo = gem_to_radeon_bo(obj);
 
-	new_radeon_fb = to_radeon_framebuffer(fb);
-	obj = new_radeon_fb->obj;
+	obj = fb->obj[0];
 	new_rbo = gem_to_radeon_bo(obj);
 
 	/* pin the new buffer */
@@ -553,26 +527,26 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		DRM_ERROR("failed to pin new rbo buffer before flip\n");
 		goto cleanup;
 	}
-	work->fence = fence_get(reservation_object_get_excl(new_rbo->tbo.resv));
+	work->fence = dma_fence_get(reservation_object_get_excl(new_rbo->tbo.resv));
 	radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(new_rbo);
 
 	if (!ASIC_IS_AVIVO(rdev)) {
 		/* crtc offset is from display base addr not FB location */
 		base -= radeon_crtc->legacy_display_base_addr;
-		pitch_pixels = fb->pitches[0] / (fb->bits_per_pixel / 8);
+		pitch_pixels = fb->pitches[0] / fb->format->cpp[0];
 
 		if (tiling_flags & RADEON_TILING_MACRO) {
 			if (ASIC_IS_R300(rdev)) {
 				base &= ~0x7ff;
 			} else {
-				int byteshift = fb->bits_per_pixel >> 4;
+				int byteshift = fb->format->cpp[0] * 8 >> 4;
 				int tile_addr = (((crtc->y >> 3) * pitch_pixels +  crtc->x) >> (8 - byteshift)) << 11;
 				base += tile_addr + ((crtc->x << byteshift) % 256) + ((crtc->y % 8) << 8);
 			}
 		} else {
 			int offset = crtc->y * pitch_pixels + crtc->x;
-			switch (fb->bits_per_pixel) {
+			switch (fb->format->cpp[0] * 8) {
 			case 8:
 			default:
 				offset *= 1;
@@ -593,12 +567,8 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		base &= ~7;
 	}
 	work->base = base;
-
-	r = drm_vblank_get(crtc->dev, radeon_crtc->crtc_id);
-	if (r) {
-		DRM_ERROR("failed to get vblank before flip\n");
-		goto pflip_cleanup;
-	}
+	work->target_vblank = target - (uint32_t)drm_crtc_vblank_count(crtc) +
+		dev->driver->get_vblank_counter(dev, work->crtc_id);
 
 	/* We borrow the event spin lock for protecting flip_work */
 	spin_lock_irqsave(&crtc->dev->event_lock, flags);
@@ -607,7 +577,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
 		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 		r = -EBUSY;
-		goto vblank_cleanup;
+		goto pflip_cleanup;
 	}
 	radeon_crtc->flip_status = RADEON_FLIP_PENDING;
 	radeon_crtc->flip_work = work;
@@ -620,9 +590,6 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 	queue_work(radeon_crtc->flip_queue, &work->flip_work);
 	return 0;
 
-vblank_cleanup:
-	drm_vblank_put(crtc->dev, radeon_crtc->crtc_id);
-
 pflip_cleanup:
 	if (unlikely(radeon_bo_reserve(new_rbo, false) != 0)) {
 		DRM_ERROR("failed to reserve new rbo in error path\n");
@@ -634,14 +601,15 @@ pflip_cleanup:
 	radeon_bo_unreserve(new_rbo);
 
 cleanup:
-	drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base);
-	fence_put(work->fence);
+	drm_gem_object_put_unlocked(&work->old_rbo->gem_base);
+	dma_fence_put(work->fence);
 	kfree(work);
 	return r;
 }
 
 static int
-radeon_crtc_set_config(struct drm_mode_set *set)
+radeon_crtc_set_config(struct drm_mode_set *set,
+		       struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_device *dev;
 	struct radeon_device *rdev;
@@ -658,7 +626,7 @@ radeon_crtc_set_config(struct drm_mode_set *set)
 	if (ret < 0)
 		return ret;
 
-	ret = drm_crtc_helper_set_config(set);
+	ret = drm_crtc_helper_set_config(set, ctx);
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
 		if (crtc->enabled)
@@ -684,13 +652,14 @@ radeon_crtc_set_config(struct drm_mode_set *set)
 	pm_runtime_put_autosuspend(dev->dev);
 	return ret;
 }
+
 static const struct drm_crtc_funcs radeon_crtc_funcs = {
 	.cursor_set2 = radeon_crtc_cursor_set2,
 	.cursor_move = radeon_crtc_cursor_move,
 	.gamma_set = radeon_crtc_gamma_set,
 	.set_config = radeon_crtc_set_config,
 	.destroy = radeon_crtc_destroy,
-	.page_flip = radeon_crtc_page_flip,
+	.page_flip_target = radeon_crtc_page_flip_target,
 };
 
 static void radeon_crtc_init(struct drm_device *dev, int index)
@@ -707,7 +676,7 @@ static void radeon_crtc_init(struct drm_device *dev, int index)
 
 	drm_mode_crtc_set_gamma_size(&radeon_crtc->base, 256);
 	radeon_crtc->crtc_id = index;
-	radeon_crtc->flip_queue = create_singlethread_workqueue("radeon-crtc");
+	radeon_crtc->flip_queue = alloc_workqueue("radeon-crtc", WQ_HIGHPRI, 0);
 	rdev->mode_info.crtcs[index] = radeon_crtc;
 
 	if (rdev->family >= CHIP_BONAIRE) {
@@ -918,7 +887,7 @@ static void avivo_reduce_ratio(unsigned *nom, unsigned *den,
 	*den /= tmp;
 
 	/* make sure nominator is large enough */
-        if (*nom < nom_min) {
+	if (*nom < nom_min) {
 		tmp = DIV_ROUND_UP(nom_min, *nom);
 		*nom *= tmp;
 		*den *= tmp;
@@ -958,7 +927,7 @@ static void avivo_get_fb_ref_div(unsigned nom, unsigned den, unsigned post_div,
 	*fb_div = DIV_ROUND_CLOSEST(nom * *ref_div * post_div, den);
 
 	/* limit fb divider to its maximum */
-        if (*fb_div > fb_div_max) {
+	if (*fb_div > fb_div_max) {
 		*ref_div = DIV_ROUND_CLOSEST(*ref_div * fb_div_max, *fb_div);
 		*fb_div = fb_div_max;
 	}
@@ -1315,43 +1284,23 @@ void radeon_compute_pll_legacy(struct radeon_pll *pll,
 
 }
 
-static void radeon_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	if (radeon_fb->obj) {
-		drm_gem_object_unreference_unlocked(radeon_fb->obj);
-	}
-	drm_framebuffer_cleanup(fb);
-	kfree(radeon_fb);
-}
-
-static int radeon_user_framebuffer_create_handle(struct drm_framebuffer *fb,
-						  struct drm_file *file_priv,
-						  unsigned int *handle)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, radeon_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs radeon_fb_funcs = {
-	.destroy = radeon_user_framebuffer_destroy,
-	.create_handle = radeon_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 int
 radeon_framebuffer_init(struct drm_device *dev,
-			struct radeon_framebuffer *rfb,
-			struct drm_mode_fb_cmd2 *mode_cmd,
+			struct drm_framebuffer *fb,
+			const struct drm_mode_fb_cmd2 *mode_cmd,
 			struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->obj = obj;
-	drm_helper_mode_fill_fb_struct(&rfb->base, mode_cmd);
-	ret = drm_framebuffer_init(dev, &rfb->base, &radeon_fb_funcs);
+	fb->obj[0] = obj;
+	drm_helper_mode_fill_fb_struct(dev, fb, mode_cmd);
+	ret = drm_framebuffer_init(dev, fb, &radeon_fb_funcs);
 	if (ret) {
-		rfb->obj = NULL;
+		fb->obj[0] = NULL;
 		return ret;
 	}
 	return 0;
@@ -1360,13 +1309,13 @@ radeon_framebuffer_init(struct drm_device *dev,
 static struct drm_framebuffer *
 radeon_user_framebuffer_create(struct drm_device *dev,
 			       struct drm_file *file_priv,
-			       struct drm_mode_fb_cmd2 *mode_cmd)
+			       const struct drm_mode_fb_cmd2 *mode_cmd)
 {
 	struct drm_gem_object *obj;
-	struct radeon_framebuffer *radeon_fb;
+	struct drm_framebuffer *fb;
 	int ret;
 
-	obj = drm_gem_object_lookup(dev, file_priv, mode_cmd->handles[0]);
+	obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[0]);
 	if (obj ==  NULL) {
 		dev_err(&dev->pdev->dev, "No GEM object associated to handle 0x%08X, "
 			"can't create framebuffer\n", mode_cmd->handles[0]);
@@ -1381,39 +1330,33 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 	}
 #endif
 
-	radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL);
-	if (radeon_fb == NULL) {
-		drm_gem_object_unreference_unlocked(obj);
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (fb == NULL) {
+		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ret = radeon_framebuffer_init(dev, radeon_fb, mode_cmd, obj);
+	ret = radeon_framebuffer_init(dev, fb, mode_cmd, obj);
 	if (ret) {
-		kfree(radeon_fb);
-		drm_gem_object_unreference_unlocked(obj);
+		kfree(fb);
+		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(ret);
 	}
 
-	return &radeon_fb->base;
-}
-
-static void radeon_output_poll_changed(struct drm_device *dev)
-{
-	struct radeon_device *rdev = dev->dev_private;
-	radeon_fb_output_poll_changed(rdev);
+	return fb;
 }
 
 static const struct drm_mode_config_funcs radeon_mode_funcs = {
 	.fb_create = radeon_user_framebuffer_create,
-	.output_poll_changed = radeon_output_poll_changed
+	.output_poll_changed = drm_fb_helper_output_poll_changed,
 };
 
-static struct drm_prop_enum_list radeon_tmds_pll_enum_list[] =
+static const struct drm_prop_enum_list radeon_tmds_pll_enum_list[] =
 {	{ 0, "driver" },
 	{ 1, "bios" },
 };
 
-static struct drm_prop_enum_list radeon_tv_std_enum_list[] =
+static const struct drm_prop_enum_list radeon_tv_std_enum_list[] =
 {	{ TV_STD_NTSC, "ntsc" },
 	{ TV_STD_PAL, "pal" },
 	{ TV_STD_PAL_M, "pal-m" },
@@ -1424,25 +1367,25 @@ static struct drm_prop_enum_list radeon_tv_std_enum_list[] =
 	{ TV_STD_SECAM, "secam" },
 };
 
-static struct drm_prop_enum_list radeon_underscan_enum_list[] =
+static const struct drm_prop_enum_list radeon_underscan_enum_list[] =
 {	{ UNDERSCAN_OFF, "off" },
 	{ UNDERSCAN_ON, "on" },
 	{ UNDERSCAN_AUTO, "auto" },
 };
 
-static struct drm_prop_enum_list radeon_audio_enum_list[] =
+static const struct drm_prop_enum_list radeon_audio_enum_list[] =
 {	{ RADEON_AUDIO_DISABLE, "off" },
 	{ RADEON_AUDIO_ENABLE, "on" },
 	{ RADEON_AUDIO_AUTO, "auto" },
 };
 
 /* XXX support different dither options? spatial, temporal, both, etc. */
-static struct drm_prop_enum_list radeon_dither_enum_list[] =
+static const struct drm_prop_enum_list radeon_dither_enum_list[] =
 {	{ RADEON_FMT_DITHER_DISABLE, "off" },
 	{ RADEON_FMT_DITHER_ENABLE, "on" },
 };
 
-static struct drm_prop_enum_list radeon_output_csc_enum_list[] =
+static const struct drm_prop_enum_list radeon_output_csc_enum_list[] =
 {	{ RADEON_OUTPUT_CSC_BYPASS, "bypass" },
 	{ RADEON_OUTPUT_CSC_TVRGB, "tvrgb" },
 	{ RADEON_OUTPUT_CSC_YCBCR601, "ycbcr601" },
@@ -1637,6 +1580,9 @@ int radeon_modeset_init(struct radeon_device *rdev)
 
 	rdev->ddev->mode_config.funcs = &radeon_mode_funcs;
 
+	if (radeon_use_pflipirq == 2 && rdev->family >= CHIP_R600)
+		rdev->ddev->mode_config.async_page_flip = true;
+
 	if (ASIC_IS_DCE5(rdev)) {
 		rdev->ddev->mode_config.max_width = 16384;
 		rdev->ddev->mode_config.max_height = 16384;
@@ -1701,16 +1647,18 @@ int radeon_modeset_init(struct radeon_device *rdev)
 
 void radeon_modeset_fini(struct radeon_device *rdev)
 {
-	radeon_fbdev_fini(rdev);
-	kfree(rdev->mode_info.bios_hardcoded_edid);
-
 	if (rdev->mode_info.mode_config_initialized) {
-		radeon_afmt_fini(rdev);
 		drm_kms_helper_poll_fini(rdev->ddev);
 		radeon_hpd_fini(rdev);
+		drm_crtc_force_disable_all(rdev->ddev);
+		radeon_fbdev_fini(rdev);
+		radeon_afmt_fini(rdev);
 		drm_mode_config_cleanup(rdev->ddev);
 		rdev->mode_info.mode_config_initialized = false;
 	}
+
+	kfree(rdev->mode_info.bios_hardcoded_edid);
+
 	/* free i2c buses */
 	radeon_i2c_fini(rdev);
 }
diff --git a/sys/dev/pci/drm/radeon/radeon_dp_auxch.c b/sys/dev/pci/drm/radeon/radeon_dp_auxch.c
index 2bb7f05fe09..12eac4e7554 100644
--- a/sys/dev/pci/drm/radeon/radeon_dp_auxch.c
+++ b/sys/dev/pci/drm/radeon/radeon_dp_auxch.c
@@ -21,8 +21,8 @@
  *
  * Authors: Dave Airlie
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "nid.h"
 
@@ -164,12 +164,12 @@ radeon_dp_aux_transfer_native(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg
 	}
 
 	if (tmp & AUX_SW_RX_TIMEOUT) {
-		DRM_DEBUG_KMS("dp_aux_ch timed out\n");
 		ret = -ETIMEDOUT;
 		goto done;
 	}
 	if (tmp & AUX_RX_ERROR_FLAGS) {
-		DRM_DEBUG_KMS("dp_aux_ch flags not zero: %08x\n", tmp);
+		DRM_DEBUG_KMS_RATELIMITED("dp_aux_ch flags not zero: %08x\n",
+					  tmp);
 		ret = -EIO;
 		goto done;
 	}
diff --git a/sys/dev/pci/drm/radeon/radeon_dp_mst.c b/sys/dev/pci/drm/radeon/radeon_dp_mst.c
index 2fcdbf480d2..84b3ad2172a 100644
--- a/sys/dev/pci/drm/radeon/radeon_dp_mst.c
+++ b/sys/dev/pci/drm/radeon/radeon_dp_mst.c
@@ -1,8 +1,8 @@
 // SPDX-License-Identifier: MIT
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_dp_mst_helper.h>
-#include <dev/pci/drm/drm_fb_helper.h>
+#include <drm/drmP.h>
+#include <drm/drm_dp_mst_helper.h>
+#include <drm/drm_fb_helper.h>
 
 #include "radeon.h"
 #include "atom.h"
@@ -90,8 +90,16 @@ static int radeon_dp_mst_set_stream_attrib(struct radeon_encoder *primary,
 	WREG32(NI_DP_MSE_SAT_UPDATE + primary->offset, 1);
 
 	do {
+		unsigned value1, value2;
+		udelay(10);
 		temp = RREG32(NI_DP_MSE_SAT_UPDATE + primary->offset);
-	} while ((temp & 0x1) && retries++ < 10000);
+
+		value1 = temp & NI_DP_MSE_SAT_UPDATE_MASK;
+		value2 = temp & NI_DP_MSE_16_MTP_KEEPOUT;
+
+		if (!value1 && !value2)
+			break;
+	} while (retries++ < 50);
 
 	if (retries == 10000)
 		DRM_ERROR("timed out waitin for SAT update %d\n", primary->offset);
@@ -151,7 +159,7 @@ static int radeon_dp_mst_update_stream_attribs(struct radeon_connector *mst_conn
 	return 0;
 }
 
-static int radeon_dp_mst_set_vcp_size(struct radeon_encoder *mst, uint32_t x, uint32_t y)
+static int radeon_dp_mst_set_vcp_size(struct radeon_encoder *mst, s64 avg_time_slots_per_mtp)
 {
 	struct drm_device *dev = mst->base.dev;
 	struct radeon_device *rdev = dev->dev_private;
@@ -159,6 +167,8 @@ static int radeon_dp_mst_set_vcp_size(struct radeon_encoder *mst, uint32_t x, ui
 	uint32_t val, temp;
 	uint32_t offset = radeon_atom_set_enc_offset(mst_enc->fe);
 	int retries = 0;
+	uint32_t x = drm_fixp2int(avg_time_slots_per_mtp);
+	uint32_t y = drm_fixp2int_ceil((avg_time_slots_per_mtp - x) << 26);
 
 	val = NI_DP_MSE_RATE_X(x) | NI_DP_MSE_RATE_Y(y);
 
@@ -166,6 +176,7 @@ static int radeon_dp_mst_set_vcp_size(struct radeon_encoder *mst, uint32_t x, ui
 
 	do {
 		temp = RREG32(NI_DP_MSE_RATE_UPDATE + offset);
+		udelay(10);
 	} while ((temp & 0x1) && (retries++ < 10000));
 
 	if (retries >= 10000)
@@ -184,12 +195,11 @@ static int radeon_dp_mst_get_ddc_modes(struct drm_connector *connector)
 	radeon_connector->edid = edid;
 	DRM_DEBUG_KMS("edid retrieved %p\n", edid);
 	if (radeon_connector->edid) {
-		drm_mode_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid);
+		drm_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid);
 		ret = drm_add_edid_modes(&radeon_connector->base, radeon_connector->edid);
-		drm_edid_to_eld(&radeon_connector->base, radeon_connector->edid);
 		return ret;
 	}
-	drm_mode_connector_update_edid_property(&radeon_connector->base, NULL);
+	drm_connector_update_edid_property(&radeon_connector->base, NULL);
 
 	return ret;
 }
@@ -213,7 +223,8 @@ radeon_dp_mst_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-struct drm_encoder *radeon_mst_best_encoder(struct drm_connector *connector)
+static struct
+drm_encoder *radeon_mst_best_encoder(struct drm_connector *connector)
 {
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 
@@ -247,14 +258,8 @@ radeon_dp_mst_connector_destroy(struct drm_connector *connector)
 	kfree(radeon_connector);
 }
 
-static int radeon_connector_dpms(struct drm_connector *connector, int mode)
-{
-	DRM_DEBUG_KMS("\n");
-	return 0;
-}
-
 static const struct drm_connector_funcs radeon_dp_mst_connector_funcs = {
-	.dpms = radeon_connector_dpms,
+	.dpms = drm_helper_connector_dpms,
 	.detect = radeon_dp_mst_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.destroy = radeon_dp_mst_connector_destroy,
@@ -285,7 +290,7 @@ static struct drm_connector *radeon_dp_add_mst_connector(struct drm_dp_mst_topol
 
 	drm_object_attach_property(&connector->base, dev->mode_config.path_property, 0);
 	drm_object_attach_property(&connector->base, dev->mode_config.tile_property, 0);
-	drm_mode_connector_set_path_property(connector, pathprop);
+	drm_connector_set_path_property(connector, pathprop);
 
 	return connector;
 }
@@ -295,9 +300,7 @@ static void radeon_dp_register_mst_connector(struct drm_connector *connector)
 	struct drm_device *dev = connector->dev;
 	struct radeon_device *rdev = dev->dev_private;
 
-	drm_modeset_lock_all(dev);
 	radeon_fb_add_connector(rdev, connector);
-	drm_modeset_unlock_all(dev);
 
 	drm_connector_register(connector);
 }
@@ -310,13 +313,8 @@ static void radeon_dp_destroy_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 	struct radeon_device *rdev = dev->dev_private;
 
 	drm_connector_unregister(connector);
-	/* need to nuke the connector */
-	drm_modeset_lock_all(dev);
-	/* dpms off */
 	radeon_fb_remove_connector(rdev, connector);
-
 	drm_connector_cleanup(connector);
-	drm_modeset_unlock_all(dev);
 
 	kfree(connector);
 	DRM_DEBUG_KMS("\n");
@@ -330,14 +328,15 @@ static void radeon_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
 	drm_kms_helper_hotplug_event(dev);
 }
 
-struct drm_dp_mst_topology_cbs mst_cbs = {
+static const struct drm_dp_mst_topology_cbs mst_cbs = {
 	.add_connector = radeon_dp_add_mst_connector,
 	.register_connector = radeon_dp_register_mst_connector,
 	.destroy_connector = radeon_dp_destroy_mst_connector,
 	.hotplug = radeon_dp_mst_hotplug,
 };
 
-struct radeon_connector *radeon_mst_find_connector(struct drm_encoder *encoder)
+static struct
+radeon_connector *radeon_mst_find_connector(struct drm_encoder *encoder)
 {
 	struct drm_device *dev = encoder->dev;
 	struct drm_connector *connector;
@@ -395,7 +394,7 @@ radeon_mst_encoder_dpms(struct drm_encoder *encoder, int mode)
 	struct drm_crtc *crtc;
 	struct radeon_crtc *radeon_crtc;
 	int ret, slots;
-
+	s64 fixed_pbn, fixed_pbn_per_slot, avg_time_slots_per_mtp;
 	if (!ASIC_IS_DCE5(rdev)) {
 		DRM_ERROR("got mst dpms on non-DCE5\n");
 		return;
@@ -447,9 +446,11 @@ radeon_mst_encoder_dpms(struct drm_encoder *encoder, int mode)
 		DRM_DEBUG_KMS("dig encoder is %d %d %d\n", dig_enc->dig_encoder,
 			      dig_enc->linkb, radeon_crtc->crtc_id);
 
+		slots = drm_dp_find_vcpi_slots(&radeon_connector->mst_port->mst_mgr,
+					       mst_enc->pbn);
 		ret = drm_dp_mst_allocate_vcpi(&radeon_connector->mst_port->mst_mgr,
 					       radeon_connector->port,
-					       mst_enc->pbn, &slots);
+					       mst_enc->pbn, slots);
 		ret = drm_dp_update_payload_part1(&radeon_connector->mst_port->mst_mgr);
 
 		radeon_dp_mst_set_be_cntl(primary, mst_enc,
@@ -457,7 +458,11 @@ radeon_mst_encoder_dpms(struct drm_encoder *encoder, int mode)
 
 		mst_enc->enc_active = true;
 		radeon_dp_mst_update_stream_attribs(radeon_connector->mst_port, primary);
-		radeon_dp_mst_set_vcp_size(radeon_encoder, slots, 0);
+
+		fixed_pbn = drm_int2fixp(mst_enc->pbn);
+		fixed_pbn_per_slot = drm_int2fixp(radeon_connector->mst_port->mst_mgr.pbn_div);
+		avg_time_slots_per_mtp = drm_fixp_div(fixed_pbn, fixed_pbn_per_slot);
+		radeon_dp_mst_set_vcp_size(radeon_encoder, avg_time_slots_per_mtp);
 
 		atombios_dig_encoder_setup2(&primary->base, ATOM_ENCODER_CMD_DP_VIDEO_ON, 0,
 					    mst_enc->fe);
@@ -511,6 +516,7 @@ static bool radeon_mst_mode_fixup(struct drm_encoder *encoder,
 {
 	struct radeon_encoder_mst *mst_enc;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+	struct radeon_connector_atom_dig *dig_connector;
 	int bpp = 24;
 
 	mst_enc = radeon_encoder->enc_priv;
@@ -524,14 +530,11 @@ static bool radeon_mst_mode_fixup(struct drm_encoder *encoder,
 
 
 	drm_mode_set_crtcinfo(adjusted_mode, 0);
-	{
-	  struct radeon_connector_atom_dig *dig_connector;
-	  dig_connector = mst_enc->connector->con_priv;
-	  dig_connector->dp_lane_count = drm_dp_max_lane_count(dig_connector->dpcd);
-	  dig_connector->dp_clock = drm_dp_max_link_rate(dig_connector->dpcd);
-	  DRM_DEBUG_KMS("dig clock %p %d %d\n", dig_connector,
-			dig_connector->dp_lane_count, dig_connector->dp_clock);
-	}
+	dig_connector = mst_enc->connector->con_priv;
+	dig_connector->dp_lane_count = drm_dp_max_lane_count(dig_connector->dpcd);
+	dig_connector->dp_clock = drm_dp_max_link_rate(dig_connector->dpcd);
+	DRM_DEBUG_KMS("dig clock %p %d %d\n", dig_connector,
+		      dig_connector->dp_lane_count, dig_connector->dp_clock);
 	return true;
 }
 
@@ -591,7 +594,7 @@ static const struct drm_encoder_helper_funcs radeon_mst_helper_funcs = {
 	.commit = radeon_mst_encoder_commit,
 };
 
-void radeon_dp_mst_encoder_destroy(struct drm_encoder *encoder)
+static void radeon_dp_mst_encoder_destroy(struct drm_encoder *encoder)
 {
 	drm_encoder_cleanup(encoder);
 	kfree(encoder);
@@ -640,7 +643,7 @@ radeon_dp_create_fake_mst_encoder(struct radeon_connector *connector)
 	}
 
 	drm_encoder_init(dev, &radeon_encoder->base, &radeon_dp_mst_enc_funcs,
-			 DRM_MODE_ENCODER_DPMST);
+			 DRM_MODE_ENCODER_DPMST, NULL);
 	drm_encoder_helper_add(encoder, &radeon_mst_helper_funcs);
 
 	mst_enc = radeon_encoder->enc_priv;
@@ -659,7 +662,7 @@ radeon_dp_mst_init(struct radeon_connector *radeon_connector)
 		return 0;
 
 	radeon_connector->mst_mgr.cbs = &mst_cbs;
-	return drm_dp_mst_topology_mgr_init(&radeon_connector->mst_mgr, dev->dev,
+	return drm_dp_mst_topology_mgr_init(&radeon_connector->mst_mgr, dev,
 					    &radeon_connector->ddc_bus->aux, 16, 6,
 					    radeon_connector->base.base.id);
 }
@@ -715,7 +718,7 @@ radeon_dp_mst_check_status(struct radeon_connector *radeon_connector)
 				       DP_SINK_COUNT_ESI, esi, 8);
 go_again:
 		if (dret == 8) {
-			DRM_DEBUG_KMS("got esi %02x %02x %02x\n", esi[0], esi[1], esi[2]);
+			DRM_DEBUG_KMS("got esi %3ph\n", esi);
 			ret = drm_dp_mst_hpd_irq(&radeon_connector->mst_mgr, esi, &handled);
 
 			if (handled) {
@@ -730,7 +733,7 @@ go_again:
 				dret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux,
 							DP_SINK_COUNT_ESI, esi, 8);
 				if (dret == 8) {
-					DRM_DEBUG_KMS("got esi2 %02x %02x %02x\n", esi[0], esi[1], esi[2]);
+					DRM_DEBUG_KMS("got esi2 %3ph\n", esi);
 					goto go_again;
 				}
 			} else
diff --git a/sys/dev/pci/drm/radeon/radeon_drv.c b/sys/dev/pci/drm/radeon/radeon_drv.c
index b57ac74a7d6..d6cd19d4a4d 100644
--- a/sys/dev/pci/drm/radeon/radeon_drv.c
+++ b/sys/dev/pci/drm/radeon/radeon_drv.c
@@ -29,14 +29,21 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_drv.h"
 #include "radeon.h"
 
-#include <dev/pci/drm/drm_pciids.h>
+#include <drm/drm_pciids.h>
+#include <linux/console.h>
+#include <linux/module.h>
+#include <linux/pm_runtime.h>
+#include <linux/vga_switcheroo.h>
+#include <linux/compat.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_fb_helper.h>
 
-#include <dev/pci/drm/drm_crtc_helper.h>
+#include <drm/drm_crtc_helper.h>
 #include "radeon_kfd.h"
 
 /*
@@ -87,27 +94,29 @@
  *   2.41.0 - evergreen/cayman: Add SET_BASE/DRAW_INDIRECT command parsing support
  *   2.42.0 - Add VCE/VUI (Video Usability Information) support
  *   2.43.0 - RADEON_INFO_GPU_RESET_COUNTER
+ *   2.44.0 - SET_APPEND_CNT packet3 support
+ *   2.45.0 - Allow setting shader registers using DMA/COPY packet3 on SI
+ *   2.46.0 - Add PFP_SYNC_ME support on evergreen
+ *   2.47.0 - Add UVD_NO_OP register support
+ *   2.48.0 - TA_CS_BC_BASE_ADDR allowed on SI
+ *   2.49.0 - DRM_RADEON_GEM_INFO ioctl returns correct vram_size/visible values
+ *   2.50.0 - Allows unaligned shader loads on CIK. (needed by OpenGL)
  */
 #define KMS_DRIVER_MAJOR	2
-#define KMS_DRIVER_MINOR	43
+#define KMS_DRIVER_MINOR	50
 #define KMS_DRIVER_PATCHLEVEL	0
 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags);
-int radeon_driver_unload_kms(struct drm_device *dev);
+void radeon_driver_unload_kms(struct drm_device *dev);
 void radeon_driver_lastclose_kms(struct drm_device *dev);
 int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv);
 void radeon_driver_postclose_kms(struct drm_device *dev,
 				 struct drm_file *file_priv);
-void radeon_driver_preclose_kms(struct drm_device *dev,
-				struct drm_file *file_priv);
-int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon);
+int radeon_suspend_kms(struct drm_device *dev, bool suspend,
+		       bool fbcon, bool freeze);
 int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
 u32 radeon_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe);
 int radeon_enable_vblank_kms(struct drm_device *dev, unsigned int pipe);
 void radeon_disable_vblank_kms(struct drm_device *dev, unsigned int pipe);
-int radeon_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags);
 void radeon_driver_irq_preinstall_kms(struct drm_device *dev);
 int radeon_driver_irq_postinstall_kms(struct drm_device *dev);
 void radeon_driver_irq_uninstall_kms(struct drm_device *dev);
@@ -141,26 +150,25 @@ struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
 							struct sg_table *sg);
 int radeon_gem_prime_pin(struct drm_gem_object *obj);
 void radeon_gem_prime_unpin(struct drm_gem_object *obj);
+#endif
 struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *);
+#ifdef notyet
 void *radeon_gem_prime_vmap(struct drm_gem_object *obj);
 void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
 #endif
-extern long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd,
-				    unsigned long arg);
-
-#if defined(CONFIG_DEBUG_FS)
-int radeon_debugfs_init(struct drm_minor *minor);
-void radeon_debugfs_cleanup(struct drm_minor *minor);
-#endif
 
 /* atpx handler */
 #if defined(CONFIG_VGA_SWITCHEROO)
 void radeon_register_atpx_handler(void);
 void radeon_unregister_atpx_handler(void);
+bool radeon_has_atpx_dgpu_power_cntl(void);
+bool radeon_is_atpx_hybrid(void);
 #else
 #ifdef notyet
 static inline void radeon_register_atpx_handler(void) {}
 static inline void radeon_unregister_atpx_handler(void) {}
+static inline bool radeon_has_atpx_dgpu_power_cntl(void) { return false; }
+static inline bool radeon_is_atpx_hybrid(void) { return false; }
 #endif
 #endif
 
@@ -168,7 +176,12 @@ int radeon_no_wb;
 int radeon_modeset = -1;
 int radeon_dynclks = -1;
 int radeon_r4xx_atom = 0;
+#ifdef __powerpc__
+/* Default to PCI on PowerPC (fdo #95017) */
+int radeon_agpmode = -1;
+#else
 int radeon_agpmode = 0;
+#endif
 int radeon_vram_limit = 0;
 int radeon_gart_size = -1; /* auto */
 int radeon_benchmarking = 0;
@@ -194,6 +207,8 @@ int radeon_bapm = -1;
 int radeon_backlight = -1;
 int radeon_auxch = -1;
 int radeon_mst = 0;
+int radeon_uvd = 1;
+int radeon_vce = 1;
 
 MODULE_PARM_DESC(no_wb, "Disable AGP writeback for scratch registers");
 module_param_named(no_wb, radeon_no_wb, int, 0444);
@@ -285,98 +300,32 @@ module_param_named(auxch, radeon_auxch, int, 0444);
 MODULE_PARM_DESC(mst, "DisplayPort MST experimental support (1 = enable, 0 = disable)");
 module_param_named(mst, radeon_mst, int, 0444);
 
-const struct drm_pcidev radeondrm_pciidlist[] = {
-	radeon_PCI_IDS
-};
+MODULE_PARM_DESC(uvd, "uvd enable/disable uvd support (1 = enable, 0 = disable)");
+module_param_named(uvd, radeon_uvd, int, 0444);
 
-MODULE_DEVICE_TABLE(pci, pciidlist);
-
-#ifdef CONFIG_DRM_RADEON_UMS
-
-static int radeon_suspend(struct drm_device *dev, pm_message_t state)
-{
-	drm_radeon_private_t *dev_priv = dev->dev_private;
-
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600)
-		return 0;
+MODULE_PARM_DESC(vce, "vce enable/disable vce support (1 = enable, 0 = disable)");
+module_param_named(vce, radeon_vce, int, 0444);
 
-	/* Disable *all* interrupts */
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RS600)
-		RADEON_WRITE(R500_DxMODE_INT_MASK, 0);
-	RADEON_WRITE(RADEON_GEN_INT_CNTL, 0);
-	return 0;
-}
-
-static int radeon_resume(struct drm_device *dev)
-{
-	drm_radeon_private_t *dev_priv = dev->dev_private;
+int radeon_si_support = 1;
+MODULE_PARM_DESC(si_support, "SI support (1 = enabled (default), 0 = disabled)");
+module_param_named(si_support, radeon_si_support, int, 0444);
 
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600)
-		return 0;
+int radeon_cik_support = 1;
+MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled)");
+module_param_named(cik_support, radeon_cik_support, int, 0444);
 
-	/* Restore interrupt registers */
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RS600)
-		RADEON_WRITE(R500_DxMODE_INT_MASK, dev_priv->r500_disp_irq_reg);
-	RADEON_WRITE(RADEON_GEN_INT_CNTL, dev_priv->irq_enable_reg);
-	return 0;
-}
-
-
-static const struct file_operations radeon_driver_old_fops = {
-	.owner = THIS_MODULE,
-	.open = drm_open,
-	.release = drm_release,
-	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_legacy_mmap,
-	.poll = drm_poll,
-	.read = drm_read,
-#ifdef CONFIG_COMPAT
-	.compat_ioctl = radeon_compat_ioctl,
-#endif
-	.llseek = noop_llseek,
-};
-
-static struct drm_driver driver_old = {
-	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_PCI_DMA | DRIVER_SG |
-	    DRIVER_HAVE_IRQ | DRIVER_HAVE_DMA | DRIVER_IRQ_SHARED,
-	.dev_priv_size = sizeof(drm_radeon_buf_priv_t),
-	.load = radeon_driver_load,
-	.firstopen = radeon_driver_firstopen,
-	.open = radeon_driver_open,
-	.preclose = radeon_driver_preclose,
-	.postclose = radeon_driver_postclose,
-	.lastclose = radeon_driver_lastclose,
-	.set_busid = drm_pci_set_busid,
-	.unload = radeon_driver_unload,
-	.suspend = radeon_suspend,
-	.resume = radeon_resume,
-	.get_vblank_counter = radeon_get_vblank_counter,
-	.enable_vblank = radeon_enable_vblank,
-	.disable_vblank = radeon_disable_vblank,
-	.master_create = radeon_master_create,
-	.master_destroy = radeon_master_destroy,
-	.irq_preinstall = radeon_driver_irq_preinstall,
-	.irq_postinstall = radeon_driver_irq_postinstall,
-	.irq_uninstall = radeon_driver_irq_uninstall,
-	.irq_handler = radeon_driver_irq_handler,
-	.ioctls = radeon_ioctls,
-	.dma_ioctl = radeon_cp_buffers,
-	.fops = &radeon_driver_old_fops,
-	.name = DRIVER_NAME,
-	.desc = DRIVER_DESC,
-	.date = DRIVER_DATE,
-	.major = DRIVER_MAJOR,
-	.minor = DRIVER_MINOR,
-	.patchlevel = DRIVER_PATCHLEVEL,
+const struct drm_pcidev radeondrm_pciidlist[] = {
+	radeon_PCI_IDS
 };
 
-#endif
+MODULE_DEVICE_TABLE(pci, pciidlist);
 
 #ifdef notyet
 static struct drm_driver kms_driver;
 #endif
 
+bool radeon_device_is_virtual(void);
+
 #ifdef notyet
 static int radeon_kick_out_firmware_fb(struct pci_dev *pdev)
 {
@@ -393,7 +342,7 @@ static int radeon_kick_out_firmware_fb(struct pci_dev *pdev)
 #ifdef CONFIG_X86
 	primary = pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW;
 #endif
-	remove_conflicting_framebuffers(ap, "radeondrmfb", primary);
+	drm_fb_helper_remove_conflicting_framebuffers(ap, "radeondrmfb", primary);
 	kfree(ap);
 
 	return 0;
@@ -406,6 +355,9 @@ static int radeon_pci_probe(struct pci_dev *pdev,
 {
 	int ret;
 
+	if (vga_switcheroo_client_probe_defer(pdev))
+		return -EPROBE_DEFER;
+
 	/* Get rid of things like offb */
 	ret = radeon_kick_out_firmware_fb(pdev);
 	if (ret)
@@ -422,17 +374,35 @@ radeon_pci_remove(struct pci_dev *pdev)
 	drm_put_dev(dev);
 }
 
+static void
+radeon_pci_shutdown(struct pci_dev *pdev)
+{
+	/* if we are running in a VM, make sure the device
+	 * torn down properly on reboot/shutdown
+	 */
+	if (radeon_device_is_virtual())
+		radeon_pci_remove(pdev);
+}
+
 static int radeon_pmops_suspend(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	return radeon_suspend_kms(drm_dev, true, true);
+	return radeon_suspend_kms(drm_dev, true, true, false);
 }
 
 static int radeon_pmops_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+
+	/* GPU comes up enabled by the bios on resume */
+	if (radeon_is_px(drm_dev)) {
+		pm_runtime_disable(dev);
+		pm_runtime_set_active(dev);
+		pm_runtime_enable(dev);
+	}
+
 	return radeon_resume_kms(drm_dev, true, true);
 }
 
@@ -440,7 +410,7 @@ static int radeon_pmops_freeze(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	return radeon_suspend_kms(drm_dev, false, true);
+	return radeon_suspend_kms(drm_dev, false, true, true);
 }
 
 static int radeon_pmops_thaw(struct device *dev)
@@ -463,13 +433,15 @@ static int radeon_pmops_runtime_suspend(struct device *dev)
 
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 	drm_kms_helper_poll_disable(drm_dev);
-	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
 
-	ret = radeon_suspend_kms(drm_dev, false, false);
+	ret = radeon_suspend_kms(drm_dev, false, false, false);
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 	pci_ignore_hotplug(pdev);
-	pci_set_power_state(pdev, PCI_D3cold);
+	if (radeon_is_atpx_hybrid())
+		pci_set_power_state(pdev, PCI_D3cold);
+	else if (!radeon_has_atpx_dgpu_power_cntl())
+		pci_set_power_state(pdev, PCI_D3hot);
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF;
 
 	return 0;
@@ -486,7 +458,9 @@ static int radeon_pmops_runtime_resume(struct device *dev)
 
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 
-	pci_set_power_state(pdev, PCI_D0);
+	if (radeon_is_atpx_hybrid() ||
+	    !radeon_has_atpx_dgpu_power_cntl())
+		pci_set_power_state(pdev, PCI_D0);
 	pci_restore_state(pdev);
 	ret = pci_enable_device(pdev);
 	if (ret)
@@ -495,7 +469,6 @@ static int radeon_pmops_runtime_resume(struct device *dev)
 
 	ret = radeon_resume_kms(drm_dev, false, false);
 	drm_kms_helper_poll_enable(drm_dev);
-	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON);
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
 	return 0;
 }
@@ -542,6 +515,21 @@ long radeon_drm_ioctl(struct file *filp,
 	return ret;
 }
 
+#ifdef CONFIG_COMPAT
+static long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+{
+	unsigned int nr = DRM_IOCTL_NR(cmd);
+	int ret;
+
+	if (nr < DRM_COMMAND_BASE)
+		return drm_compat_ioctl(filp, cmd, arg);
+
+	ret = radeon_drm_ioctl(filp, cmd, arg);
+
+	return ret;
+}
+#endif
+
 static const struct dev_pm_ops radeon_pm_ops = {
 	.suspend = radeon_pmops_suspend,
 	.resume = radeon_pmops_resume,
@@ -568,6 +556,16 @@ static const struct file_operations radeon_driver_kms_fops = {
 };
 #endif /* __linux__ */
 
+static bool
+radeon_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
+				 bool in_vblank_irq, int *vpos, int *hpos,
+				 ktime_t *stime, ktime_t *etime,
+				 const struct drm_display_mode *mode)
+{
+	return radeon_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos,
+					  stime, etime, mode);
+}
+
 struct drm_driver kms_driver = {
 	.driver_features =
 	    DRIVER_USE_AGP |
@@ -578,34 +576,27 @@ struct drm_driver kms_driver = {
 #endif
 	.open = radeon_driver_open_kms,
 	.mmap = radeon_mmap,
-	.preclose = radeon_driver_preclose_kms,
 	.postclose = radeon_driver_postclose_kms,
 	.lastclose = radeon_driver_lastclose_kms,
 #ifdef notyet
-	.set_busid = drm_pci_set_busid,
 	.unload = radeon_driver_unload_kms,
 #endif
 	.get_vblank_counter = radeon_get_vblank_counter_kms,
 	.enable_vblank = radeon_enable_vblank_kms,
 	.disable_vblank = radeon_disable_vblank_kms,
-	.get_vblank_timestamp = radeon_get_vblank_timestamp_kms,
-	.get_scanout_position = radeon_get_crtc_scanoutpos,
-#if defined(CONFIG_DEBUG_FS)
-	.debugfs_init = radeon_debugfs_init,
-	.debugfs_cleanup = radeon_debugfs_cleanup,
-#endif
+	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
+	.get_scanout_position = radeon_get_crtc_scanout_position,
 	.irq_preinstall = radeon_driver_irq_preinstall_kms,
 	.irq_postinstall = radeon_driver_irq_postinstall_kms,
 	.irq_uninstall = radeon_driver_irq_uninstall_kms,
 	.irq_handler = radeon_driver_irq_handler_kms,
 	.ioctls = radeon_ioctls_kms,
-	.gem_free_object = radeon_gem_object_free,
+	.gem_free_object_unlocked = radeon_gem_object_free,
 	.gem_open_object = radeon_gem_object_open,
 	.gem_close_object = radeon_gem_object_close,
 	.gem_size = sizeof(struct radeon_bo),
 	.dumb_create = radeon_mode_dumb_create,
 	.dumb_map_offset = radeon_mode_dumb_mmap,
-	.dumb_destroy = drm_gem_dumb_destroy,
 #ifdef __linux__
 	.fops = &radeon_driver_kms_fops,
 #endif
@@ -617,7 +608,9 @@ struct drm_driver kms_driver = {
 #ifdef notyet
 	.gem_prime_pin = radeon_gem_prime_pin,
 	.gem_prime_unpin = radeon_gem_prime_unpin,
+#endif
 	.gem_prime_res_obj = radeon_gem_prime_res_obj,
+#ifdef notyet
 	.gem_prime_get_sg_table = radeon_gem_prime_get_sg_table,
 	.gem_prime_import_sg_table = radeon_gem_prime_import_sg_table,
 	.gem_prime_vmap = radeon_gem_prime_vmap,
@@ -638,18 +631,12 @@ static struct drm_driver *driver;
 #ifdef __linux__
 static struct pci_driver *pdriver;
 
-#ifdef CONFIG_DRM_RADEON_UMS
-static struct pci_driver radeon_pci_driver = {
-	.name = DRIVER_NAME,
-	.id_table = pciidlist,
-};
-#endif
-
 static struct pci_driver radeon_kms_pci_driver = {
 	.name = DRIVER_NAME,
 	.id_table = pciidlist,
 	.probe = radeon_pci_probe,
 	.remove = radeon_pci_remove,
+	.shutdown = radeon_pci_shutdown,
 	.driver.pm = &radeon_pm_ops,
 };
 #endif /* __linux__ */
@@ -657,12 +644,10 @@ static struct pci_driver radeon_kms_pci_driver = {
 #ifdef notyet
 static int __init radeon_init(void)
 {
-#ifdef CONFIG_VGA_CONSOLE
 	if (vgacon_text_force() && radeon_modeset == -1) {
 		DRM_INFO("VGACON disable radeon kernel modesetting.\n");
 		radeon_modeset = 0;
 	}
-#endif
 	/* set to modesetting by default if not nomodeset */
 	if (radeon_modeset == -1)
 		radeon_modeset = 1;
@@ -678,23 +663,12 @@ static int __init radeon_init(void)
 		radeon_register_atpx_handler();
 
 	} else {
-#ifdef CONFIG_DRM_RADEON_UMS
-		DRM_INFO("radeon userspace modesetting enabled.\n");
-		driver = &driver_old;
-		pdriver = &radeon_pci_driver;
-		driver->driver_features &= ~DRIVER_MODESET;
-		driver->num_ioctls = radeon_max_ioctl;
-#else
 		DRM_ERROR("No UMS support in radeon module!\n");
 		return -EINVAL;
-#endif
 	}
 
-	radeon_kfd_init();
-
-	/* let modprobe override vga console setting */
 #ifdef notyet
-	return drm_pci_init(driver, pdriver);
+	return pci_register_driver(pdriver);
 #else
 	STUB();
 	return -ENOSYS;
@@ -704,9 +678,8 @@ static int __init radeon_init(void)
 static void __exit radeon_exit(void)
 {
 	STUB();
-	radeon_kfd_fini();
 #ifdef notyet
-	drm_pci_exit(driver, pdriver);
+	pci_unregister_driver(pdriver);
 #endif
 	radeon_unregister_atpx_handler();
 }
diff --git a/sys/dev/pci/drm/radeon/radeon_drv.h b/sys/dev/pci/drm/radeon/radeon_drv.h
index 1b962db8dfc..9bd0e9c0e64 100644
--- a/sys/dev/pci/drm/radeon/radeon_drv.h
+++ b/sys/dev/pci/drm/radeon/radeon_drv.h
@@ -31,6 +31,13 @@
 #ifndef __RADEON_DRV_H__
 #define __RADEON_DRV_H__
 
+#include <linux/firmware.h>
+#include <linux/platform_device.h>
+#ifdef __linux__
+#include <drm/drm_legacy.h>
+
+#include <drm/ati_pcigart.h>
+#endif
 #include "radeon_family.h"
 
 /* General customization:
@@ -114,2052 +121,4 @@
 long radeon_drm_ioctl(struct file *filp,
 		      unsigned int cmd, unsigned long arg);
 
-/* The rest of the file is DEPRECATED! */
-#ifdef CONFIG_DRM_RADEON_UMS
-
-enum radeon_cp_microcode_version {
-	UCODE_R100,
-	UCODE_R200,
-	UCODE_R300,
-};
-
-typedef struct drm_radeon_freelist {
-	unsigned int age;
-	struct drm_buf *buf;
-	struct drm_radeon_freelist *next;
-	struct drm_radeon_freelist *prev;
-} drm_radeon_freelist_t;
-
-typedef struct drm_radeon_ring_buffer {
-	u32 *start;
-	u32 *end;
-	int size;
-	int size_l2qw;
-
-	int rptr_update; /* Double Words */
-	int rptr_update_l2qw; /* log2 Quad Words */
-
-	int fetch_size; /* Double Words */
-	int fetch_size_l2ow; /* log2 Oct Words */
-
-	u32 tail;
-	u32 tail_mask;
-	int space;
-
-	int high_mark;
-} drm_radeon_ring_buffer_t;
-
-typedef struct drm_radeon_depth_clear_t {
-	u32 rb3d_cntl;
-	u32 rb3d_zstencilcntl;
-	u32 se_cntl;
-} drm_radeon_depth_clear_t;
-
-struct drm_radeon_driver_file_fields {
-	int64_t radeon_fb_delta;
-};
-
-struct mem_block {
-	struct mem_block *next;
-	struct mem_block *prev;
-	int start;
-	int size;
-	struct drm_file *file_priv; /* NULL: free, -1: heap, other: real files */
-};
-
-struct radeon_surface {
-	int refcount;
-	u32 lower;
-	u32 upper;
-	u32 flags;
-};
-
-struct radeon_virt_surface {
-	int surface_index;
-	u32 lower;
-	u32 upper;
-	u32 flags;
-	struct drm_file *file_priv;
-#define PCIGART_FILE_PRIV	((void *) -1L)
-};
-
-#define RADEON_FLUSH_EMITED	(1 << 0)
-#define RADEON_PURGE_EMITED	(1 << 1)
-
-struct drm_radeon_master_private {
-	drm_local_map_t *sarea;
-	drm_radeon_sarea_t *sarea_priv;
-};
-
-typedef struct drm_radeon_private {
-	drm_radeon_ring_buffer_t ring;
-
-	u32 fb_location;
-	u32 fb_size;
-	int new_memmap;
-
-	int gart_size;
-	u32 gart_vm_start;
-	unsigned long gart_buffers_offset;
-
-	int cp_mode;
-	int cp_running;
-
-	drm_radeon_freelist_t *head;
-	drm_radeon_freelist_t *tail;
-	int last_buf;
-	int writeback_works;
-
-	int usec_timeout;
-
-	int microcode_version;
-
-	struct {
-		u32 boxes;
-		int freelist_timeouts;
-		int freelist_loops;
-		int requested_bufs;
-		int last_frame_reads;
-		int last_clear_reads;
-		int clears;
-		int texture_uploads;
-	} stats;
-
-	int do_boxes;
-	int page_flipping;
-
-	u32 color_fmt;
-	unsigned int front_offset;
-	unsigned int front_pitch;
-	unsigned int back_offset;
-	unsigned int back_pitch;
-
-	u32 depth_fmt;
-	unsigned int depth_offset;
-	unsigned int depth_pitch;
-
-	u32 front_pitch_offset;
-	u32 back_pitch_offset;
-	u32 depth_pitch_offset;
-
-	drm_radeon_depth_clear_t depth_clear;
-
-	unsigned long ring_offset;
-	unsigned long ring_rptr_offset;
-	unsigned long buffers_offset;
-	unsigned long gart_textures_offset;
-
-	drm_local_map_t *sarea;
-	drm_local_map_t *cp_ring;
-	drm_local_map_t *ring_rptr;
-	drm_local_map_t *gart_textures;
-
-	struct mem_block *gart_heap;
-	struct mem_block *fb_heap;
-
-	/* SW interrupt */
-	wait_queue_head_t swi_queue;
-	atomic_t swi_emitted;
-	int vblank_crtc;
-	uint32_t irq_enable_reg;
-	uint32_t r500_disp_irq_reg;
-
-	struct radeon_surface surfaces[RADEON_MAX_SURFACES];
-	struct radeon_virt_surface virt_surfaces[2 * RADEON_MAX_SURFACES];
-
-	unsigned long pcigart_offset;
-	unsigned int pcigart_offset_set;
-	struct drm_ati_pcigart_info gart_info;
-
-	u32 scratch_ages[5];
-
-	int have_z_offset;
-
-	/* starting from here on, data is preserved across an open */
-	uint32_t flags;		/* see radeon_chip_flags */
-	resource_size_t fb_aper_offset;
-
-	int num_gb_pipes;
-	int num_z_pipes;
-	int track_flush;
-	drm_local_map_t *mmio;
-
-	/* r6xx/r7xx pipe/shader config */
-	int r600_max_pipes;
-	int r600_max_tile_pipes;
-	int r600_max_simds;
-	int r600_max_backends;
-	int r600_max_gprs;
-	int r600_max_threads;
-	int r600_max_stack_entries;
-	int r600_max_hw_contexts;
-	int r600_max_gs_threads;
-	int r600_sx_max_export_size;
-	int r600_sx_max_export_pos_size;
-	int r600_sx_max_export_smx_size;
-	int r600_sq_num_cf_insts;
-	int r700_sx_num_of_sets;
-	int r700_sc_prim_fifo_size;
-	int r700_sc_hiz_tile_fifo_size;
-	int r700_sc_earlyz_tile_fifo_fize;
-	int r600_group_size;
-	int r600_npipes;
-	int r600_nbanks;
-
-	struct rwlock cs_mutex;
-	u32 cs_id_scnt;
-	u32 cs_id_wcnt;
-	/* r6xx/r7xx drm blit vertex buffer */
-	struct drm_buf *blit_vb;
-
-	/* firmware */
-	const struct firmware *me_fw, *pfp_fw;
-} drm_radeon_private_t;
-
-typedef struct drm_radeon_buf_priv {
-	u32 age;
-} drm_radeon_buf_priv_t;
-
-struct drm_buffer;
-
-typedef struct drm_radeon_kcmd_buffer {
-	int bufsz;
-	struct drm_buffer *buffer;
-	int nbox;
-	struct drm_clip_rect __user *boxes;
-} drm_radeon_kcmd_buffer_t;
-
-extern int radeon_no_wb;
-extern struct drm_ioctl_desc radeon_ioctls[];
-extern int radeon_max_ioctl;
-
-extern u32 radeon_get_ring_head(drm_radeon_private_t *dev_priv);
-extern void radeon_set_ring_head(drm_radeon_private_t *dev_priv, u32 val);
-
-#define GET_RING_HEAD(dev_priv)	radeon_get_ring_head(dev_priv)
-#define SET_RING_HEAD(dev_priv, val) radeon_set_ring_head(dev_priv, val)
-
-/* Check whether the given hardware address is inside the framebuffer or the
- * GART area.
- */
-static __inline__ int radeon_check_offset(drm_radeon_private_t *dev_priv,
-					  u64 off)
-{
-	u32 fb_start = dev_priv->fb_location;
-	u32 fb_end = fb_start + dev_priv->fb_size - 1;
-	u32 gart_start = dev_priv->gart_vm_start;
-	u32 gart_end = gart_start + dev_priv->gart_size - 1;
-
-	return ((off >= fb_start && off <= fb_end) ||
-		(off >= gart_start && off <= gart_end));
-}
-
-/* radeon_state.c */
-extern void radeon_cp_discard_buffer(struct drm_device *dev, struct drm_master *master, struct drm_buf *buf);
-
-				/* radeon_cp.c */
-extern int radeon_cp_init(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_start(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_stop(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_reset(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_idle(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_resume(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_engine_reset(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_fullscreen(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_cp_buffers(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern u32 radeon_read_fb_location(drm_radeon_private_t *dev_priv);
-extern void radeon_write_agp_location(drm_radeon_private_t *dev_priv, u32 agp_loc);
-extern void radeon_write_agp_base(drm_radeon_private_t *dev_priv, u64 agp_base);
-
-extern void radeon_freelist_reset(struct drm_device * dev);
-extern struct drm_buf *radeon_freelist_get(struct drm_device * dev);
-
-extern int radeon_wait_ring(drm_radeon_private_t * dev_priv, int n);
-
-extern int radeon_do_cp_idle(drm_radeon_private_t * dev_priv);
-
-extern int radeon_driver_preinit(struct drm_device *dev, unsigned long flags);
-extern int radeon_presetup(struct drm_device *dev);
-extern int radeon_driver_postcleanup(struct drm_device *dev);
-
-extern int radeon_mem_alloc(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_mem_free(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_mem_init_heap(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern void radeon_mem_takedown(struct mem_block **heap);
-extern void radeon_mem_release(struct drm_file *file_priv,
-			       struct mem_block *heap);
-
-extern void radeon_enable_bm(struct drm_radeon_private *dev_priv);
-extern u32 radeon_read_ring_rptr(drm_radeon_private_t *dev_priv, u32 off);
-extern void radeon_write_ring_rptr(drm_radeon_private_t *dev_priv, u32 off, u32 val);
-
-				/* radeon_irq.c */
-extern void radeon_irq_set_state(struct drm_device *dev, u32 mask, int state);
-extern int radeon_irq_emit(struct drm_device *dev, void *data, struct drm_file *file_priv);
-extern int radeon_irq_wait(struct drm_device *dev, void *data, struct drm_file *file_priv);
-
-extern void radeon_do_release(struct drm_device * dev);
-extern u32 radeon_get_vblank_counter(struct drm_device *dev, unsigned int pipe);
-extern int radeon_enable_vblank(struct drm_device *dev, unsigned int pipe);
-extern void radeon_disable_vblank(struct drm_device *dev, unsigned int pipe);
-extern irqreturn_t radeon_driver_irq_handler(int irq, void *arg);
-extern void radeon_driver_irq_preinstall(struct drm_device * dev);
-extern int radeon_driver_irq_postinstall(struct drm_device *dev);
-extern void radeon_driver_irq_uninstall(struct drm_device * dev);
-extern void radeon_enable_interrupt(struct drm_device *dev);
-extern int radeon_vblank_crtc_get(struct drm_device *dev);
-extern int radeon_vblank_crtc_set(struct drm_device *dev, int64_t value);
-
-extern int radeon_driver_load(struct drm_device *dev, unsigned long flags);
-extern int radeon_driver_unload(struct drm_device *dev);
-extern int radeon_driver_firstopen(struct drm_device *dev);
-extern void radeon_driver_preclose(struct drm_device *dev,
-				   struct drm_file *file_priv);
-extern void radeon_driver_postclose(struct drm_device *dev,
-				    struct drm_file *file_priv);
-extern void radeon_driver_lastclose(struct drm_device * dev);
-extern int radeon_driver_open(struct drm_device *dev,
-			      struct drm_file *file_priv);
-extern long radeon_compat_ioctl(struct file *filp, unsigned int cmd,
-				unsigned long arg);
-
-extern int radeon_master_create(struct drm_device *dev, struct drm_master *master);
-extern void radeon_master_destroy(struct drm_device *dev, struct drm_master *master);
-extern void radeon_cp_dispatch_flip(struct drm_device *dev, struct drm_master *master);
-/* r300_cmdbuf.c */
-extern void r300_init_reg_flags(struct drm_device *dev);
-
-extern int r300_do_cp_cmdbuf(struct drm_device *dev,
-			     struct drm_file *file_priv,
-			     drm_radeon_kcmd_buffer_t *cmdbuf);
-
-/* r600_cp.c */
-extern int r600_do_engine_reset(struct drm_device *dev);
-extern int r600_do_cleanup_cp(struct drm_device *dev);
-extern int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init,
-			   struct drm_file *file_priv);
-extern int r600_do_resume_cp(struct drm_device *dev, struct drm_file *file_priv);
-extern int r600_do_cp_idle(drm_radeon_private_t *dev_priv);
-extern void r600_do_cp_start(drm_radeon_private_t *dev_priv);
-extern void r600_do_cp_reset(drm_radeon_private_t *dev_priv);
-extern void r600_do_cp_stop(drm_radeon_private_t *dev_priv);
-extern int r600_cp_dispatch_indirect(struct drm_device *dev,
-				     struct drm_buf *buf, int start, int end);
-extern int r600_page_table_init(struct drm_device *dev);
-extern void r600_page_table_cleanup(struct drm_device *dev, struct drm_ati_pcigart_info *gart_info);
-extern int r600_cs_legacy_ioctl(struct drm_device *dev, void *data, struct drm_file *fpriv);
-extern void r600_cp_dispatch_swap(struct drm_device *dev, struct drm_file *file_priv);
-extern int r600_cp_dispatch_texture(struct drm_device *dev,
-				    struct drm_file *file_priv,
-				    drm_radeon_texture_t *tex,
-				    drm_radeon_tex_image_t *image);
-/* r600_blit.c */
-extern int r600_prepare_blit_copy(struct drm_device *dev, struct drm_file *file_priv);
-extern void r600_done_blit_copy(struct drm_device *dev);
-extern void r600_blit_copy(struct drm_device *dev,
-			   uint64_t src_gpu_addr, uint64_t dst_gpu_addr,
-			   int size_bytes);
-extern void r600_blit_swap(struct drm_device *dev,
-			   uint64_t src_gpu_addr, uint64_t dst_gpu_addr,
-			   int sx, int sy, int dx, int dy,
-			   int w, int h, int src_pitch, int dst_pitch, int cpp);
-
-/* Flags for stats.boxes
- */
-#define RADEON_BOX_DMA_IDLE      0x1
-#define RADEON_BOX_RING_FULL     0x2
-#define RADEON_BOX_FLIP          0x4
-#define RADEON_BOX_WAIT_IDLE     0x8
-#define RADEON_BOX_TEXTURE_LOAD  0x10
-
-/* Register definitions, register access macros and drmAddMap constants
- * for Radeon kernel driver.
- */
-#define RADEON_MM_INDEX		        0x0000
-#define RADEON_MM_DATA		        0x0004
-
-#define RADEON_AGP_COMMAND		0x0f60
-#define RADEON_AGP_COMMAND_PCI_CONFIG   0x0060	/* offset in PCI config */
-#	define RADEON_AGP_ENABLE	(1<<8)
-#define RADEON_AUX_SCISSOR_CNTL		0x26f0
-#	define RADEON_EXCLUSIVE_SCISSOR_0	(1 << 24)
-#	define RADEON_EXCLUSIVE_SCISSOR_1	(1 << 25)
-#	define RADEON_EXCLUSIVE_SCISSOR_2	(1 << 26)
-#	define RADEON_SCISSOR_0_ENABLE		(1 << 28)
-#	define RADEON_SCISSOR_1_ENABLE		(1 << 29)
-#	define RADEON_SCISSOR_2_ENABLE		(1 << 30)
-
-/*
- * PCIE radeons (rv370/rv380, rv410, r423/r430/r480, r5xx)
- * don't have an explicit bus mastering disable bit.  It's handled
- * by the PCI D-states.  PMI_BM_DIS disables D-state bus master
- * handling, not bus mastering itself.
- */
-#define RADEON_BUS_CNTL			0x0030
-/* r1xx, r2xx, r300, r(v)350, r420/r481, rs400/rs480 */
-#	define RADEON_BUS_MASTER_DIS		(1 << 6)
-/* rs600/rs690/rs740 */
-#	define RS600_BUS_MASTER_DIS		(1 << 14)
-#	define RS600_MSI_REARM		        (1 << 20)
-/* see RS400_MSI_REARM in AIC_CNTL for rs480 */
-
-#define RADEON_BUS_CNTL1		0x0034
-#	define RADEON_PMI_BM_DIS		(1 << 2)
-#	define RADEON_PMI_INT_DIS		(1 << 3)
-
-#define RV370_BUS_CNTL			0x004c
-#	define RV370_PMI_BM_DIS		        (1 << 5)
-#	define RV370_PMI_INT_DIS		(1 << 6)
-
-#define RADEON_MSI_REARM_EN		0x0160
-/* rv370/rv380, rv410, r423/r430/r480, r5xx */
-#	define RV370_MSI_REARM_EN		(1 << 0)
-
-#define RADEON_CLOCK_CNTL_DATA		0x000c
-#	define RADEON_PLL_WR_EN			(1 << 7)
-#define RADEON_CLOCK_CNTL_INDEX		0x0008
-#define RADEON_CONFIG_APER_SIZE		0x0108
-#define RADEON_CONFIG_MEMSIZE		0x00f8
-#define RADEON_CRTC_OFFSET		0x0224
-#define RADEON_CRTC_OFFSET_CNTL		0x0228
-#	define RADEON_CRTC_TILE_EN		(1 << 15)
-#	define RADEON_CRTC_OFFSET_FLIP_CNTL	(1 << 16)
-#define RADEON_CRTC2_OFFSET		0x0324
-#define RADEON_CRTC2_OFFSET_CNTL	0x0328
-
-#define RADEON_PCIE_INDEX               0x0030
-#define RADEON_PCIE_DATA                0x0034
-#define RADEON_PCIE_TX_GART_CNTL	0x10
-#	define RADEON_PCIE_TX_GART_EN		(1 << 0)
-#	define RADEON_PCIE_TX_GART_UNMAPPED_ACCESS_PASS_THRU (0 << 1)
-#	define RADEON_PCIE_TX_GART_UNMAPPED_ACCESS_CLAMP_LO  (1 << 1)
-#	define RADEON_PCIE_TX_GART_UNMAPPED_ACCESS_DISCARD   (3 << 1)
-#	define RADEON_PCIE_TX_GART_MODE_32_128_CACHE	(0 << 3)
-#	define RADEON_PCIE_TX_GART_MODE_8_4_128_CACHE	(1 << 3)
-#	define RADEON_PCIE_TX_GART_CHK_RW_VALID_EN      (1 << 5)
-#	define RADEON_PCIE_TX_GART_INVALIDATE_TLB	(1 << 8)
-#define RADEON_PCIE_TX_DISCARD_RD_ADDR_LO 0x11
-#define RADEON_PCIE_TX_DISCARD_RD_ADDR_HI 0x12
-#define RADEON_PCIE_TX_GART_BASE	0x13
-#define RADEON_PCIE_TX_GART_START_LO	0x14
-#define RADEON_PCIE_TX_GART_START_HI	0x15
-#define RADEON_PCIE_TX_GART_END_LO	0x16
-#define RADEON_PCIE_TX_GART_END_HI	0x17
-
-#define RS480_NB_MC_INDEX               0x168
-#	define RS480_NB_MC_IND_WR_EN	(1 << 8)
-#define RS480_NB_MC_DATA                0x16c
-
-#define RS690_MC_INDEX                  0x78
-#   define RS690_MC_INDEX_MASK          0x1ff
-#   define RS690_MC_INDEX_WR_EN         (1 << 9)
-#   define RS690_MC_INDEX_WR_ACK        0x7f
-#define RS690_MC_DATA                   0x7c
-
-/* MC indirect registers */
-#define RS480_MC_MISC_CNTL              0x18
-#	define RS480_DISABLE_GTW	(1 << 1)
-/* switch between MCIND GART and MM GART registers. 0 = mmgart, 1 = mcind gart */
-#	define RS480_GART_INDEX_REG_EN	(1 << 12)
-#	define RS690_BLOCK_GFX_D3_EN	(1 << 14)
-#define RS480_K8_FB_LOCATION            0x1e
-#define RS480_GART_FEATURE_ID           0x2b
-#	define RS480_HANG_EN	        (1 << 11)
-#	define RS480_TLB_ENABLE	        (1 << 18)
-#	define RS480_P2P_ENABLE	        (1 << 19)
-#	define RS480_GTW_LAC_EN	        (1 << 25)
-#	define RS480_2LEVEL_GART	(0 << 30)
-#	define RS480_1LEVEL_GART	(1 << 30)
-#	define RS480_PDC_EN	        (1 << 31)
-#define RS480_GART_BASE                 0x2c
-#define RS480_GART_CACHE_CNTRL          0x2e
-#	define RS480_GART_CACHE_INVALIDATE (1 << 0) /* wait for it to clear */
-#define RS480_AGP_ADDRESS_SPACE_SIZE    0x38
-#	define RS480_GART_EN	        (1 << 0)
-#	define RS480_VA_SIZE_32MB	(0 << 1)
-#	define RS480_VA_SIZE_64MB	(1 << 1)
-#	define RS480_VA_SIZE_128MB	(2 << 1)
-#	define RS480_VA_SIZE_256MB	(3 << 1)
-#	define RS480_VA_SIZE_512MB	(4 << 1)
-#	define RS480_VA_SIZE_1GB	(5 << 1)
-#	define RS480_VA_SIZE_2GB	(6 << 1)
-#define RS480_AGP_MODE_CNTL             0x39
-#	define RS480_POST_GART_Q_SIZE	(1 << 18)
-#	define RS480_NONGART_SNOOP	(1 << 19)
-#	define RS480_AGP_RD_BUF_SIZE	(1 << 20)
-#	define RS480_REQ_TYPE_SNOOP_SHIFT 22
-#	define RS480_REQ_TYPE_SNOOP_MASK  0x3
-#	define RS480_REQ_TYPE_SNOOP_DIS	(1 << 24)
-#define RS480_MC_MISC_UMA_CNTL          0x5f
-#define RS480_MC_MCLK_CNTL              0x7a
-#define RS480_MC_UMA_DUALCH_CNTL        0x86
-
-#define RS690_MC_FB_LOCATION            0x100
-#define RS690_MC_AGP_LOCATION           0x101
-#define RS690_MC_AGP_BASE               0x102
-#define RS690_MC_AGP_BASE_2             0x103
-
-#define RS600_MC_INDEX                          0x70
-#       define RS600_MC_ADDR_MASK               0xffff
-#       define RS600_MC_IND_SEQ_RBS_0           (1 << 16)
-#       define RS600_MC_IND_SEQ_RBS_1           (1 << 17)
-#       define RS600_MC_IND_SEQ_RBS_2           (1 << 18)
-#       define RS600_MC_IND_SEQ_RBS_3           (1 << 19)
-#       define RS600_MC_IND_AIC_RBS             (1 << 20)
-#       define RS600_MC_IND_CITF_ARB0           (1 << 21)
-#       define RS600_MC_IND_CITF_ARB1           (1 << 22)
-#       define RS600_MC_IND_WR_EN               (1 << 23)
-#define RS600_MC_DATA                           0x74
-
-#define RS600_MC_STATUS                         0x0
-#       define RS600_MC_IDLE                    (1 << 1)
-#define RS600_MC_FB_LOCATION                    0x4
-#define RS600_MC_AGP_LOCATION                   0x5
-#define RS600_AGP_BASE                          0x6
-#define RS600_AGP_BASE_2                        0x7
-#define RS600_MC_CNTL1                          0x9
-#       define RS600_ENABLE_PAGE_TABLES         (1 << 26)
-#define RS600_MC_PT0_CNTL                       0x100
-#       define RS600_ENABLE_PT                  (1 << 0)
-#       define RS600_EFFECTIVE_L2_CACHE_SIZE(x) ((x) << 15)
-#       define RS600_EFFECTIVE_L2_QUEUE_SIZE(x) ((x) << 21)
-#       define RS600_INVALIDATE_ALL_L1_TLBS     (1 << 28)
-#       define RS600_INVALIDATE_L2_CACHE        (1 << 29)
-#define RS600_MC_PT0_CONTEXT0_CNTL              0x102
-#       define RS600_ENABLE_PAGE_TABLE          (1 << 0)
-#       define RS600_PAGE_TABLE_TYPE_FLAT       (0 << 1)
-#define RS600_MC_PT0_SYSTEM_APERTURE_LOW_ADDR   0x112
-#define RS600_MC_PT0_SYSTEM_APERTURE_HIGH_ADDR  0x114
-#define RS600_MC_PT0_CONTEXT0_DEFAULT_READ_ADDR 0x11c
-#define RS600_MC_PT0_CONTEXT0_FLAT_BASE_ADDR    0x12c
-#define RS600_MC_PT0_CONTEXT0_FLAT_START_ADDR   0x13c
-#define RS600_MC_PT0_CONTEXT0_FLAT_END_ADDR     0x14c
-#define RS600_MC_PT0_CLIENT0_CNTL               0x16c
-#       define RS600_ENABLE_TRANSLATION_MODE_OVERRIDE       (1 << 0)
-#       define RS600_TRANSLATION_MODE_OVERRIDE              (1 << 1)
-#       define RS600_SYSTEM_ACCESS_MODE_MASK                (3 << 8)
-#       define RS600_SYSTEM_ACCESS_MODE_PA_ONLY             (0 << 8)
-#       define RS600_SYSTEM_ACCESS_MODE_USE_SYS_MAP         (1 << 8)
-#       define RS600_SYSTEM_ACCESS_MODE_IN_SYS              (2 << 8)
-#       define RS600_SYSTEM_ACCESS_MODE_NOT_IN_SYS          (3 << 8)
-#       define RS600_SYSTEM_APERTURE_UNMAPPED_ACCESS_PASSTHROUGH        (0 << 10)
-#       define RS600_SYSTEM_APERTURE_UNMAPPED_ACCESS_DEFAULT_PAGE       (1 << 10)
-#       define RS600_EFFECTIVE_L1_CACHE_SIZE(x) ((x) << 11)
-#       define RS600_ENABLE_FRAGMENT_PROCESSING (1 << 14)
-#       define RS600_EFFECTIVE_L1_QUEUE_SIZE(x) ((x) << 15)
-#       define RS600_INVALIDATE_L1_TLB          (1 << 20)
-
-#define R520_MC_IND_INDEX 0x70
-#define R520_MC_IND_WR_EN (1 << 24)
-#define R520_MC_IND_DATA  0x74
-
-#define RV515_MC_FB_LOCATION 0x01
-#define RV515_MC_AGP_LOCATION 0x02
-#define RV515_MC_AGP_BASE     0x03
-#define RV515_MC_AGP_BASE_2   0x04
-
-#define R520_MC_FB_LOCATION 0x04
-#define R520_MC_AGP_LOCATION 0x05
-#define R520_MC_AGP_BASE     0x06
-#define R520_MC_AGP_BASE_2   0x07
-
-#define RADEON_MPP_TB_CONFIG		0x01c0
-#define RADEON_MEM_CNTL			0x0140
-#define RADEON_MEM_SDRAM_MODE_REG	0x0158
-#define RADEON_AGP_BASE_2		0x015c /* r200+ only */
-#define RS480_AGP_BASE_2		0x0164
-#define RADEON_AGP_BASE			0x0170
-
-/* pipe config regs */
-#define R400_GB_PIPE_SELECT             0x402c
-#define RV530_GB_PIPE_SELECT2           0x4124
-#define R500_DYN_SCLK_PWMEM_PIPE        0x000d /* PLL */
-#define R300_GB_TILE_CONFIG             0x4018
-#       define R300_ENABLE_TILING       (1 << 0)
-#       define R300_PIPE_COUNT_RV350    (0 << 1)
-#       define R300_PIPE_COUNT_R300     (3 << 1)
-#       define R300_PIPE_COUNT_R420_3P  (6 << 1)
-#       define R300_PIPE_COUNT_R420     (7 << 1)
-#       define R300_TILE_SIZE_8         (0 << 4)
-#       define R300_TILE_SIZE_16        (1 << 4)
-#       define R300_TILE_SIZE_32        (2 << 4)
-#       define R300_SUBPIXEL_1_12       (0 << 16)
-#       define R300_SUBPIXEL_1_16       (1 << 16)
-#define R300_DST_PIPE_CONFIG            0x170c
-#       define R300_PIPE_AUTO_CONFIG    (1 << 31)
-#define R300_RB2D_DSTCACHE_MODE         0x3428
-#       define R300_DC_AUTOFLUSH_ENABLE (1 << 8)
-#       define R300_DC_DC_DISABLE_IGNORE_PE (1 << 17)
-
-#define RADEON_RB3D_COLOROFFSET		0x1c40
-#define RADEON_RB3D_COLORPITCH		0x1c48
-
-#define	RADEON_SRC_X_Y			0x1590
-
-#define RADEON_DP_GUI_MASTER_CNTL	0x146c
-#	define RADEON_GMC_SRC_PITCH_OFFSET_CNTL	(1 << 0)
-#	define RADEON_GMC_DST_PITCH_OFFSET_CNTL	(1 << 1)
-#	define RADEON_GMC_BRUSH_SOLID_COLOR	(13 << 4)
-#	define RADEON_GMC_BRUSH_NONE		(15 << 4)
-#	define RADEON_GMC_DST_16BPP		(4 << 8)
-#	define RADEON_GMC_DST_24BPP		(5 << 8)
-#	define RADEON_GMC_DST_32BPP		(6 << 8)
-#	define RADEON_GMC_DST_DATATYPE_SHIFT	8
-#	define RADEON_GMC_SRC_DATATYPE_COLOR	(3 << 12)
-#	define RADEON_DP_SRC_SOURCE_MEMORY	(2 << 24)
-#	define RADEON_DP_SRC_SOURCE_HOST_DATA	(3 << 24)
-#	define RADEON_GMC_CLR_CMP_CNTL_DIS	(1 << 28)
-#	define RADEON_GMC_WR_MSK_DIS		(1 << 30)
-#	define RADEON_ROP3_S			0x00cc0000
-#	define RADEON_ROP3_P			0x00f00000
-#define RADEON_DP_WRITE_MASK		0x16cc
-#define RADEON_SRC_PITCH_OFFSET		0x1428
-#define RADEON_DST_PITCH_OFFSET		0x142c
-#define RADEON_DST_PITCH_OFFSET_C	0x1c80
-#	define RADEON_DST_TILE_LINEAR		(0 << 30)
-#	define RADEON_DST_TILE_MACRO		(1 << 30)
-#	define RADEON_DST_TILE_MICRO		(2 << 30)
-#	define RADEON_DST_TILE_BOTH		(3 << 30)
-
-#define RADEON_SCRATCH_REG0		0x15e0
-#define RADEON_SCRATCH_REG1		0x15e4
-#define RADEON_SCRATCH_REG2		0x15e8
-#define RADEON_SCRATCH_REG3		0x15ec
-#define RADEON_SCRATCH_REG4		0x15f0
-#define RADEON_SCRATCH_REG5		0x15f4
-#define RADEON_SCRATCH_UMSK		0x0770
-#define RADEON_SCRATCH_ADDR		0x0774
-
-#define RADEON_SCRATCHOFF( x )		(RADEON_SCRATCH_REG_OFFSET + 4*(x))
-
-extern u32 radeon_get_scratch(drm_radeon_private_t *dev_priv, int index);
-
-#define GET_SCRATCH(dev_priv, x) radeon_get_scratch(dev_priv, x)
-
-#define R600_SCRATCH_REG0		0x8500
-#define R600_SCRATCH_REG1		0x8504
-#define R600_SCRATCH_REG2		0x8508
-#define R600_SCRATCH_REG3		0x850c
-#define R600_SCRATCH_REG4		0x8510
-#define R600_SCRATCH_REG5		0x8514
-#define R600_SCRATCH_REG6		0x8518
-#define R600_SCRATCH_REG7		0x851c
-#define R600_SCRATCH_UMSK		0x8540
-#define R600_SCRATCH_ADDR		0x8544
-
-#define R600_SCRATCHOFF(x)		(R600_SCRATCH_REG_OFFSET + 4*(x))
-
-#define RADEON_GEN_INT_CNTL		0x0040
-#	define RADEON_CRTC_VBLANK_MASK		(1 << 0)
-#	define RADEON_CRTC2_VBLANK_MASK		(1 << 9)
-#	define RADEON_GUI_IDLE_INT_ENABLE	(1 << 19)
-#	define RADEON_SW_INT_ENABLE		(1 << 25)
-
-#define RADEON_GEN_INT_STATUS		0x0044
-#	define RADEON_CRTC_VBLANK_STAT		(1 << 0)
-#	define RADEON_CRTC_VBLANK_STAT_ACK	(1 << 0)
-#	define RADEON_CRTC2_VBLANK_STAT		(1 << 9)
-#	define RADEON_CRTC2_VBLANK_STAT_ACK	(1 << 9)
-#	define RADEON_GUI_IDLE_INT_TEST_ACK     (1 << 19)
-#	define RADEON_SW_INT_TEST		(1 << 25)
-#	define RADEON_SW_INT_TEST_ACK		(1 << 25)
-#	define RADEON_SW_INT_FIRE		(1 << 26)
-#       define R500_DISPLAY_INT_STATUS          (1 << 0)
-
-#define RADEON_HOST_PATH_CNTL		0x0130
-#	define RADEON_HDP_SOFT_RESET		(1 << 26)
-#	define RADEON_HDP_WC_TIMEOUT_MASK	(7 << 28)
-#	define RADEON_HDP_WC_TIMEOUT_28BCLK	(7 << 28)
-
-#define RADEON_ISYNC_CNTL		0x1724
-#	define RADEON_ISYNC_ANY2D_IDLE3D	(1 << 0)
-#	define RADEON_ISYNC_ANY3D_IDLE2D	(1 << 1)
-#	define RADEON_ISYNC_TRIG2D_IDLE3D	(1 << 2)
-#	define RADEON_ISYNC_TRIG3D_IDLE2D	(1 << 3)
-#	define RADEON_ISYNC_WAIT_IDLEGUI	(1 << 4)
-#	define RADEON_ISYNC_CPSCRATCH_IDLEGUI	(1 << 5)
-
-#define RADEON_RBBM_GUICNTL		0x172c
-#	define RADEON_HOST_DATA_SWAP_NONE	(0 << 0)
-#	define RADEON_HOST_DATA_SWAP_16BIT	(1 << 0)
-#	define RADEON_HOST_DATA_SWAP_32BIT	(2 << 0)
-#	define RADEON_HOST_DATA_SWAP_HDW	(3 << 0)
-
-#define RADEON_MC_AGP_LOCATION		0x014c
-#define RADEON_MC_FB_LOCATION		0x0148
-#define RADEON_MCLK_CNTL		0x0012
-#	define RADEON_FORCEON_MCLKA		(1 << 16)
-#	define RADEON_FORCEON_MCLKB		(1 << 17)
-#	define RADEON_FORCEON_YCLKA		(1 << 18)
-#	define RADEON_FORCEON_YCLKB		(1 << 19)
-#	define RADEON_FORCEON_MC		(1 << 20)
-#	define RADEON_FORCEON_AIC		(1 << 21)
-
-#define RADEON_PP_BORDER_COLOR_0	0x1d40
-#define RADEON_PP_BORDER_COLOR_1	0x1d44
-#define RADEON_PP_BORDER_COLOR_2	0x1d48
-#define RADEON_PP_CNTL			0x1c38
-#	define RADEON_SCISSOR_ENABLE		(1 <<  1)
-#define RADEON_PP_LUM_MATRIX		0x1d00
-#define RADEON_PP_MISC			0x1c14
-#define RADEON_PP_ROT_MATRIX_0		0x1d58
-#define RADEON_PP_TXFILTER_0		0x1c54
-#define RADEON_PP_TXOFFSET_0		0x1c5c
-#define RADEON_PP_TXFILTER_1		0x1c6c
-#define RADEON_PP_TXFILTER_2		0x1c84
-
-#define R300_RB2D_DSTCACHE_CTLSTAT	0x342c /* use R300_DSTCACHE_CTLSTAT */
-#define R300_DSTCACHE_CTLSTAT		0x1714
-#	define R300_RB2D_DC_FLUSH		(3 << 0)
-#	define R300_RB2D_DC_FREE		(3 << 2)
-#	define R300_RB2D_DC_FLUSH_ALL		0xf
-#	define R300_RB2D_DC_BUSY		(1 << 31)
-#define RADEON_RB3D_CNTL		0x1c3c
-#	define RADEON_ALPHA_BLEND_ENABLE	(1 << 0)
-#	define RADEON_PLANE_MASK_ENABLE		(1 << 1)
-#	define RADEON_DITHER_ENABLE		(1 << 2)
-#	define RADEON_ROUND_ENABLE		(1 << 3)
-#	define RADEON_SCALE_DITHER_ENABLE	(1 << 4)
-#	define RADEON_DITHER_INIT		(1 << 5)
-#	define RADEON_ROP_ENABLE		(1 << 6)
-#	define RADEON_STENCIL_ENABLE		(1 << 7)
-#	define RADEON_Z_ENABLE			(1 << 8)
-#	define RADEON_ZBLOCK16			(1 << 15)
-#define RADEON_RB3D_DEPTHOFFSET		0x1c24
-#define RADEON_RB3D_DEPTHCLEARVALUE	0x3230
-#define RADEON_RB3D_DEPTHPITCH		0x1c28
-#define RADEON_RB3D_PLANEMASK		0x1d84
-#define RADEON_RB3D_STENCILREFMASK	0x1d7c
-#define RADEON_RB3D_ZCACHE_MODE		0x3250
-#define RADEON_RB3D_ZCACHE_CTLSTAT	0x3254
-#	define RADEON_RB3D_ZC_FLUSH		(1 << 0)
-#	define RADEON_RB3D_ZC_FREE		(1 << 2)
-#	define RADEON_RB3D_ZC_FLUSH_ALL		0x5
-#	define RADEON_RB3D_ZC_BUSY		(1 << 31)
-#define R300_ZB_ZCACHE_CTLSTAT                  0x4f18
-#	define R300_ZC_FLUSH		        (1 << 0)
-#	define R300_ZC_FREE		        (1 << 1)
-#	define R300_ZC_BUSY		        (1 << 31)
-#define RADEON_RB3D_DSTCACHE_CTLSTAT	0x325c
-#	define RADEON_RB3D_DC_FLUSH		(3 << 0)
-#	define RADEON_RB3D_DC_FREE		(3 << 2)
-#	define RADEON_RB3D_DC_FLUSH_ALL		0xf
-#	define RADEON_RB3D_DC_BUSY		(1 << 31)
-#define R300_RB3D_DSTCACHE_CTLSTAT              0x4e4c
-#	define R300_RB3D_DC_FLUSH		(2 << 0)
-#	define R300_RB3D_DC_FREE		(2 << 2)
-#	define R300_RB3D_DC_FINISH		(1 << 4)
-#define RADEON_RB3D_ZSTENCILCNTL	0x1c2c
-#	define RADEON_Z_TEST_MASK		(7 << 4)
-#	define RADEON_Z_TEST_ALWAYS		(7 << 4)
-#	define RADEON_Z_HIERARCHY_ENABLE	(1 << 8)
-#	define RADEON_STENCIL_TEST_ALWAYS	(7 << 12)
-#	define RADEON_STENCIL_S_FAIL_REPLACE	(2 << 16)
-#	define RADEON_STENCIL_ZPASS_REPLACE	(2 << 20)
-#	define RADEON_STENCIL_ZFAIL_REPLACE	(2 << 24)
-#	define RADEON_Z_COMPRESSION_ENABLE	(1 << 28)
-#	define RADEON_FORCE_Z_DIRTY		(1 << 29)
-#	define RADEON_Z_WRITE_ENABLE		(1 << 30)
-#	define RADEON_Z_DECOMPRESSION_ENABLE	(1 << 31)
-#define RADEON_RBBM_SOFT_RESET		0x00f0
-#	define RADEON_SOFT_RESET_CP		(1 <<  0)
-#	define RADEON_SOFT_RESET_HI		(1 <<  1)
-#	define RADEON_SOFT_RESET_SE		(1 <<  2)
-#	define RADEON_SOFT_RESET_RE		(1 <<  3)
-#	define RADEON_SOFT_RESET_PP		(1 <<  4)
-#	define RADEON_SOFT_RESET_E2		(1 <<  5)
-#	define RADEON_SOFT_RESET_RB		(1 <<  6)
-#	define RADEON_SOFT_RESET_HDP		(1 <<  7)
-/*
- *   6:0  Available slots in the FIFO
- *   8    Host Interface active
- *   9    CP request active
- *   10   FIFO request active
- *   11   Host Interface retry active
- *   12   CP retry active
- *   13   FIFO retry active
- *   14   FIFO pipeline busy
- *   15   Event engine busy
- *   16   CP command stream busy
- *   17   2D engine busy
- *   18   2D portion of render backend busy
- *   20   3D setup engine busy
- *   26   GA engine busy
- *   27   CBA 2D engine busy
- *   31   2D engine busy or 3D engine busy or FIFO not empty or CP busy or
- *           command stream queue not empty or Ring Buffer not empty
- */
-#define RADEON_RBBM_STATUS		0x0e40
-/* Same as the previous RADEON_RBBM_STATUS; this is a mirror of that register.  */
-/* #define RADEON_RBBM_STATUS		0x1740 */
-/* bits 6:0 are dword slots available in the cmd fifo */
-#	define RADEON_RBBM_FIFOCNT_MASK		0x007f
-#	define RADEON_HIRQ_ON_RBB	(1 <<  8)
-#	define RADEON_CPRQ_ON_RBB	(1 <<  9)
-#	define RADEON_CFRQ_ON_RBB	(1 << 10)
-#	define RADEON_HIRQ_IN_RTBUF	(1 << 11)
-#	define RADEON_CPRQ_IN_RTBUF	(1 << 12)
-#	define RADEON_CFRQ_IN_RTBUF	(1 << 13)
-#	define RADEON_PIPE_BUSY		(1 << 14)
-#	define RADEON_ENG_EV_BUSY	(1 << 15)
-#	define RADEON_CP_CMDSTRM_BUSY	(1 << 16)
-#	define RADEON_E2_BUSY		(1 << 17)
-#	define RADEON_RB2D_BUSY		(1 << 18)
-#	define RADEON_RB3D_BUSY		(1 << 19) /* not used on r300 */
-#	define RADEON_VAP_BUSY		(1 << 20)
-#	define RADEON_RE_BUSY		(1 << 21) /* not used on r300 */
-#	define RADEON_TAM_BUSY		(1 << 22) /* not used on r300 */
-#	define RADEON_TDM_BUSY		(1 << 23) /* not used on r300 */
-#	define RADEON_PB_BUSY		(1 << 24) /* not used on r300 */
-#	define RADEON_TIM_BUSY		(1 << 25) /* not used on r300 */
-#	define RADEON_GA_BUSY		(1 << 26)
-#	define RADEON_CBA2D_BUSY	(1 << 27)
-#	define RADEON_RBBM_ACTIVE	(1 << 31)
-#define RADEON_RE_LINE_PATTERN		0x1cd0
-#define RADEON_RE_MISC			0x26c4
-#define RADEON_RE_TOP_LEFT		0x26c0
-#define RADEON_RE_WIDTH_HEIGHT		0x1c44
-#define RADEON_RE_STIPPLE_ADDR		0x1cc8
-#define RADEON_RE_STIPPLE_DATA		0x1ccc
-
-#define RADEON_SCISSOR_TL_0		0x1cd8
-#define RADEON_SCISSOR_BR_0		0x1cdc
-#define RADEON_SCISSOR_TL_1		0x1ce0
-#define RADEON_SCISSOR_BR_1		0x1ce4
-#define RADEON_SCISSOR_TL_2		0x1ce8
-#define RADEON_SCISSOR_BR_2		0x1cec
-#define RADEON_SE_COORD_FMT		0x1c50
-#define RADEON_SE_CNTL			0x1c4c
-#	define RADEON_FFACE_CULL_CW		(0 << 0)
-#	define RADEON_BFACE_SOLID		(3 << 1)
-#	define RADEON_FFACE_SOLID		(3 << 3)
-#	define RADEON_FLAT_SHADE_VTX_LAST	(3 << 6)
-#	define RADEON_DIFFUSE_SHADE_FLAT	(1 << 8)
-#	define RADEON_DIFFUSE_SHADE_GOURAUD	(2 << 8)
-#	define RADEON_ALPHA_SHADE_FLAT		(1 << 10)
-#	define RADEON_ALPHA_SHADE_GOURAUD	(2 << 10)
-#	define RADEON_SPECULAR_SHADE_FLAT	(1 << 12)
-#	define RADEON_SPECULAR_SHADE_GOURAUD	(2 << 12)
-#	define RADEON_FOG_SHADE_FLAT		(1 << 14)
-#	define RADEON_FOG_SHADE_GOURAUD		(2 << 14)
-#	define RADEON_VPORT_XY_XFORM_ENABLE	(1 << 24)
-#	define RADEON_VPORT_Z_XFORM_ENABLE	(1 << 25)
-#	define RADEON_VTX_PIX_CENTER_OGL	(1 << 27)
-#	define RADEON_ROUND_MODE_TRUNC		(0 << 28)
-#	define RADEON_ROUND_PREC_8TH_PIX	(1 << 30)
-#define RADEON_SE_CNTL_STATUS		0x2140
-#define RADEON_SE_LINE_WIDTH		0x1db8
-#define RADEON_SE_VPORT_XSCALE		0x1d98
-#define RADEON_SE_ZBIAS_FACTOR		0x1db0
-#define RADEON_SE_TCL_MATERIAL_EMMISSIVE_RED 0x2210
-#define RADEON_SE_TCL_OUTPUT_VTX_FMT         0x2254
-#define RADEON_SE_TCL_VECTOR_INDX_REG        0x2200
-#       define RADEON_VEC_INDX_OCTWORD_STRIDE_SHIFT  16
-#       define RADEON_VEC_INDX_DWORD_COUNT_SHIFT     28
-#define RADEON_SE_TCL_VECTOR_DATA_REG       0x2204
-#define RADEON_SE_TCL_SCALAR_INDX_REG       0x2208
-#       define RADEON_SCAL_INDX_DWORD_STRIDE_SHIFT  16
-#define RADEON_SE_TCL_SCALAR_DATA_REG       0x220C
-#define RADEON_SURFACE_ACCESS_FLAGS	0x0bf8
-#define RADEON_SURFACE_ACCESS_CLR	0x0bfc
-#define RADEON_SURFACE_CNTL		0x0b00
-#	define RADEON_SURF_TRANSLATION_DIS	(1 << 8)
-#	define RADEON_NONSURF_AP0_SWP_MASK	(3 << 20)
-#	define RADEON_NONSURF_AP0_SWP_LITTLE	(0 << 20)
-#	define RADEON_NONSURF_AP0_SWP_BIG16	(1 << 20)
-#	define RADEON_NONSURF_AP0_SWP_BIG32	(2 << 20)
-#	define RADEON_NONSURF_AP1_SWP_MASK	(3 << 22)
-#	define RADEON_NONSURF_AP1_SWP_LITTLE	(0 << 22)
-#	define RADEON_NONSURF_AP1_SWP_BIG16	(1 << 22)
-#	define RADEON_NONSURF_AP1_SWP_BIG32	(2 << 22)
-#define RADEON_SURFACE0_INFO		0x0b0c
-#	define RADEON_SURF_PITCHSEL_MASK	(0x1ff << 0)
-#	define RADEON_SURF_TILE_MODE_MASK	(3 << 16)
-#	define RADEON_SURF_TILE_MODE_MACRO	(0 << 16)
-#	define RADEON_SURF_TILE_MODE_MICRO	(1 << 16)
-#	define RADEON_SURF_TILE_MODE_32BIT_Z	(2 << 16)
-#	define RADEON_SURF_TILE_MODE_16BIT_Z	(3 << 16)
-#define RADEON_SURFACE0_LOWER_BOUND	0x0b04
-#define RADEON_SURFACE0_UPPER_BOUND	0x0b08
-#	define RADEON_SURF_ADDRESS_FIXED_MASK	(0x3ff << 0)
-#define RADEON_SURFACE1_INFO		0x0b1c
-#define RADEON_SURFACE1_LOWER_BOUND	0x0b14
-#define RADEON_SURFACE1_UPPER_BOUND	0x0b18
-#define RADEON_SURFACE2_INFO		0x0b2c
-#define RADEON_SURFACE2_LOWER_BOUND	0x0b24
-#define RADEON_SURFACE2_UPPER_BOUND	0x0b28
-#define RADEON_SURFACE3_INFO		0x0b3c
-#define RADEON_SURFACE3_LOWER_BOUND	0x0b34
-#define RADEON_SURFACE3_UPPER_BOUND	0x0b38
-#define RADEON_SURFACE4_INFO		0x0b4c
-#define RADEON_SURFACE4_LOWER_BOUND	0x0b44
-#define RADEON_SURFACE4_UPPER_BOUND	0x0b48
-#define RADEON_SURFACE5_INFO		0x0b5c
-#define RADEON_SURFACE5_LOWER_BOUND	0x0b54
-#define RADEON_SURFACE5_UPPER_BOUND	0x0b58
-#define RADEON_SURFACE6_INFO		0x0b6c
-#define RADEON_SURFACE6_LOWER_BOUND	0x0b64
-#define RADEON_SURFACE6_UPPER_BOUND	0x0b68
-#define RADEON_SURFACE7_INFO		0x0b7c
-#define RADEON_SURFACE7_LOWER_BOUND	0x0b74
-#define RADEON_SURFACE7_UPPER_BOUND	0x0b78
-#define RADEON_SW_SEMAPHORE		0x013c
-
-#define RADEON_WAIT_UNTIL		0x1720
-#	define RADEON_WAIT_CRTC_PFLIP		(1 << 0)
-#	define RADEON_WAIT_2D_IDLE		(1 << 14)
-#	define RADEON_WAIT_3D_IDLE		(1 << 15)
-#	define RADEON_WAIT_2D_IDLECLEAN		(1 << 16)
-#	define RADEON_WAIT_3D_IDLECLEAN		(1 << 17)
-#	define RADEON_WAIT_HOST_IDLECLEAN	(1 << 18)
-
-#define RADEON_RB3D_ZMASKOFFSET		0x3234
-#define RADEON_RB3D_ZSTENCILCNTL	0x1c2c
-#	define RADEON_DEPTH_FORMAT_16BIT_INT_Z	(0 << 0)
-#	define RADEON_DEPTH_FORMAT_24BIT_INT_Z	(2 << 0)
-
-/* CP registers */
-#define RADEON_CP_ME_RAM_ADDR		0x07d4
-#define RADEON_CP_ME_RAM_RADDR		0x07d8
-#define RADEON_CP_ME_RAM_DATAH		0x07dc
-#define RADEON_CP_ME_RAM_DATAL		0x07e0
-
-#define RADEON_CP_RB_BASE		0x0700
-#define RADEON_CP_RB_CNTL		0x0704
-#	define RADEON_BUF_SWAP_32BIT		(2 << 16)
-#	define RADEON_RB_NO_UPDATE		(1 << 27)
-#	define RADEON_RB_RPTR_WR_ENA		(1 << 31)
-#define RADEON_CP_RB_RPTR_ADDR		0x070c
-#define RADEON_CP_RB_RPTR		0x0710
-#define RADEON_CP_RB_WPTR		0x0714
-
-#define RADEON_CP_RB_WPTR_DELAY		0x0718
-#	define RADEON_PRE_WRITE_TIMER_SHIFT	0
-#	define RADEON_PRE_WRITE_LIMIT_SHIFT	23
-
-#define RADEON_CP_IB_BASE		0x0738
-
-#define RADEON_CP_CSQ_CNTL		0x0740
-#	define RADEON_CSQ_CNT_PRIMARY_MASK	(0xff << 0)
-#	define RADEON_CSQ_PRIDIS_INDDIS		(0 << 28)
-#	define RADEON_CSQ_PRIPIO_INDDIS		(1 << 28)
-#	define RADEON_CSQ_PRIBM_INDDIS		(2 << 28)
-#	define RADEON_CSQ_PRIPIO_INDBM		(3 << 28)
-#	define RADEON_CSQ_PRIBM_INDBM		(4 << 28)
-#	define RADEON_CSQ_PRIPIO_INDPIO		(15 << 28)
-
-#define R300_CP_RESYNC_ADDR		0x0778
-#define R300_CP_RESYNC_DATA		0x077c
-
-#define RADEON_AIC_CNTL			0x01d0
-#	define RADEON_PCIGART_TRANSLATE_EN	(1 << 0)
-#	define RS400_MSI_REARM	                (1 << 3)
-#define RADEON_AIC_STAT			0x01d4
-#define RADEON_AIC_PT_BASE		0x01d8
-#define RADEON_AIC_LO_ADDR		0x01dc
-#define RADEON_AIC_HI_ADDR		0x01e0
-#define RADEON_AIC_TLB_ADDR		0x01e4
-#define RADEON_AIC_TLB_DATA		0x01e8
-
-/* CP command packets */
-#define RADEON_CP_PACKET0		0x00000000
-#	define RADEON_ONE_REG_WR		(1 << 15)
-#define RADEON_CP_PACKET1		0x40000000
-#define RADEON_CP_PACKET2		0x80000000
-#define RADEON_CP_PACKET3		0xC0000000
-#       define RADEON_CP_NOP                    0x00001000
-#       define RADEON_CP_NEXT_CHAR              0x00001900
-#       define RADEON_CP_PLY_NEXTSCAN           0x00001D00
-#       define RADEON_CP_SET_SCISSORS           0x00001E00
-	     /* GEN_INDX_PRIM is unsupported starting with R300 */
-#	define RADEON_3D_RNDR_GEN_INDX_PRIM	0x00002300
-#	define RADEON_WAIT_FOR_IDLE		0x00002600
-#	define RADEON_3D_DRAW_VBUF		0x00002800
-#	define RADEON_3D_DRAW_IMMD		0x00002900
-#	define RADEON_3D_DRAW_INDX		0x00002A00
-#       define RADEON_CP_LOAD_PALETTE           0x00002C00
-#	define RADEON_3D_LOAD_VBPNTR		0x00002F00
-#	define RADEON_MPEG_IDCT_MACROBLOCK	0x00003000
-#	define RADEON_MPEG_IDCT_MACROBLOCK_REV	0x00003100
-#	define RADEON_3D_CLEAR_ZMASK		0x00003200
-#	define RADEON_CP_INDX_BUFFER		0x00003300
-#       define RADEON_CP_3D_DRAW_VBUF_2         0x00003400
-#       define RADEON_CP_3D_DRAW_IMMD_2         0x00003500
-#       define RADEON_CP_3D_DRAW_INDX_2         0x00003600
-#	define RADEON_3D_CLEAR_HIZ		0x00003700
-#       define RADEON_CP_3D_CLEAR_CMASK         0x00003802
-#	define RADEON_CNTL_HOSTDATA_BLT		0x00009400
-#	define RADEON_CNTL_PAINT_MULTI		0x00009A00
-#	define RADEON_CNTL_BITBLT_MULTI		0x00009B00
-#	define RADEON_CNTL_SET_SCISSORS		0xC0001E00
-
-#       define R600_IT_INDIRECT_BUFFER_END      0x00001700
-#       define R600_IT_SET_PREDICATION          0x00002000
-#       define R600_IT_REG_RMW                  0x00002100
-#       define R600_IT_COND_EXEC                0x00002200
-#       define R600_IT_PRED_EXEC                0x00002300
-#       define R600_IT_START_3D_CMDBUF          0x00002400
-#       define R600_IT_DRAW_INDEX_2             0x00002700
-#       define R600_IT_CONTEXT_CONTROL          0x00002800
-#       define R600_IT_DRAW_INDEX_IMMD_BE       0x00002900
-#       define R600_IT_INDEX_TYPE               0x00002A00
-#       define R600_IT_DRAW_INDEX               0x00002B00
-#       define R600_IT_DRAW_INDEX_AUTO          0x00002D00
-#       define R600_IT_DRAW_INDEX_IMMD          0x00002E00
-#       define R600_IT_NUM_INSTANCES            0x00002F00
-#       define R600_IT_STRMOUT_BUFFER_UPDATE    0x00003400
-#       define R600_IT_INDIRECT_BUFFER_MP       0x00003800
-#       define R600_IT_MEM_SEMAPHORE            0x00003900
-#       define R600_IT_MPEG_INDEX               0x00003A00
-#       define R600_IT_WAIT_REG_MEM             0x00003C00
-#       define R600_IT_MEM_WRITE                0x00003D00
-#       define R600_IT_INDIRECT_BUFFER          0x00003200
-#       define R600_IT_SURFACE_SYNC             0x00004300
-#              define R600_CB0_DEST_BASE_ENA    (1 << 6)
-#              define R600_TC_ACTION_ENA        (1 << 23)
-#              define R600_VC_ACTION_ENA        (1 << 24)
-#              define R600_CB_ACTION_ENA        (1 << 25)
-#              define R600_DB_ACTION_ENA        (1 << 26)
-#              define R600_SH_ACTION_ENA        (1 << 27)
-#              define R600_SMX_ACTION_ENA       (1 << 28)
-#       define R600_IT_ME_INITIALIZE            0x00004400
-#	       define R600_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16)
-#       define R600_IT_COND_WRITE               0x00004500
-#       define R600_IT_EVENT_WRITE              0x00004600
-#       define R600_IT_EVENT_WRITE_EOP          0x00004700
-#       define R600_IT_ONE_REG_WRITE            0x00005700
-#       define R600_IT_SET_CONFIG_REG           0x00006800
-#              define R600_SET_CONFIG_REG_OFFSET 0x00008000
-#              define R600_SET_CONFIG_REG_END   0x0000ac00
-#       define R600_IT_SET_CONTEXT_REG          0x00006900
-#              define R600_SET_CONTEXT_REG_OFFSET 0x00028000
-#              define R600_SET_CONTEXT_REG_END  0x00029000
-#       define R600_IT_SET_ALU_CONST            0x00006A00
-#              define R600_SET_ALU_CONST_OFFSET 0x00030000
-#              define R600_SET_ALU_CONST_END    0x00032000
-#       define R600_IT_SET_BOOL_CONST           0x00006B00
-#              define R600_SET_BOOL_CONST_OFFSET 0x0003e380
-#              define R600_SET_BOOL_CONST_END   0x00040000
-#       define R600_IT_SET_LOOP_CONST           0x00006C00
-#              define R600_SET_LOOP_CONST_OFFSET 0x0003e200
-#              define R600_SET_LOOP_CONST_END   0x0003e380
-#       define R600_IT_SET_RESOURCE             0x00006D00
-#              define R600_SET_RESOURCE_OFFSET  0x00038000
-#              define R600_SET_RESOURCE_END     0x0003c000
-#              define R600_SQ_TEX_VTX_INVALID_TEXTURE  0x0
-#              define R600_SQ_TEX_VTX_INVALID_BUFFER   0x1
-#              define R600_SQ_TEX_VTX_VALID_TEXTURE    0x2
-#              define R600_SQ_TEX_VTX_VALID_BUFFER     0x3
-#       define R600_IT_SET_SAMPLER              0x00006E00
-#              define R600_SET_SAMPLER_OFFSET   0x0003c000
-#              define R600_SET_SAMPLER_END      0x0003cff0
-#       define R600_IT_SET_CTL_CONST            0x00006F00
-#              define R600_SET_CTL_CONST_OFFSET 0x0003cff0
-#              define R600_SET_CTL_CONST_END    0x0003e200
-#       define R600_IT_SURFACE_BASE_UPDATE      0x00007300
-
-#define RADEON_CP_PACKET_MASK		0xC0000000
-#define RADEON_CP_PACKET_COUNT_MASK	0x3fff0000
-#define RADEON_CP_PACKET0_REG_MASK	0x000007ff
-#define RADEON_CP_PACKET1_REG0_MASK	0x000007ff
-#define RADEON_CP_PACKET1_REG1_MASK	0x003ff800
-
-#define RADEON_VTX_Z_PRESENT			(1 << 31)
-#define RADEON_VTX_PKCOLOR_PRESENT		(1 << 3)
-
-#define RADEON_PRIM_TYPE_NONE			(0 << 0)
-#define RADEON_PRIM_TYPE_POINT			(1 << 0)
-#define RADEON_PRIM_TYPE_LINE			(2 << 0)
-#define RADEON_PRIM_TYPE_LINE_STRIP		(3 << 0)
-#define RADEON_PRIM_TYPE_TRI_LIST		(4 << 0)
-#define RADEON_PRIM_TYPE_TRI_FAN		(5 << 0)
-#define RADEON_PRIM_TYPE_TRI_STRIP		(6 << 0)
-#define RADEON_PRIM_TYPE_TRI_TYPE2		(7 << 0)
-#define RADEON_PRIM_TYPE_RECT_LIST		(8 << 0)
-#define RADEON_PRIM_TYPE_3VRT_POINT_LIST	(9 << 0)
-#define RADEON_PRIM_TYPE_3VRT_LINE_LIST		(10 << 0)
-#define RADEON_PRIM_TYPE_MASK                   0xf
-#define RADEON_PRIM_WALK_IND			(1 << 4)
-#define RADEON_PRIM_WALK_LIST			(2 << 4)
-#define RADEON_PRIM_WALK_RING			(3 << 4)
-#define RADEON_COLOR_ORDER_BGRA			(0 << 6)
-#define RADEON_COLOR_ORDER_RGBA			(1 << 6)
-#define RADEON_MAOS_ENABLE			(1 << 7)
-#define RADEON_VTX_FMT_R128_MODE		(0 << 8)
-#define RADEON_VTX_FMT_RADEON_MODE		(1 << 8)
-#define RADEON_NUM_VERTICES_SHIFT		16
-
-#define RADEON_COLOR_FORMAT_CI8		2
-#define RADEON_COLOR_FORMAT_ARGB1555	3
-#define RADEON_COLOR_FORMAT_RGB565	4
-#define RADEON_COLOR_FORMAT_ARGB8888	6
-#define RADEON_COLOR_FORMAT_RGB332	7
-#define RADEON_COLOR_FORMAT_RGB8	9
-#define RADEON_COLOR_FORMAT_ARGB4444	15
-
-#define RADEON_TXFORMAT_I8		0
-#define RADEON_TXFORMAT_AI88		1
-#define RADEON_TXFORMAT_RGB332		2
-#define RADEON_TXFORMAT_ARGB1555	3
-#define RADEON_TXFORMAT_RGB565		4
-#define RADEON_TXFORMAT_ARGB4444	5
-#define RADEON_TXFORMAT_ARGB8888	6
-#define RADEON_TXFORMAT_RGBA8888	7
-#define RADEON_TXFORMAT_Y8		8
-#define RADEON_TXFORMAT_VYUY422         10
-#define RADEON_TXFORMAT_YVYU422         11
-#define RADEON_TXFORMAT_DXT1            12
-#define RADEON_TXFORMAT_DXT23           14
-#define RADEON_TXFORMAT_DXT45           15
-
-#define R200_PP_TXCBLEND_0                0x2f00
-#define R200_PP_TXCBLEND_1                0x2f10
-#define R200_PP_TXCBLEND_2                0x2f20
-#define R200_PP_TXCBLEND_3                0x2f30
-#define R200_PP_TXCBLEND_4                0x2f40
-#define R200_PP_TXCBLEND_5                0x2f50
-#define R200_PP_TXCBLEND_6                0x2f60
-#define R200_PP_TXCBLEND_7                0x2f70
-#define R200_SE_TCL_LIGHT_MODEL_CTL_0     0x2268
-#define R200_PP_TFACTOR_0                 0x2ee0
-#define R200_SE_VTX_FMT_0                 0x2088
-#define R200_SE_VAP_CNTL                  0x2080
-#define R200_SE_TCL_MATRIX_SEL_0          0x2230
-#define R200_SE_TCL_TEX_PROC_CTL_2        0x22a8
-#define R200_SE_TCL_UCP_VERT_BLEND_CTL    0x22c0
-#define R200_PP_TXFILTER_5                0x2ca0
-#define R200_PP_TXFILTER_4                0x2c80
-#define R200_PP_TXFILTER_3                0x2c60
-#define R200_PP_TXFILTER_2                0x2c40
-#define R200_PP_TXFILTER_1                0x2c20
-#define R200_PP_TXFILTER_0                0x2c00
-#define R200_PP_TXOFFSET_5                0x2d78
-#define R200_PP_TXOFFSET_4                0x2d60
-#define R200_PP_TXOFFSET_3                0x2d48
-#define R200_PP_TXOFFSET_2                0x2d30
-#define R200_PP_TXOFFSET_1                0x2d18
-#define R200_PP_TXOFFSET_0                0x2d00
-
-#define R200_PP_CUBIC_FACES_0             0x2c18
-#define R200_PP_CUBIC_FACES_1             0x2c38
-#define R200_PP_CUBIC_FACES_2             0x2c58
-#define R200_PP_CUBIC_FACES_3             0x2c78
-#define R200_PP_CUBIC_FACES_4             0x2c98
-#define R200_PP_CUBIC_FACES_5             0x2cb8
-#define R200_PP_CUBIC_OFFSET_F1_0         0x2d04
-#define R200_PP_CUBIC_OFFSET_F2_0         0x2d08
-#define R200_PP_CUBIC_OFFSET_F3_0         0x2d0c
-#define R200_PP_CUBIC_OFFSET_F4_0         0x2d10
-#define R200_PP_CUBIC_OFFSET_F5_0         0x2d14
-#define R200_PP_CUBIC_OFFSET_F1_1         0x2d1c
-#define R200_PP_CUBIC_OFFSET_F2_1         0x2d20
-#define R200_PP_CUBIC_OFFSET_F3_1         0x2d24
-#define R200_PP_CUBIC_OFFSET_F4_1         0x2d28
-#define R200_PP_CUBIC_OFFSET_F5_1         0x2d2c
-#define R200_PP_CUBIC_OFFSET_F1_2         0x2d34
-#define R200_PP_CUBIC_OFFSET_F2_2         0x2d38
-#define R200_PP_CUBIC_OFFSET_F3_2         0x2d3c
-#define R200_PP_CUBIC_OFFSET_F4_2         0x2d40
-#define R200_PP_CUBIC_OFFSET_F5_2         0x2d44
-#define R200_PP_CUBIC_OFFSET_F1_3         0x2d4c
-#define R200_PP_CUBIC_OFFSET_F2_3         0x2d50
-#define R200_PP_CUBIC_OFFSET_F3_3         0x2d54
-#define R200_PP_CUBIC_OFFSET_F4_3         0x2d58
-#define R200_PP_CUBIC_OFFSET_F5_3         0x2d5c
-#define R200_PP_CUBIC_OFFSET_F1_4         0x2d64
-#define R200_PP_CUBIC_OFFSET_F2_4         0x2d68
-#define R200_PP_CUBIC_OFFSET_F3_4         0x2d6c
-#define R200_PP_CUBIC_OFFSET_F4_4         0x2d70
-#define R200_PP_CUBIC_OFFSET_F5_4         0x2d74
-#define R200_PP_CUBIC_OFFSET_F1_5         0x2d7c
-#define R200_PP_CUBIC_OFFSET_F2_5         0x2d80
-#define R200_PP_CUBIC_OFFSET_F3_5         0x2d84
-#define R200_PP_CUBIC_OFFSET_F4_5         0x2d88
-#define R200_PP_CUBIC_OFFSET_F5_5         0x2d8c
-
-#define R200_RE_AUX_SCISSOR_CNTL          0x26f0
-#define R200_SE_VTE_CNTL                  0x20b0
-#define R200_SE_TCL_OUTPUT_VTX_COMP_SEL   0x2250
-#define R200_PP_TAM_DEBUG3                0x2d9c
-#define R200_PP_CNTL_X                    0x2cc4
-#define R200_SE_VAP_CNTL_STATUS           0x2140
-#define R200_RE_SCISSOR_TL_0              0x1cd8
-#define R200_RE_SCISSOR_TL_1              0x1ce0
-#define R200_RE_SCISSOR_TL_2              0x1ce8
-#define R200_RB3D_DEPTHXY_OFFSET          0x1d60
-#define R200_RE_AUX_SCISSOR_CNTL          0x26f0
-#define R200_SE_VTX_STATE_CNTL            0x2180
-#define R200_RE_POINTSIZE                 0x2648
-#define R200_SE_TCL_INPUT_VTX_VECTOR_ADDR_0 0x2254
-
-#define RADEON_PP_TEX_SIZE_0                0x1d04	/* NPOT */
-#define RADEON_PP_TEX_SIZE_1                0x1d0c
-#define RADEON_PP_TEX_SIZE_2                0x1d14
-
-#define RADEON_PP_CUBIC_FACES_0             0x1d24
-#define RADEON_PP_CUBIC_FACES_1             0x1d28
-#define RADEON_PP_CUBIC_FACES_2             0x1d2c
-#define RADEON_PP_CUBIC_OFFSET_T0_0         0x1dd0	/* bits [31:5] */
-#define RADEON_PP_CUBIC_OFFSET_T1_0         0x1e00
-#define RADEON_PP_CUBIC_OFFSET_T2_0         0x1e14
-
-#define RADEON_SE_TCL_STATE_FLUSH           0x2284
-
-#define SE_VAP_CNTL__TCL_ENA_MASK                          0x00000001
-#define SE_VAP_CNTL__FORCE_W_TO_ONE_MASK                   0x00010000
-#define SE_VAP_CNTL__VF_MAX_VTX_NUM__SHIFT                 0x00000012
-#define SE_VTE_CNTL__VTX_XY_FMT_MASK                       0x00000100
-#define SE_VTE_CNTL__VTX_Z_FMT_MASK                        0x00000200
-#define SE_VTX_FMT_0__VTX_Z0_PRESENT_MASK                  0x00000001
-#define SE_VTX_FMT_0__VTX_W0_PRESENT_MASK                  0x00000002
-#define SE_VTX_FMT_0__VTX_COLOR_0_FMT__SHIFT               0x0000000b
-#define R200_3D_DRAW_IMMD_2      0xC0003500
-#define R200_SE_VTX_FMT_1                 0x208c
-#define R200_RE_CNTL                      0x1c50
-
-#define R200_RB3D_BLENDCOLOR              0x3218
-
-#define R200_SE_TCL_POINT_SPRITE_CNTL     0x22c4
-
-#define R200_PP_TRI_PERF 0x2cf8
-
-#define R200_PP_AFS_0                     0x2f80
-#define R200_PP_AFS_1                     0x2f00	/* same as txcblend_0 */
-
-#define R200_VAP_PVS_CNTL_1               0x22D0
-
-#define RADEON_CRTC_CRNT_FRAME 0x0214
-#define RADEON_CRTC2_CRNT_FRAME 0x0314
-
-#define R500_D1CRTC_STATUS 0x609c
-#define R500_D2CRTC_STATUS 0x689c
-#define R500_CRTC_V_BLANK (1<<0)
-
-#define R500_D1CRTC_FRAME_COUNT 0x60a4
-#define R500_D2CRTC_FRAME_COUNT 0x68a4
-
-#define R500_D1MODE_V_COUNTER 0x6530
-#define R500_D2MODE_V_COUNTER 0x6d30
-
-#define R500_D1MODE_VBLANK_STATUS 0x6534
-#define R500_D2MODE_VBLANK_STATUS 0x6d34
-#define R500_VBLANK_OCCURED (1<<0)
-#define R500_VBLANK_ACK     (1<<4)
-#define R500_VBLANK_STAT    (1<<12)
-#define R500_VBLANK_INT     (1<<16)
-
-#define R500_DxMODE_INT_MASK 0x6540
-#define R500_D1MODE_INT_MASK (1<<0)
-#define R500_D2MODE_INT_MASK (1<<8)
-
-#define R500_DISP_INTERRUPT_STATUS 0x7edc
-#define R500_D1_VBLANK_INTERRUPT (1 << 4)
-#define R500_D2_VBLANK_INTERRUPT (1 << 5)
-
-/* R6xx/R7xx registers */
-#define R600_MC_VM_FB_LOCATION                                 0x2180
-#define R600_MC_VM_AGP_TOP                                     0x2184
-#define R600_MC_VM_AGP_BOT                                     0x2188
-#define R600_MC_VM_AGP_BASE                                    0x218c
-#define R600_MC_VM_SYSTEM_APERTURE_LOW_ADDR                    0x2190
-#define R600_MC_VM_SYSTEM_APERTURE_HIGH_ADDR                   0x2194
-#define R600_MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR                0x2198
-
-#define R700_MC_VM_FB_LOCATION                                 0x2024
-#define R700_MC_VM_AGP_TOP                                     0x2028
-#define R700_MC_VM_AGP_BOT                                     0x202c
-#define R700_MC_VM_AGP_BASE                                    0x2030
-#define R700_MC_VM_SYSTEM_APERTURE_LOW_ADDR                    0x2034
-#define R700_MC_VM_SYSTEM_APERTURE_HIGH_ADDR                   0x2038
-#define R700_MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR                0x203c
-
-#define R600_MCD_RD_A_CNTL                                     0x219c
-#define R600_MCD_RD_B_CNTL                                     0x21a0
-
-#define R600_MCD_WR_A_CNTL                                     0x21a4
-#define R600_MCD_WR_B_CNTL                                     0x21a8
-
-#define R600_MCD_RD_SYS_CNTL                                   0x2200
-#define R600_MCD_WR_SYS_CNTL                                   0x2214
-
-#define R600_MCD_RD_GFX_CNTL                                   0x21fc
-#define R600_MCD_RD_HDP_CNTL                                   0x2204
-#define R600_MCD_RD_PDMA_CNTL                                  0x2208
-#define R600_MCD_RD_SEM_CNTL                                   0x220c
-#define R600_MCD_WR_GFX_CNTL                                   0x2210
-#define R600_MCD_WR_HDP_CNTL                                   0x2218
-#define R600_MCD_WR_PDMA_CNTL                                  0x221c
-#define R600_MCD_WR_SEM_CNTL                                   0x2220
-
-#       define R600_MCD_L1_TLB                                 (1 << 0)
-#       define R600_MCD_L1_FRAG_PROC                           (1 << 1)
-#       define R600_MCD_L1_STRICT_ORDERING                     (1 << 2)
-
-#       define R600_MCD_SYSTEM_ACCESS_MODE_MASK                (3 << 6)
-#       define R600_MCD_SYSTEM_ACCESS_MODE_PA_ONLY             (0 << 6)
-#       define R600_MCD_SYSTEM_ACCESS_MODE_USE_SYS_MAP         (1 << 6)
-#       define R600_MCD_SYSTEM_ACCESS_MODE_IN_SYS              (2 << 6)
-#       define R600_MCD_SYSTEM_ACCESS_MODE_NOT_IN_SYS          (3 << 6)
-
-#       define R600_MCD_SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU    (0 << 8)
-#       define R600_MCD_SYSTEM_APERTURE_UNMAPPED_ACCESS_DEFAULT_PAGE (1 << 8)
-
-#       define R600_MCD_SEMAPHORE_MODE                         (1 << 10)
-#       define R600_MCD_WAIT_L2_QUERY                          (1 << 11)
-#       define R600_MCD_EFFECTIVE_L1_TLB_SIZE(x)               ((x) << 12)
-#       define R600_MCD_EFFECTIVE_L1_QUEUE_SIZE(x)             ((x) << 15)
-
-#define R700_MC_VM_MD_L1_TLB0_CNTL                             0x2654
-#define R700_MC_VM_MD_L1_TLB1_CNTL                             0x2658
-#define R700_MC_VM_MD_L1_TLB2_CNTL                             0x265c
-
-#define R700_MC_VM_MB_L1_TLB0_CNTL                             0x2234
-#define R700_MC_VM_MB_L1_TLB1_CNTL                             0x2238
-#define R700_MC_VM_MB_L1_TLB2_CNTL                             0x223c
-#define R700_MC_VM_MB_L1_TLB3_CNTL                             0x2240
-
-#       define R700_ENABLE_L1_TLB                              (1 << 0)
-#       define R700_ENABLE_L1_FRAGMENT_PROCESSING              (1 << 1)
-#       define R700_SYSTEM_ACCESS_MODE_IN_SYS                  (2 << 3)
-#       define R700_SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU  (0 << 5)
-#       define R700_EFFECTIVE_L1_TLB_SIZE(x)                   ((x) << 15)
-#       define R700_EFFECTIVE_L1_QUEUE_SIZE(x)                 ((x) << 18)
-
-#define R700_MC_ARB_RAMCFG                                     0x2760
-#       define R700_NOOFBANK_SHIFT                             0
-#       define R700_NOOFBANK_MASK                              0x3
-#       define R700_NOOFRANK_SHIFT                             2
-#       define R700_NOOFRANK_MASK                              0x1
-#       define R700_NOOFROWS_SHIFT                             3
-#       define R700_NOOFROWS_MASK                              0x7
-#       define R700_NOOFCOLS_SHIFT                             6
-#       define R700_NOOFCOLS_MASK                              0x3
-#       define R700_CHANSIZE_SHIFT                             8
-#       define R700_CHANSIZE_MASK                              0x1
-#       define R700_BURSTLENGTH_SHIFT                          9
-#       define R700_BURSTLENGTH_MASK                           0x1
-#define R600_RAMCFG                                            0x2408
-#       define R600_NOOFBANK_SHIFT                             0
-#       define R600_NOOFBANK_MASK                              0x1
-#       define R600_NOOFRANK_SHIFT                             1
-#       define R600_NOOFRANK_MASK                              0x1
-#       define R600_NOOFROWS_SHIFT                             2
-#       define R600_NOOFROWS_MASK                              0x7
-#       define R600_NOOFCOLS_SHIFT                             5
-#       define R600_NOOFCOLS_MASK                              0x3
-#       define R600_CHANSIZE_SHIFT                             7
-#       define R600_CHANSIZE_MASK                              0x1
-#       define R600_BURSTLENGTH_SHIFT                          8
-#       define R600_BURSTLENGTH_MASK                           0x1
-
-#define R600_VM_L2_CNTL                                        0x1400
-#       define R600_VM_L2_CACHE_EN                             (1 << 0)
-#       define R600_VM_L2_FRAG_PROC                            (1 << 1)
-#       define R600_VM_ENABLE_PTE_CACHE_LRU_W                  (1 << 9)
-#       define R600_VM_L2_CNTL_QUEUE_SIZE(x)                   ((x) << 13)
-#       define R700_VM_L2_CNTL_QUEUE_SIZE(x)                   ((x) << 14)
-
-#define R600_VM_L2_CNTL2                                       0x1404
-#       define R600_VM_L2_CNTL2_INVALIDATE_ALL_L1_TLBS         (1 << 0)
-#       define R600_VM_L2_CNTL2_INVALIDATE_L2_CACHE            (1 << 1)
-#define R600_VM_L2_CNTL3                                       0x1408
-#       define R600_VM_L2_CNTL3_BANK_SELECT_0(x)               ((x) << 0)
-#       define R600_VM_L2_CNTL3_BANK_SELECT_1(x)               ((x) << 5)
-#       define R600_VM_L2_CNTL3_CACHE_UPDATE_MODE(x)           ((x) << 10)
-#       define R700_VM_L2_CNTL3_BANK_SELECT(x)                 ((x) << 0)
-#       define R700_VM_L2_CNTL3_CACHE_UPDATE_MODE(x)           ((x) << 6)
-
-#define R600_VM_L2_STATUS                                      0x140c
-
-#define R600_VM_CONTEXT0_CNTL                                  0x1410
-#       define R600_VM_ENABLE_CONTEXT                          (1 << 0)
-#       define R600_VM_PAGE_TABLE_DEPTH_FLAT                   (0 << 1)
-
-#define R600_VM_CONTEXT0_CNTL2                                 0x1430
-#define R600_VM_CONTEXT0_REQUEST_RESPONSE                      0x1470
-#define R600_VM_CONTEXT0_INVALIDATION_LOW_ADDR                 0x1490
-#define R600_VM_CONTEXT0_INVALIDATION_HIGH_ADDR                0x14b0
-#define R600_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR                  0x1574
-#define R600_VM_CONTEXT0_PAGE_TABLE_START_ADDR                 0x1594
-#define R600_VM_CONTEXT0_PAGE_TABLE_END_ADDR                   0x15b4
-
-#define R700_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR                  0x153c
-#define R700_VM_CONTEXT0_PAGE_TABLE_START_ADDR                 0x155c
-#define R700_VM_CONTEXT0_PAGE_TABLE_END_ADDR                   0x157c
-
-#define R600_HDP_HOST_PATH_CNTL                                0x2c00
-
-#define R600_GRBM_CNTL                                         0x8000
-#       define R600_GRBM_READ_TIMEOUT(x)                       ((x) << 0)
-
-#define R600_GRBM_STATUS                                       0x8010
-#       define R600_CMDFIFO_AVAIL_MASK                         0x1f
-#       define R700_CMDFIFO_AVAIL_MASK                         0xf
-#       define R600_GUI_ACTIVE                                 (1 << 31)
-#define R600_GRBM_STATUS2                                      0x8014
-#define R600_GRBM_SOFT_RESET                                   0x8020
-#       define R600_SOFT_RESET_CP                              (1 << 0)
-#define R600_WAIT_UNTIL		                               0x8040
-
-#define R600_CP_SEM_WAIT_TIMER                                 0x85bc
-#define R600_CP_ME_CNTL                                        0x86d8
-#       define R600_CP_ME_HALT                                 (1 << 28)
-#define R600_CP_QUEUE_THRESHOLDS                               0x8760
-#       define R600_ROQ_IB1_START(x)                           ((x) << 0)
-#       define R600_ROQ_IB2_START(x)                           ((x) << 8)
-#define R600_CP_MEQ_THRESHOLDS                                 0x8764
-#       define R700_STQ_SPLIT(x)                               ((x) << 0)
-#       define R600_MEQ_END(x)                                 ((x) << 16)
-#       define R600_ROQ_END(x)                                 ((x) << 24)
-#define R600_CP_PERFMON_CNTL                                   0x87fc
-#define R600_CP_RB_BASE                                        0xc100
-#define R600_CP_RB_CNTL                                        0xc104
-#       define R600_RB_BUFSZ(x)                                ((x) << 0)
-#       define R600_RB_BLKSZ(x)                                ((x) << 8)
-#	define R600_BUF_SWAP_32BIT		               (2 << 16)
-#       define R600_RB_NO_UPDATE                               (1 << 27)
-#       define R600_RB_RPTR_WR_ENA                             (1 << 31)
-#define R600_CP_RB_RPTR_WR                                     0xc108
-#define R600_CP_RB_RPTR_ADDR                                   0xc10c
-#define R600_CP_RB_RPTR_ADDR_HI                                0xc110
-#define R600_CP_RB_WPTR                                        0xc114
-#define R600_CP_RB_WPTR_ADDR                                   0xc118
-#define R600_CP_RB_WPTR_ADDR_HI                                0xc11c
-#define R600_CP_RB_RPTR                                        0x8700
-#define R600_CP_RB_WPTR_DELAY                                  0x8704
-#define R600_CP_PFP_UCODE_ADDR                                 0xc150
-#define R600_CP_PFP_UCODE_DATA                                 0xc154
-#define R600_CP_ME_RAM_RADDR                                   0xc158
-#define R600_CP_ME_RAM_WADDR                                   0xc15c
-#define R600_CP_ME_RAM_DATA                                    0xc160
-#define R600_CP_DEBUG                                          0xc1fc
-
-#define R600_PA_CL_ENHANCE                                     0x8a14
-#       define R600_CLIP_VTX_REORDER_ENA                       (1 << 0)
-#       define R600_NUM_CLIP_SEQ(x)                            ((x) << 1)
-#define R600_PA_SC_LINE_STIPPLE_STATE                          0x8b10
-#define R600_PA_SC_MULTI_CHIP_CNTL                             0x8b20
-#define R700_PA_SC_FORCE_EOV_MAX_CNTS                          0x8b24
-#       define R700_FORCE_EOV_MAX_CLK_CNT(x)                   ((x) << 0)
-#       define R700_FORCE_EOV_MAX_REZ_CNT(x)                   ((x) << 16)
-#define R600_PA_SC_AA_SAMPLE_LOCS_2S                           0x8b40
-#define R600_PA_SC_AA_SAMPLE_LOCS_4S                           0x8b44
-#define R600_PA_SC_AA_SAMPLE_LOCS_8S_WD0                       0x8b48
-#define R600_PA_SC_AA_SAMPLE_LOCS_8S_WD1                       0x8b4c
-#       define R600_S0_X(x)                                    ((x) << 0)
-#       define R600_S0_Y(x)                                    ((x) << 4)
-#       define R600_S1_X(x)                                    ((x) << 8)
-#       define R600_S1_Y(x)                                    ((x) << 12)
-#       define R600_S2_X(x)                                    ((x) << 16)
-#       define R600_S2_Y(x)                                    ((x) << 20)
-#       define R600_S3_X(x)                                    ((x) << 24)
-#       define R600_S3_Y(x)                                    ((x) << 28)
-#       define R600_S4_X(x)                                    ((x) << 0)
-#       define R600_S4_Y(x)                                    ((x) << 4)
-#       define R600_S5_X(x)                                    ((x) << 8)
-#       define R600_S5_Y(x)                                    ((x) << 12)
-#       define R600_S6_X(x)                                    ((x) << 16)
-#       define R600_S6_Y(x)                                    ((x) << 20)
-#       define R600_S7_X(x)                                    ((x) << 24)
-#       define R600_S7_Y(x)                                    ((x) << 28)
-#define R600_PA_SC_FIFO_SIZE                                   0x8bd0
-#       define R600_SC_PRIM_FIFO_SIZE(x)                       ((x) << 0)
-#       define R600_SC_HIZ_TILE_FIFO_SIZE(x)                   ((x) << 8)
-#       define R600_SC_EARLYZ_TILE_FIFO_SIZE(x)                ((x) << 16)
-#define R700_PA_SC_FIFO_SIZE_R7XX                              0x8bcc
-#       define R700_SC_PRIM_FIFO_SIZE(x)                       ((x) << 0)
-#       define R700_SC_HIZ_TILE_FIFO_SIZE(x)                   ((x) << 12)
-#       define R700_SC_EARLYZ_TILE_FIFO_SIZE(x)                ((x) << 20)
-#define R600_PA_SC_ENHANCE                                     0x8bf0
-#       define R600_FORCE_EOV_MAX_CLK_CNT(x)                   ((x) << 0)
-#       define R600_FORCE_EOV_MAX_TILE_CNT(x)                  ((x) << 12)
-#define R600_PA_SC_CLIPRECT_RULE                               0x2820c
-#define R700_PA_SC_EDGERULE                                    0x28230
-#define R600_PA_SC_LINE_STIPPLE                                0x28a0c
-#define R600_PA_SC_MODE_CNTL                                   0x28a4c
-#define R600_PA_SC_AA_CONFIG                                   0x28c04
-
-#define R600_SX_EXPORT_BUFFER_SIZES                            0x900c
-#       define R600_COLOR_BUFFER_SIZE(x)                       ((x) << 0)
-#       define R600_POSITION_BUFFER_SIZE(x)                    ((x) << 8)
-#       define R600_SMX_BUFFER_SIZE(x)                         ((x) << 16)
-#define R600_SX_DEBUG_1                                        0x9054
-#       define R600_SMX_EVENT_RELEASE                          (1 << 0)
-#       define R600_ENABLE_NEW_SMX_ADDRESS                     (1 << 16)
-#define R700_SX_DEBUG_1                                        0x9058
-#       define R700_ENABLE_NEW_SMX_ADDRESS                     (1 << 16)
-#define R600_SX_MISC                                           0x28350
-
-#define R600_DB_DEBUG                                          0x9830
-#       define R600_PREZ_MUST_WAIT_FOR_POSTZ_DONE              (1 << 31)
-#define R600_DB_WATERMARKS                                     0x9838
-#       define R600_DEPTH_FREE(x)                              ((x) << 0)
-#       define R600_DEPTH_FLUSH(x)                             ((x) << 5)
-#       define R600_DEPTH_PENDING_FREE(x)                      ((x) << 15)
-#       define R600_DEPTH_CACHELINE_FREE(x)                    ((x) << 20)
-#define R700_DB_DEBUG3                                         0x98b0
-#       define R700_DB_CLK_OFF_DELAY(x)                        ((x) << 11)
-#define RV700_DB_DEBUG4                                        0x9b8c
-#       define RV700_DISABLE_TILE_COVERED_FOR_PS_ITER          (1 << 6)
-
-#define R600_VGT_CACHE_INVALIDATION                            0x88c4
-#       define R600_CACHE_INVALIDATION(x)                      ((x) << 0)
-#       define R600_VC_ONLY                                    0
-#       define R600_TC_ONLY                                    1
-#       define R600_VC_AND_TC                                  2
-#       define R700_AUTO_INVLD_EN(x)                           ((x) << 6)
-#       define R700_NO_AUTO                                    0
-#       define R700_ES_AUTO                                    1
-#       define R700_GS_AUTO                                    2
-#       define R700_ES_AND_GS_AUTO                             3
-#define R600_VGT_GS_PER_ES                                     0x88c8
-#define R600_VGT_ES_PER_GS                                     0x88cc
-#define R600_VGT_GS_PER_VS                                     0x88e8
-#define R600_VGT_GS_VERTEX_REUSE                               0x88d4
-#define R600_VGT_NUM_INSTANCES                                 0x8974
-#define R600_VGT_STRMOUT_EN                                    0x28ab0
-#define R600_VGT_EVENT_INITIATOR                               0x28a90
-#       define R600_CACHE_FLUSH_AND_INV_EVENT                  (0x16 << 0)
-#define R600_VGT_VERTEX_REUSE_BLOCK_CNTL                       0x28c58
-#       define R600_VTX_REUSE_DEPTH_MASK                       0xff
-#define R600_VGT_OUT_DEALLOC_CNTL                              0x28c5c
-#       define R600_DEALLOC_DIST_MASK                          0x7f
-
-#define R600_CB_COLOR0_BASE                                    0x28040
-#define R600_CB_COLOR1_BASE                                    0x28044
-#define R600_CB_COLOR2_BASE                                    0x28048
-#define R600_CB_COLOR3_BASE                                    0x2804c
-#define R600_CB_COLOR4_BASE                                    0x28050
-#define R600_CB_COLOR5_BASE                                    0x28054
-#define R600_CB_COLOR6_BASE                                    0x28058
-#define R600_CB_COLOR7_BASE                                    0x2805c
-#define R600_CB_COLOR7_FRAG                                    0x280fc
-
-#define R600_CB_COLOR0_SIZE                                    0x28060
-#define R600_CB_COLOR0_VIEW                                    0x28080
-#define R600_CB_COLOR0_INFO                                    0x280a0
-#define R600_CB_COLOR0_TILE                                    0x280c0
-#define R600_CB_COLOR0_FRAG                                    0x280e0
-#define R600_CB_COLOR0_MASK                                    0x28100
-
-#define AVIVO_D1MODE_VLINE_START_END                           0x6538
-#define AVIVO_D2MODE_VLINE_START_END                           0x6d38
-#define R600_CP_COHER_BASE                                     0x85f8
-#define R600_DB_DEPTH_BASE                                     0x2800c
-#define R600_SQ_PGM_START_FS                                   0x28894
-#define R600_SQ_PGM_START_ES                                   0x28880
-#define R600_SQ_PGM_START_VS                                   0x28858
-#define R600_SQ_PGM_RESOURCES_VS                               0x28868
-#define R600_SQ_PGM_CF_OFFSET_VS                               0x288d0
-#define R600_SQ_PGM_START_GS                                   0x2886c
-#define R600_SQ_PGM_START_PS                                   0x28840
-#define R600_SQ_PGM_RESOURCES_PS                               0x28850
-#define R600_SQ_PGM_EXPORTS_PS                                 0x28854
-#define R600_SQ_PGM_CF_OFFSET_PS                               0x288cc
-#define R600_VGT_DMA_BASE                                      0x287e8
-#define R600_VGT_DMA_BASE_HI                                   0x287e4
-#define R600_VGT_STRMOUT_BASE_OFFSET_0                         0x28b10
-#define R600_VGT_STRMOUT_BASE_OFFSET_1                         0x28b14
-#define R600_VGT_STRMOUT_BASE_OFFSET_2                         0x28b18
-#define R600_VGT_STRMOUT_BASE_OFFSET_3                         0x28b1c
-#define R600_VGT_STRMOUT_BASE_OFFSET_HI_0                      0x28b44
-#define R600_VGT_STRMOUT_BASE_OFFSET_HI_1                      0x28b48
-#define R600_VGT_STRMOUT_BASE_OFFSET_HI_2                      0x28b4c
-#define R600_VGT_STRMOUT_BASE_OFFSET_HI_3                      0x28b50
-#define R600_VGT_STRMOUT_BUFFER_BASE_0                         0x28ad8
-#define R600_VGT_STRMOUT_BUFFER_BASE_1                         0x28ae8
-#define R600_VGT_STRMOUT_BUFFER_BASE_2                         0x28af8
-#define R600_VGT_STRMOUT_BUFFER_BASE_3                         0x28b08
-#define R600_VGT_STRMOUT_BUFFER_OFFSET_0                       0x28adc
-#define R600_VGT_STRMOUT_BUFFER_OFFSET_1                       0x28aec
-#define R600_VGT_STRMOUT_BUFFER_OFFSET_2                       0x28afc
-#define R600_VGT_STRMOUT_BUFFER_OFFSET_3                       0x28b0c
-
-#define R600_VGT_PRIMITIVE_TYPE                                0x8958
-
-#define R600_PA_SC_SCREEN_SCISSOR_TL                           0x28030
-#define R600_PA_SC_GENERIC_SCISSOR_TL                          0x28240
-#define R600_PA_SC_WINDOW_SCISSOR_TL                           0x28204
-
-#define R600_TC_CNTL                                           0x9608
-#       define R600_TC_L2_SIZE(x)                              ((x) << 5)
-#       define R600_L2_DISABLE_LATE_HIT                        (1 << 9)
-
-#define R600_ARB_POP                                           0x2418
-#       define R600_ENABLE_TC128                               (1 << 30)
-#define R600_ARB_GDEC_RD_CNTL                                  0x246c
-
-#define R600_TA_CNTL_AUX                                       0x9508
-#       define R600_DISABLE_CUBE_WRAP                          (1 << 0)
-#       define R600_DISABLE_CUBE_ANISO                         (1 << 1)
-#       define R700_GETLOD_SELECT(x)                           ((x) << 2)
-#       define R600_SYNC_GRADIENT                              (1 << 24)
-#       define R600_SYNC_WALKER                                (1 << 25)
-#       define R600_SYNC_ALIGNER                               (1 << 26)
-#       define R600_BILINEAR_PRECISION_6_BIT                   (0 << 31)
-#       define R600_BILINEAR_PRECISION_8_BIT                   (1 << 31)
-
-#define R700_TCP_CNTL                                          0x9610
-
-#define R600_SMX_DC_CTL0                                       0xa020
-#       define R700_USE_HASH_FUNCTION                          (1 << 0)
-#       define R700_CACHE_DEPTH(x)                             ((x) << 1)
-#       define R700_FLUSH_ALL_ON_EVENT                         (1 << 10)
-#       define R700_STALL_ON_EVENT                             (1 << 11)
-#define R700_SMX_EVENT_CTL                                     0xa02c
-#       define R700_ES_FLUSH_CTL(x)                            ((x) << 0)
-#       define R700_GS_FLUSH_CTL(x)                            ((x) << 3)
-#       define R700_ACK_FLUSH_CTL(x)                           ((x) << 6)
-#       define R700_SYNC_FLUSH_CTL                             (1 << 8)
-
-#define R600_SQ_CONFIG                                         0x8c00
-#       define R600_VC_ENABLE                                  (1 << 0)
-#       define R600_EXPORT_SRC_C                               (1 << 1)
-#       define R600_DX9_CONSTS                                 (1 << 2)
-#       define R600_ALU_INST_PREFER_VECTOR                     (1 << 3)
-#       define R600_DX10_CLAMP                                 (1 << 4)
-#       define R600_CLAUSE_SEQ_PRIO(x)                         ((x) << 8)
-#       define R600_PS_PRIO(x)                                 ((x) << 24)
-#       define R600_VS_PRIO(x)                                 ((x) << 26)
-#       define R600_GS_PRIO(x)                                 ((x) << 28)
-#       define R600_ES_PRIO(x)                                 ((x) << 30)
-#define R600_SQ_GPR_RESOURCE_MGMT_1                            0x8c04
-#       define R600_NUM_PS_GPRS(x)                             ((x) << 0)
-#       define R600_NUM_VS_GPRS(x)                             ((x) << 16)
-#       define R700_DYN_GPR_ENABLE                             (1 << 27)
-#       define R600_NUM_CLAUSE_TEMP_GPRS(x)                    ((x) << 28)
-#define R600_SQ_GPR_RESOURCE_MGMT_2                            0x8c08
-#       define R600_NUM_GS_GPRS(x)                             ((x) << 0)
-#       define R600_NUM_ES_GPRS(x)                             ((x) << 16)
-#define R600_SQ_THREAD_RESOURCE_MGMT                           0x8c0c
-#       define R600_NUM_PS_THREADS(x)                          ((x) << 0)
-#       define R600_NUM_VS_THREADS(x)                          ((x) << 8)
-#       define R600_NUM_GS_THREADS(x)                          ((x) << 16)
-#       define R600_NUM_ES_THREADS(x)                          ((x) << 24)
-#define R600_SQ_STACK_RESOURCE_MGMT_1                          0x8c10
-#       define R600_NUM_PS_STACK_ENTRIES(x)                    ((x) << 0)
-#       define R600_NUM_VS_STACK_ENTRIES(x)                    ((x) << 16)
-#define R600_SQ_STACK_RESOURCE_MGMT_2                          0x8c14
-#       define R600_NUM_GS_STACK_ENTRIES(x)                    ((x) << 0)
-#       define R600_NUM_ES_STACK_ENTRIES(x)                    ((x) << 16)
-#define R600_SQ_MS_FIFO_SIZES                                  0x8cf0
-#       define R600_CACHE_FIFO_SIZE(x)                         ((x) << 0)
-#       define R600_FETCH_FIFO_HIWATER(x)                      ((x) << 8)
-#       define R600_DONE_FIFO_HIWATER(x)                       ((x) << 16)
-#       define R600_ALU_UPDATE_FIFO_HIWATER(x)                 ((x) << 24)
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_0                         0x8db0
-#       define R700_SIMDA_RING0(x)                             ((x) << 0)
-#       define R700_SIMDA_RING1(x)                             ((x) << 8)
-#       define R700_SIMDB_RING0(x)                             ((x) << 16)
-#       define R700_SIMDB_RING1(x)                             ((x) << 24)
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_1                         0x8db4
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_2                         0x8db8
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_3                         0x8dbc
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_4                         0x8dc0
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_5                         0x8dc4
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_6                         0x8dc8
-#define R700_SQ_DYN_GPR_SIZE_SIMD_AB_7                         0x8dcc
-
-#define R600_SPI_PS_IN_CONTROL_0                               0x286cc
-#       define R600_NUM_INTERP(x)                              ((x) << 0)
-#       define R600_POSITION_ENA                               (1 << 8)
-#       define R600_POSITION_CENTROID                          (1 << 9)
-#       define R600_POSITION_ADDR(x)                           ((x) << 10)
-#       define R600_PARAM_GEN(x)                               ((x) << 15)
-#       define R600_PARAM_GEN_ADDR(x)                          ((x) << 19)
-#       define R600_BARYC_SAMPLE_CNTL(x)                       ((x) << 26)
-#       define R600_PERSP_GRADIENT_ENA                         (1 << 28)
-#       define R600_LINEAR_GRADIENT_ENA                        (1 << 29)
-#       define R600_POSITION_SAMPLE                            (1 << 30)
-#       define R600_BARYC_AT_SAMPLE_ENA                        (1 << 31)
-#define R600_SPI_PS_IN_CONTROL_1                               0x286d0
-#       define R600_GEN_INDEX_PIX                              (1 << 0)
-#       define R600_GEN_INDEX_PIX_ADDR(x)                      ((x) << 1)
-#       define R600_FRONT_FACE_ENA                             (1 << 8)
-#       define R600_FRONT_FACE_CHAN(x)                         ((x) << 9)
-#       define R600_FRONT_FACE_ALL_BITS                        (1 << 11)
-#       define R600_FRONT_FACE_ADDR(x)                         ((x) << 12)
-#       define R600_FOG_ADDR(x)                                ((x) << 17)
-#       define R600_FIXED_PT_POSITION_ENA                      (1 << 24)
-#       define R600_FIXED_PT_POSITION_ADDR(x)                  ((x) << 25)
-#       define R700_POSITION_ULC                               (1 << 30)
-#define R600_SPI_INPUT_Z                                       0x286d8
-
-#define R600_SPI_CONFIG_CNTL                                   0x9100
-#       define R600_GPR_WRITE_PRIORITY(x)                      ((x) << 0)
-#       define R600_DISABLE_INTERP_1                           (1 << 5)
-#define R600_SPI_CONFIG_CNTL_1                                 0x913c
-#       define R600_VTX_DONE_DELAY(x)                          ((x) << 0)
-#       define R600_INTERP_ONE_PRIM_PER_ROW                    (1 << 4)
-
-#define R600_GB_TILING_CONFIG                                  0x98f0
-#       define R600_PIPE_TILING(x)                             ((x) << 1)
-#       define R600_BANK_TILING(x)                             ((x) << 4)
-#       define R600_GROUP_SIZE(x)                              ((x) << 6)
-#       define R600_ROW_TILING(x)                              ((x) << 8)
-#       define R600_BANK_SWAPS(x)                              ((x) << 11)
-#       define R600_SAMPLE_SPLIT(x)                            ((x) << 14)
-#       define R600_BACKEND_MAP(x)                             ((x) << 16)
-#define R600_DCP_TILING_CONFIG                                 0x6ca0
-#define R600_HDP_TILING_CONFIG                                 0x2f3c
-
-#define R600_CC_RB_BACKEND_DISABLE                             0x98f4
-#define R700_CC_SYS_RB_BACKEND_DISABLE                         0x3f88
-#       define R600_BACKEND_DISABLE(x)                         ((x) << 16)
-
-#define R600_CC_GC_SHADER_PIPE_CONFIG                          0x8950
-#define R600_GC_USER_SHADER_PIPE_CONFIG                        0x8954
-#       define R600_INACTIVE_QD_PIPES(x)                       ((x) << 8)
-#       define R600_INACTIVE_QD_PIPES_MASK                     (0xff << 8)
-#       define R600_INACTIVE_SIMDS(x)                          ((x) << 16)
-#       define R600_INACTIVE_SIMDS_MASK                        (0xff << 16)
-
-#define R700_CGTS_SYS_TCC_DISABLE                              0x3f90
-#define R700_CGTS_USER_SYS_TCC_DISABLE                         0x3f94
-#define R700_CGTS_TCC_DISABLE                                  0x9148
-#define R700_CGTS_USER_TCC_DISABLE                             0x914c
-
-/* Constants */
-#define RADEON_MAX_USEC_TIMEOUT		100000	/* 100 ms */
-
-#define RADEON_LAST_FRAME_REG		RADEON_SCRATCH_REG0
-#define RADEON_LAST_DISPATCH_REG	RADEON_SCRATCH_REG1
-#define RADEON_LAST_CLEAR_REG		RADEON_SCRATCH_REG2
-#define RADEON_LAST_SWI_REG		RADEON_SCRATCH_REG3
-#define RADEON_LAST_DISPATCH		1
-
-#define R600_LAST_FRAME_REG		R600_SCRATCH_REG0
-#define R600_LAST_DISPATCH_REG	        R600_SCRATCH_REG1
-#define R600_LAST_CLEAR_REG		R600_SCRATCH_REG2
-#define R600_LAST_SWI_REG		R600_SCRATCH_REG3
-
-#define RADEON_MAX_VB_AGE		0x7fffffff
-#define RADEON_MAX_VB_VERTS		(0xffff)
-
-#define RADEON_RING_HIGH_MARK		128
-
-#define RADEON_PCIGART_TABLE_SIZE      (32*1024)
-
-#define RADEON_READ(reg)	DRM_READ32(  dev_priv->mmio, (reg) )
-#define RADEON_WRITE(reg, val)                                          \
-do {									\
-	if (reg < 0x10000) {				                \
-		DRM_WRITE32(dev_priv->mmio, (reg), (val));		\
-	} else {                                                        \
-		DRM_WRITE32(dev_priv->mmio, RADEON_MM_INDEX, (reg));	\
-		DRM_WRITE32(dev_priv->mmio, RADEON_MM_DATA, (val));	\
-	}                                                               \
-} while (0)
-#define RADEON_READ8(reg)	DRM_READ8(  dev_priv->mmio, (reg) )
-#define RADEON_WRITE8(reg,val)	DRM_WRITE8( dev_priv->mmio, (reg), (val) )
-
-#define RADEON_WRITE_PLL(addr, val)					\
-do {									\
-	RADEON_WRITE8(RADEON_CLOCK_CNTL_INDEX,				\
-		       ((addr) & 0x1f) | RADEON_PLL_WR_EN );		\
-	RADEON_WRITE(RADEON_CLOCK_CNTL_DATA, (val));			\
-} while (0)
-
-#define RADEON_WRITE_PCIE(addr, val)					\
-do {									\
-	RADEON_WRITE8(RADEON_PCIE_INDEX,				\
-			((addr) & 0xff));				\
-	RADEON_WRITE(RADEON_PCIE_DATA, (val));			\
-} while (0)
-
-#define R500_WRITE_MCIND(addr, val)					\
-do {								\
-	RADEON_WRITE(R520_MC_IND_INDEX, 0xff0000 | ((addr) & 0xff));	\
-	RADEON_WRITE(R520_MC_IND_DATA, (val));			\
-	RADEON_WRITE(R520_MC_IND_INDEX, 0);	\
-} while (0)
-
-#define RS480_WRITE_MCIND(addr, val)				\
-do {									\
-	RADEON_WRITE(RS480_NB_MC_INDEX,				\
-			((addr) & 0xff) | RS480_NB_MC_IND_WR_EN);	\
-	RADEON_WRITE(RS480_NB_MC_DATA, (val));			\
-	RADEON_WRITE(RS480_NB_MC_INDEX, 0xff);			\
-} while (0)
-
-#define RS690_WRITE_MCIND(addr, val)					\
-do {								\
-	RADEON_WRITE(RS690_MC_INDEX, RS690_MC_INDEX_WR_EN | ((addr) & RS690_MC_INDEX_MASK));	\
-	RADEON_WRITE(RS690_MC_DATA, val);			\
-	RADEON_WRITE(RS690_MC_INDEX, RS690_MC_INDEX_WR_ACK);	\
-} while (0)
-
-#define RS600_WRITE_MCIND(addr, val)				\
-do {							        \
-	RADEON_WRITE(RS600_MC_INDEX, RS600_MC_IND_WR_EN | RS600_MC_IND_CITF_ARB0 | ((addr) & RS600_MC_ADDR_MASK)); \
-	RADEON_WRITE(RS600_MC_DATA, val);                       \
-} while (0)
-
-#define IGP_WRITE_MCIND(addr, val)				\
-do {									\
-	if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS690) ||   \
-	    ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS740))      \
-		RS690_WRITE_MCIND(addr, val);				\
-	else if ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RS600)  \
-		RS600_WRITE_MCIND(addr, val);				\
-	else								\
-		RS480_WRITE_MCIND(addr, val);				\
-} while (0)
-
-#define CP_PACKET0( reg, n )						\
-	(RADEON_CP_PACKET0 | ((n) << 16) | ((reg) >> 2))
-#define CP_PACKET0_TABLE( reg, n )					\
-	(RADEON_CP_PACKET0 | RADEON_ONE_REG_WR | ((n) << 16) | ((reg) >> 2))
-#define CP_PACKET1( reg0, reg1 )					\
-	(RADEON_CP_PACKET1 | (((reg1) >> 2) << 15) | ((reg0) >> 2))
-#define CP_PACKET2()							\
-	(RADEON_CP_PACKET2)
-#define CP_PACKET3( pkt, n )						\
-	(RADEON_CP_PACKET3 | (pkt) | ((n) << 16))
-
-/* ================================================================
- * Engine control helper macros
- */
-
-#define RADEON_WAIT_UNTIL_2D_IDLE() do {				\
-	OUT_RING( CP_PACKET0( RADEON_WAIT_UNTIL, 0 ) );			\
-	OUT_RING( (RADEON_WAIT_2D_IDLECLEAN |				\
-		   RADEON_WAIT_HOST_IDLECLEAN) );			\
-} while (0)
-
-#define RADEON_WAIT_UNTIL_3D_IDLE() do {				\
-	OUT_RING( CP_PACKET0( RADEON_WAIT_UNTIL, 0 ) );			\
-	OUT_RING( (RADEON_WAIT_3D_IDLECLEAN |				\
-		   RADEON_WAIT_HOST_IDLECLEAN) );			\
-} while (0)
-
-#define RADEON_WAIT_UNTIL_IDLE() do {					\
-	OUT_RING( CP_PACKET0( RADEON_WAIT_UNTIL, 0 ) );			\
-	OUT_RING( (RADEON_WAIT_2D_IDLECLEAN |				\
-		   RADEON_WAIT_3D_IDLECLEAN |				\
-		   RADEON_WAIT_HOST_IDLECLEAN) );			\
-} while (0)
-
-#define RADEON_WAIT_UNTIL_PAGE_FLIPPED() do {				\
-	OUT_RING( CP_PACKET0( RADEON_WAIT_UNTIL, 0 ) );			\
-	OUT_RING( RADEON_WAIT_CRTC_PFLIP );				\
-} while (0)
-
-#define RADEON_FLUSH_CACHE() do {					\
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) <= CHIP_RV280) {	\
-		OUT_RING(CP_PACKET0(RADEON_RB3D_DSTCACHE_CTLSTAT, 0));	\
-		OUT_RING(RADEON_RB3D_DC_FLUSH);				\
-	} else {                                                        \
-		OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));	\
-		OUT_RING(R300_RB3D_DC_FLUSH);				\
-	}                                                               \
-} while (0)
-
-#define RADEON_PURGE_CACHE() do {					\
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) <= CHIP_RV280) {	\
-		OUT_RING(CP_PACKET0(RADEON_RB3D_DSTCACHE_CTLSTAT, 0));	\
-		OUT_RING(RADEON_RB3D_DC_FLUSH | RADEON_RB3D_DC_FREE);	\
-	} else {                                                        \
-		OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));	\
-		OUT_RING(R300_RB3D_DC_FLUSH | R300_RB3D_DC_FREE);	\
-	}                                                               \
-} while (0)
-
-#define RADEON_FLUSH_ZCACHE() do {					\
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) <= CHIP_RV280) {	\
-		OUT_RING(CP_PACKET0(RADEON_RB3D_ZCACHE_CTLSTAT, 0));	\
-		OUT_RING(RADEON_RB3D_ZC_FLUSH);				\
-	} else {                                                        \
-		OUT_RING(CP_PACKET0(R300_ZB_ZCACHE_CTLSTAT, 0));	\
-		OUT_RING(R300_ZC_FLUSH);				\
-	}                                                               \
-} while (0)
-
-#define RADEON_PURGE_ZCACHE() do {					\
-	if ((dev_priv->flags & RADEON_FAMILY_MASK) <= CHIP_RV280) {	\
-		OUT_RING(CP_PACKET0(RADEON_RB3D_ZCACHE_CTLSTAT, 0));	\
-		OUT_RING(RADEON_RB3D_ZC_FLUSH | RADEON_RB3D_ZC_FREE);			\
-	} else {                                                        \
-		OUT_RING(CP_PACKET0(R300_ZB_ZCACHE_CTLSTAT, 0));	\
-		OUT_RING(R300_ZC_FLUSH | R300_ZC_FREE);				\
-	}                                                               \
-} while (0)
-
-/* ================================================================
- * Misc helper macros
- */
-
-/* Perfbox functionality only.
- */
-#define RING_SPACE_TEST_WITH_RETURN( dev_priv )				\
-do {									\
-	if (!(dev_priv->stats.boxes & RADEON_BOX_DMA_IDLE)) {		\
-		u32 head = GET_RING_HEAD( dev_priv );			\
-		if (head == dev_priv->ring.tail)			\
-			dev_priv->stats.boxes |= RADEON_BOX_DMA_IDLE;	\
-	}								\
-} while (0)
-
-#define VB_AGE_TEST_WITH_RETURN( dev_priv )				\
-do {								\
-	struct drm_radeon_master_private *master_priv = file_priv->master->driver_priv;	\
-	drm_radeon_sarea_t *sarea_priv = master_priv->sarea_priv;	\
-	if ( sarea_priv->last_dispatch >= RADEON_MAX_VB_AGE ) {		\
-		int __ret;						\
-		if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600) \
-			__ret = r600_do_cp_idle(dev_priv);		\
-		else							\
-			__ret = radeon_do_cp_idle(dev_priv);		\
-		if ( __ret ) return __ret;				\
-		sarea_priv->last_dispatch = 0;				\
-		radeon_freelist_reset( dev );				\
-	}								\
-} while (0)
-
-#define RADEON_DISPATCH_AGE( age ) do {					\
-	OUT_RING( CP_PACKET0( RADEON_LAST_DISPATCH_REG, 0 ) );		\
-	OUT_RING( age );						\
-} while (0)
-
-#define RADEON_FRAME_AGE( age ) do {					\
-	OUT_RING( CP_PACKET0( RADEON_LAST_FRAME_REG, 0 ) );		\
-	OUT_RING( age );						\
-} while (0)
-
-#define RADEON_CLEAR_AGE( age ) do {					\
-	OUT_RING( CP_PACKET0( RADEON_LAST_CLEAR_REG, 0 ) );		\
-	OUT_RING( age );						\
-} while (0)
-
-#define R600_DISPATCH_AGE(age) do {					\
-	OUT_RING(CP_PACKET3(R600_IT_SET_CONFIG_REG, 1));		\
-	OUT_RING((R600_LAST_DISPATCH_REG - R600_SET_CONFIG_REG_OFFSET) >> 2);  \
-	OUT_RING(age);							\
-} while (0)
-
-#define R600_FRAME_AGE(age) do {					\
-	OUT_RING(CP_PACKET3(R600_IT_SET_CONFIG_REG, 1));		\
-	OUT_RING((R600_LAST_FRAME_REG - R600_SET_CONFIG_REG_OFFSET) >> 2);  \
-	OUT_RING(age);							\
-} while (0)
-
-#define R600_CLEAR_AGE(age) do {					\
-	OUT_RING(CP_PACKET3(R600_IT_SET_CONFIG_REG, 1));		\
-	OUT_RING((R600_LAST_CLEAR_REG - R600_SET_CONFIG_REG_OFFSET) >> 2);  \
-	OUT_RING(age);							\
-} while (0)
-
-/* ================================================================
- * Ring control
- */
-
-#define RADEON_VERBOSE	0
-
-#define RING_LOCALS	int write, _nr, _align_nr; unsigned int mask; u32 *ring;
-
-#define RADEON_RING_ALIGN 16
-
-#define BEGIN_RING( n ) do {						\
-	if ( RADEON_VERBOSE ) {						\
-		DRM_INFO( "BEGIN_RING( %d )\n", (n));			\
-	}								\
-	_align_nr = RADEON_RING_ALIGN - ((dev_priv->ring.tail + n) & (RADEON_RING_ALIGN-1));	\
-	_align_nr += n;							\
-	if (dev_priv->ring.space <= (_align_nr * sizeof(u32))) {	\
-                COMMIT_RING();						\
-		radeon_wait_ring( dev_priv, _align_nr * sizeof(u32));	\
-	}								\
-	_nr = n; dev_priv->ring.space -= (n) * sizeof(u32);		\
-	ring = dev_priv->ring.start;					\
-	write = dev_priv->ring.tail;					\
-	mask = dev_priv->ring.tail_mask;				\
-} while (0)
-
-#define ADVANCE_RING() do {						\
-	if ( RADEON_VERBOSE ) {						\
-		DRM_INFO( "ADVANCE_RING() wr=0x%06x tail=0x%06x\n",	\
-			  write, dev_priv->ring.tail );			\
-	}								\
-	if (((dev_priv->ring.tail + _nr) & mask) != write) {		\
-		DRM_ERROR(						\
-			"ADVANCE_RING(): mismatch: nr: %x write: %x line: %d\n",	\
-			((dev_priv->ring.tail + _nr) & mask),		\
-			write, __LINE__);				\
-	} else								\
-		dev_priv->ring.tail = write;				\
-} while (0)
-
-extern void radeon_commit_ring(drm_radeon_private_t *dev_priv);
-
-#define COMMIT_RING() do {						\
-		radeon_commit_ring(dev_priv);				\
-	} while(0)
-
-#define OUT_RING( x ) do {						\
-	if ( RADEON_VERBOSE ) {						\
-		DRM_INFO( "   OUT_RING( 0x%08x ) at 0x%x\n",		\
-			   (unsigned int)(x), write );			\
-	}								\
-	ring[write++] = (x);						\
-	write &= mask;							\
-} while (0)
-
-#define OUT_RING_REG( reg, val ) do {					\
-	OUT_RING( CP_PACKET0( reg, 0 ) );				\
-	OUT_RING( val );						\
-} while (0)
-
-#define OUT_RING_TABLE( tab, sz ) do {					\
-	int _size = (sz);					\
-	int *_tab = (int *)(tab);				\
-								\
-	if (write + _size > mask) {				\
-		int _i = (mask+1) - write;			\
-		_size -= _i;					\
-		while (_i > 0 ) {				\
-			*(int *)(ring + write) = *_tab++;	\
-			write++;				\
-			_i--;					\
-		}						\
-		write = 0;					\
-		_tab += _i;					\
-	}							\
-	while (_size > 0) {					\
-		*(ring + write) = *_tab++;			\
-		write++;					\
-		_size--;					\
-	}							\
-	write &= mask;						\
-} while (0)
-
-/**
- * Copy given number of dwords from drm buffer to the ring buffer.
- */
-#define OUT_RING_DRM_BUFFER(buf, sz) do {				\
-	int _size = (sz) * 4;						\
-	struct drm_buffer *_buf = (buf);				\
-	int _part_size;							\
-	while (_size > 0) {						\
-		_part_size = _size;					\
-									\
-		if (write + _part_size/4 > mask)			\
-			_part_size = ((mask + 1) - write)*4;		\
-									\
-		if (drm_buffer_index(_buf) + _part_size > PAGE_SIZE)	\
-			_part_size = PAGE_SIZE - drm_buffer_index(_buf);\
-									\
-									\
-									\
-		memcpy(ring + write, &_buf->data[drm_buffer_page(_buf)]	\
-			[drm_buffer_index(_buf)], _part_size);		\
-									\
-		_size -= _part_size;					\
-		write = (write + _part_size/4) & mask;			\
-		drm_buffer_advance(_buf, _part_size);			\
-	}								\
-} while (0)
-
-
-#endif				/* CONFIG_DRM_RADEON_UMS */
-
 #endif				/* __RADEON_DRV_H__ */
diff --git a/sys/dev/pci/drm/radeon/radeon_encoders.c b/sys/dev/pci/drm/radeon/radeon_encoders.c
index b1fb98e1e93..c341fb2a5b5 100644
--- a/sys/dev/pci/drm/radeon/radeon_encoders.c
+++ b/sys/dev/pci/drm/radeon/radeon_encoders.c
@@ -23,9 +23,9 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "atom.h"
 
@@ -211,7 +211,7 @@ radeon_link_encoder_connector(struct drm_device *dev)
 		list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 			radeon_encoder = to_radeon_encoder(encoder);
 			if (radeon_encoder->devices & radeon_connector->devices) {
-				drm_mode_connector_attach_encoder(connector, encoder);
+				drm_connector_attach_encoder(connector, encoder);
 				if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
 					radeon_encoder_add_backlight(radeon_encoder, connector);
 			}
diff --git a/sys/dev/pci/drm/radeon/radeon_fb.c b/sys/dev/pci/drm/radeon/radeon_fb.c
index d6ee3157a3f..d43d99ad8b3 100644
--- a/sys/dev/pci/drm/radeon/radeon_fb.c
+++ b/sys/dev/pci/drm/radeon/radeon_fb.c
@@ -23,52 +23,76 @@
  * Authors:
  *     David Airlie
  */
-
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/pm_runtime.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
-#include <dev/pci/drm/drm_fb_helper.h>
+#include <drm/drm_fb_helper.h>
 
+#include <linux/vga_switcheroo.h>
 
 /* object hierarchy -
-   this contains a helper + a radeon fb
-   the helper contains a pointer to radeon framebuffer baseclass.
-*/
+ * this contains a helper + a radeon fb
+ * the helper contains a pointer to radeon framebuffer baseclass.
+ */
 struct radeon_fbdev {
 	struct drm_fb_helper helper;
-	struct radeon_framebuffer rfb;
-	struct list_head fbdev_list;
+	struct drm_framebuffer fb;
 	struct radeon_device *rdev;
 };
 
-#ifdef notyet
+#ifdef __linux__
+static int
+radeonfb_open(struct fb_info *info, int user)
+{
+	struct radeon_fbdev *rfbdev = info->par;
+	struct radeon_device *rdev = rfbdev->rdev;
+	int ret = pm_runtime_get_sync(rdev->ddev->dev);
+	if (ret < 0 && ret != -EACCES) {
+		pm_runtime_mark_last_busy(rdev->ddev->dev);
+		pm_runtime_put_autosuspend(rdev->ddev->dev);
+		return ret;
+	}
+	return 0;
+}
+
+static int
+radeonfb_release(struct fb_info *info, int user)
+{
+	struct radeon_fbdev *rfbdev = info->par;
+	struct radeon_device *rdev = rfbdev->rdev;
+
+	pm_runtime_mark_last_busy(rdev->ddev->dev);
+	pm_runtime_put_autosuspend(rdev->ddev->dev);
+	return 0;
+}
+
 static struct fb_ops radeonfb_ops = {
 	.owner = THIS_MODULE,
-	.fb_check_var = drm_fb_helper_check_var,
-	.fb_set_par = drm_fb_helper_set_par,
+	DRM_FB_HELPER_DEFAULT_OPS,
+	.fb_open = radeonfb_open,
+	.fb_release = radeonfb_release,
 	.fb_fillrect = drm_fb_helper_cfb_fillrect,
 	.fb_copyarea = drm_fb_helper_cfb_copyarea,
 	.fb_imageblit = drm_fb_helper_cfb_imageblit,
-	.fb_pan_display = drm_fb_helper_pan_display,
-	.fb_blank = drm_fb_helper_blank,
-	.fb_setcmap = drm_fb_helper_setcmap,
-	.fb_debug_enter = drm_fb_helper_debug_enter,
-	.fb_debug_leave = drm_fb_helper_debug_leave,
 };
 #endif
 
 void radeondrm_burner_cb(void *);
 
-int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled)
+int radeon_align_pitch(struct radeon_device *rdev, int width, int cpp, bool tiled)
 {
 	int aligned = width;
 	int align_large = (ASIC_IS_AVIVO(rdev)) || tiled;
 	int pitch_mask = 0;
 
-	switch (bpp / 8) {
+	switch (cpp) {
 	case 1:
 		pitch_mask = align_large ? 255 : 127;
 		break;
@@ -83,7 +107,7 @@ int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tile
 
 	aligned += pitch_mask;
 	aligned &= ~pitch_mask;
-	return aligned;
+	return aligned * cpp;
 }
 
 static void radeonfb_destroy_pinned_object(struct drm_gem_object *gobj)
@@ -97,7 +121,7 @@ static void radeonfb_destroy_pinned_object(struct drm_gem_object *gobj)
 		radeon_bo_unpin(rbo);
 		radeon_bo_unreserve(rbo);
 	}
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 }
 
 static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev,
@@ -112,13 +136,13 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev,
 	int ret;
 	int aligned_size, size;
 	int height = mode_cmd->height;
-	u32 bpp, depth;
+	u32 cpp;
 
-	drm_fb_get_bpp_depth(mode_cmd->pixel_format, &depth, &bpp);
+	cpp = drm_format_plane_cpp(mode_cmd->pixel_format, 0);
 
 	/* need to align pitch with crtc limits */
-	mode_cmd->pitches[0] = radeon_align_pitch(rdev, mode_cmd->width, bpp,
-						  fb_tiled) * ((bpp + 1) / 8);
+	mode_cmd->pitches[0] = radeon_align_pitch(rdev, mode_cmd->width, cpp,
+						  fb_tiled);
 
 	if (rdev->family >= CHIP_R600)
 		height = roundup2(mode_cmd->height, 8);
@@ -128,8 +152,7 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev,
 				       RADEON_GEM_DOMAIN_VRAM,
 				       0, true, &gobj);
 	if (ret) {
-		printk(KERN_ERR "failed to allocate framebuffer (%d)\n",
-		       aligned_size);
+		pr_err("failed to allocate framebuffer (%d)\n", aligned_size);
 		return -ENOMEM;
 	}
 	rbo = gem_to_radeon_bo(gobj);
@@ -138,11 +161,11 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev,
 		tiling_flags = RADEON_TILING_MACRO;
 
 #ifdef __BIG_ENDIAN
-	switch (bpp) {
-	case 32:
+	switch (cpp) {
+	case 4:
 		tiling_flags |= RADEON_TILING_SWAP_32BIT;
 		break;
-	case 16:
+	case 2:
 		tiling_flags |= RADEON_TILING_SWAP_16BIT;
 	default:
 		break;
@@ -222,18 +245,18 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 	info = drm_fb_helper_alloc_fbi(helper);
 	if (IS_ERR(info)) {
 		ret = PTR_ERR(info);
-		goto out_unref;
+		goto out;
 	}
 
 	info->par = rfbdev;
 
-	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->rfb, &mode_cmd, gobj);
+	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->fb, &mode_cmd, gobj);
 	if (ret) {
 		DRM_ERROR("failed to initialize framebuffer %d\n", ret);
-		goto out_destroy_fbi;
+		goto out;
 	}
 
-	fb = &rfbdev->rfb.base;
+	fb = &rfbdev->fb;
 
 	/* setup helper */
 	rfbdev->helper.fb = fb;
@@ -243,9 +266,8 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 #ifdef __linux__
 	strcpy(info->fix.id, "radeondrmfb");
 
-	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->depth);
+	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->format->depth);
 
-	info->flags = FBINFO_DEFAULT | FBINFO_CAN_FORCE_OUTPUT;
 	info->fbops = &radeonfb_ops;
 #endif
 
@@ -268,23 +290,23 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 #ifdef __linux__
 	if (info->screen_base == NULL) {
 		ret = -ENOSPC;
-		goto out_destroy_fbi;
+		goto out;
 	}
 
 	DRM_INFO("fb mappable at 0x%lX\n",  info->fix.smem_start);
 #endif
 	DRM_INFO("vram apper at 0x%lX\n",  (unsigned long)rdev->mc.aper_base);
 	DRM_INFO("size %lu\n", (unsigned long)radeon_bo_size(rbo));
-	DRM_INFO("fb depth is %d\n", fb->depth);
+	DRM_INFO("fb depth is %d\n", fb->format->depth);
 	DRM_INFO("   pitch is %d\n", fb->pitches[0]);
 
 	ri->ri_bits = rbo->kptr;
-	ri->ri_depth = fb->bits_per_pixel;
+	ri->ri_depth = fb->format->cpp[0] * 8;
 	ri->ri_stride = fb->pitches[0];
 	ri->ri_width = sizes->fb_width;
 	ri->ri_height = sizes->fb_height;
 
-	switch (fb->pixel_format) {
+	switch (fb->format->format) {
 	case DRM_FORMAT_XRGB8888:
 		ri->ri_rnum = 8;
 		ri->ri_rpos = 16;
@@ -308,16 +330,12 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 #endif
 	return 0;
 
-out_destroy_fbi:
-#ifdef __linux__
-	drm_fb_helper_release_fbi(helper);
-#endif
-out_unref:
+out:
 	if (rbo) {
 
 	}
 	if (fb && ret) {
-		drm_gem_object_unreference(gobj);
+		drm_gem_object_put_unlocked(gobj);
 		drm_framebuffer_unregister_private(fb);
 		drm_framebuffer_cleanup(fb);
 		kfree(fb);
@@ -325,34 +343,26 @@ out_unref:
 	return ret;
 }
 
-void radeon_fb_output_poll_changed(struct radeon_device *rdev)
-{
-	drm_fb_helper_hotplug_event(&rdev->mode_info.rfbdev->helper);
-}
-
 #ifdef notyet
 static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfbdev)
 {
-	struct radeon_framebuffer *rfb = &rfbdev->rfb;
+	struct drm_framebuffer *fb = &rfbdev->fb;
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
-	drm_fb_helper_release_fbi(&rfbdev->helper);
 
-	if (rfb->obj) {
-		radeonfb_destroy_pinned_object(rfb->obj);
-		rfb->obj = NULL;
+	if (fb->obj[0]) {
+		radeonfb_destroy_pinned_object(fb->obj[0]);
+		fb->obj[0] = NULL;
+		drm_framebuffer_unregister_private(fb);
+		drm_framebuffer_cleanup(fb);
 	}
 	drm_fb_helper_fini(&rfbdev->helper);
-	drm_framebuffer_unregister_private(&rfb->base);
-	drm_framebuffer_cleanup(&rfb->base);
 
 	return 0;
 }
 #endif
 
 static const struct drm_fb_helper_funcs radeon_fb_helper_funcs = {
-	.gamma_set = radeon_crtc_fb_gamma_set,
-	.gamma_get = radeon_crtc_fb_gamma_get,
 	.fb_probe = radeonfb_create,
 };
 
@@ -362,9 +372,16 @@ int radeon_fbdev_init(struct radeon_device *rdev)
 	int bpp_sel = 32;
 	int ret;
 
-	/* select 8 bpp console on RN50 or 16MB cards */
-	if (ASIC_IS_RN50(rdev) || rdev->mc.real_vram_size <= (32*1024*1024))
+	/* don't enable fbdev if no connectors */
+	if (list_empty(&rdev->ddev->mode_config.connector_list))
+		return 0;
+
+	/* select 8 bpp console on 8MB cards, or 16 bpp on RN50 or 32MB */
+	if (rdev->mc.real_vram_size <= (8*1024*1024))
 		bpp_sel = 8;
+	else if (ASIC_IS_RN50(rdev) ||
+		 rdev->mc.real_vram_size <= (32*1024*1024))
+		bpp_sel = 16;
 
 	rfbdev = kzalloc(sizeof(struct radeon_fbdev), GFP_KERNEL);
 	if (!rfbdev)
@@ -377,7 +394,6 @@ int radeon_fbdev_init(struct radeon_device *rdev)
 			      &radeon_fb_helper_funcs);
 
 	ret = drm_fb_helper_init(rdev->ddev, &rfbdev->helper,
-				 rdev->num_crtc,
 				 RADEONFB_CONN_LIMIT);
 	if (ret)
 		goto free;
@@ -444,41 +460,31 @@ void radeon_fbdev_fini(struct radeon_device *rdev)
 void radeon_fbdev_set_suspend(struct radeon_device *rdev, int state)
 {
 #ifdef __linux__
-	fb_set_suspend(rdev->mode_info.rfbdev->helper.fbdev, state);
+	if (rdev->mode_info.rfbdev)
+		drm_fb_helper_set_suspend(&rdev->mode_info.rfbdev->helper, state);
 #endif
 }
 
 bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj)
 {
-	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj))
+	if (!rdev->mode_info.rfbdev)
+		return false;
+
+	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->fb.obj[0]))
 		return true;
 	return false;
 }
 
 void radeon_fb_add_connector(struct radeon_device *rdev, struct drm_connector *connector)
 {
-	drm_fb_helper_add_one_connector(&rdev->mode_info.rfbdev->helper, connector);
+	if (rdev->mode_info.rfbdev)
+		drm_fb_helper_add_one_connector(&rdev->mode_info.rfbdev->helper, connector);
 }
 
 void radeon_fb_remove_connector(struct radeon_device *rdev, struct drm_connector *connector)
 {
-	drm_fb_helper_remove_one_connector(&rdev->mode_info.rfbdev->helper, connector);
-}
-
-void radeon_fbdev_restore_mode(struct radeon_device *rdev)
-{
-	struct radeon_fbdev *rfbdev = rdev->mode_info.rfbdev;
-	struct drm_fb_helper *fb_helper;
-	int ret;
-
-	if (!rfbdev)
-		return;
-
-	fb_helper = &rfbdev->helper;
-
-	ret = drm_fb_helper_restore_fbdev_mode_unlocked(fb_helper);
-	if (ret)
-		DRM_DEBUG("failed to restore crtc mode\n");
+	if (rdev->mode_info.rfbdev)
+		drm_fb_helper_remove_one_connector(&rdev->mode_info.rfbdev->helper, connector);
 }
 
 void
diff --git a/sys/dev/pci/drm/radeon/radeon_fence.c b/sys/dev/pci/drm/radeon/radeon_fence.c
index 6fd9cf59975..1d1904fa15d 100644
--- a/sys/dev/pci/drm/radeon/radeon_fence.c
+++ b/sys/dev/pci/drm/radeon/radeon_fence.c
@@ -28,8 +28,13 @@
  *    Jerome Glisse <glisse@freedesktop.org>
  *    Dave Airlie
  */
-#include <dev/pci/drm/drm_linux.h>
-#include <dev/pci/drm/drmP.h>
+#include <linux/seq_file.h>
+#include <linux/atomic.h>
+#include <linux/wait.h>
+#include <linux/kref.h>
+#include <linux/slab.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_trace.h"
@@ -125,7 +130,7 @@ int radeon_fence_emit(struct radeon_device *rdev,
 		      struct radeon_fence **fence,
 		      int ring)
 {
-	u64 seq = ++rdev->fence_drv[ring].sync_seq[ring];
+	u64 seq;
 
 	/* we are protected by the ring emission mutex */
 	*fence = kmalloc(sizeof(struct radeon_fence), GFP_KERNEL);
@@ -133,11 +138,13 @@ int radeon_fence_emit(struct radeon_device *rdev,
 		return -ENOMEM;
 	}
 	(*fence)->rdev = rdev;
-	(*fence)->seq = seq;
+	(*fence)->seq = seq = ++rdev->fence_drv[ring].sync_seq[ring];
 	(*fence)->ring = ring;
 	(*fence)->is_vm_update = false;
-	fence_init(&(*fence)->base, &radeon_fence_ops,
-		   &rdev->fence_queue.lock, rdev->fence_context + ring, seq);
+	dma_fence_init(&(*fence)->base, &radeon_fence_ops,
+		       &rdev->fence_queue.lock,
+		       rdev->fence_context + ring,
+		       seq);
 	radeon_fence_ring_emit(rdev, ring, *fence);
 	trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq);
 	radeon_fence_schedule_check(rdev, ring);
@@ -151,7 +158,7 @@ int radeon_fence_emit(struct radeon_device *rdev,
  * for the fence locking itself, so unlocked variants are used for
  * fence_signal, and remove_wait_queue.
  */
-static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int flags, void *key)
+static int radeon_fence_check_signaled(wait_queue_entry_t *wait, unsigned mode, int flags, void *key)
 {
 	struct radeon_fence *fence;
 	u64 seq;
@@ -164,18 +171,18 @@ static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int fl
 	 */
 	seq = atomic64_read(&fence->rdev->fence_drv[fence->ring].last_seq);
 	if (seq >= fence->seq) {
-		int ret = fence_signal_locked(&fence->base);
+		int ret = dma_fence_signal_locked(&fence->base);
 
 		if (!ret)
-			FENCE_TRACE(&fence->base, "signaled from irq context\n");
+			DMA_FENCE_TRACE(&fence->base, "signaled from irq context\n");
 		else
-			FENCE_TRACE(&fence->base, "was already signaled\n");
+			DMA_FENCE_TRACE(&fence->base, "was already signaled\n");
 
 		radeon_irq_kms_sw_irq_put(fence->rdev, fence->ring);
 		__remove_wait_queue(&fence->rdev->fence_queue, &fence->fence_wake);
-		fence_put(&fence->base);
+		dma_fence_put(&fence->base);
 	} else
-		FENCE_TRACE(&fence->base, "pending\n");
+		DMA_FENCE_TRACE(&fence->base, "pending\n");
 	return 0;
 }
 
@@ -346,7 +353,7 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev,
 	return false;
 }
 
-static bool radeon_fence_is_signaled(struct fence *f)
+static bool radeon_fence_is_signaled(struct dma_fence *f)
 {
 	struct radeon_fence *fence = to_radeon_fence(f);
 	struct radeon_device *rdev = fence->rdev;
@@ -376,7 +383,7 @@ static bool radeon_fence_is_signaled(struct fence *f)
  * to fence_queue that checks if this fence is signaled, and if so it
  * signals the fence and removes itself.
  */
-static bool radeon_fence_enable_signaling(struct fence *f)
+static bool radeon_fence_enable_signaling(struct dma_fence *f)
 {
 	struct radeon_fence *fence = to_radeon_fence(f);
 	struct radeon_device *rdev = fence->rdev;
@@ -409,9 +416,9 @@ static bool radeon_fence_enable_signaling(struct fence *f)
 	fence->fence_wake.private = NULL;
 	fence->fence_wake.func = radeon_fence_check_signaled;
 	__add_wait_queue(&rdev->fence_queue, &fence->fence_wake);
-	fence_get(f);
+	dma_fence_get(f);
 
-	FENCE_TRACE(&fence->base, "armed on ring %i!\n", fence->ring);
+	DMA_FENCE_TRACE(&fence->base, "armed on ring %i!\n", fence->ring);
 	return true;
 }
 
@@ -431,9 +438,9 @@ bool radeon_fence_signaled(struct radeon_fence *fence)
 	if (radeon_fence_seq_signaled(fence->rdev, fence->seq, fence->ring)) {
 		int ret;
 
-		ret = fence_signal(&fence->base);
+		ret = dma_fence_signal(&fence->base);
 		if (!ret)
-			FENCE_TRACE(&fence->base, "signaled from radeon_fence_signaled\n");
+			DMA_FENCE_TRACE(&fence->base, "signaled from radeon_fence_signaled\n");
 		return true;
 	}
 	return false;
@@ -522,7 +529,7 @@ static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
 }
 
 /**
- * radeon_fence_wait - wait for a fence to signal
+ * radeon_fence_wait_timeout - wait for a fence to signal with timeout
  *
  * @fence: radeon fence object
  * @intr: use interruptible sleep
@@ -530,12 +537,15 @@ static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev,
  * Wait for the requested fence to signal (all asics).
  * @intr selects whether to use interruptable (true) or non-interruptable
  * (false) sleep when waiting for the fence.
- * Returns 0 if the fence has passed, error for all other cases.
+ * @timeout: maximum time to wait, or MAX_SCHEDULE_TIMEOUT for infinite wait
+ * Returns remaining time if the sequence number has passed, 0 when
+ * the wait timeout, or an error for all other cases.
  */
-int radeon_fence_wait(struct radeon_fence *fence, bool intr)
+long radeon_fence_wait_timeout(struct radeon_fence *fence, bool intr, long timeout)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
 	long r;
+	int r_sig;
 
 	/*
 	 * This function should not be called on !radeon fences.
@@ -544,18 +554,39 @@ int radeon_fence_wait(struct radeon_fence *fence, bool intr)
 	 * exclusive_lock is not held in that case.
 	 */
 	if (WARN_ON_ONCE(!to_radeon_fence(&fence->base)))
-		return fence_wait(&fence->base, intr);
+		return dma_fence_wait(&fence->base, intr);
 
 	seq[fence->ring] = fence->seq;
-	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, MAX_SCHEDULE_TIMEOUT);
-	if (r < 0) {
+	r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, timeout);
+	if (r <= 0) {
 		return r;
 	}
 
-	r = fence_signal(&fence->base);
-	if (!r)
-		FENCE_TRACE(&fence->base, "signaled from fence_wait\n");
-	return 0;
+	r_sig = dma_fence_signal(&fence->base);
+	if (!r_sig)
+		DMA_FENCE_TRACE(&fence->base, "signaled from fence_wait\n");
+	return r;
+}
+
+/**
+ * radeon_fence_wait - wait for a fence to signal
+ *
+ * @fence: radeon fence object
+ * @intr: use interruptible sleep
+ *
+ * Wait for the requested fence to signal (all asics).
+ * @intr selects whether to use interruptable (true) or non-interruptable
+ * (false) sleep when waiting for the fence.
+ * Returns 0 if the fence has passed, error for all other cases.
+ */
+int radeon_fence_wait(struct radeon_fence *fence, bool intr)
+{
+	long r = radeon_fence_wait_timeout(fence, intr, MAX_SCHEDULE_TIMEOUT);
+	if (r > 0) {
+		return 0;
+	} else {
+		return r;
+	}
 }
 
 /**
@@ -668,7 +699,7 @@ int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
  */
 struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence)
 {
-	fence_get(&fence->base);
+	dma_fence_get(&fence->base);
 	return fence;
 }
 
@@ -685,7 +716,7 @@ void radeon_fence_unref(struct radeon_fence **fence)
 
 	*fence = NULL;
 	if (tmp) {
-		fence_put(&tmp->base);
+		dma_fence_put(&tmp->base);
 	}
 }
 
@@ -999,12 +1030,12 @@ int radeon_debugfs_fence_init(struct radeon_device *rdev)
 #endif
 }
 
-static const char *radeon_fence_get_driver_name(struct fence *fence)
+static const char *radeon_fence_get_driver_name(struct dma_fence *fence)
 {
 	return "radeon";
 }
 
-static const char *radeon_fence_get_timeline_name(struct fence *f)
+static const char *radeon_fence_get_timeline_name(struct dma_fence *f)
 {
 	struct radeon_fence *fence = to_radeon_fence(f);
 	switch (fence->ring) {
@@ -1022,23 +1053,24 @@ static const char *radeon_fence_get_timeline_name(struct fence *f)
 
 static inline bool radeon_test_signaled(struct radeon_fence *fence)
 {
-	return test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags);
+	return test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->base.flags);
 }
 
 struct radeon_wait_cb {
-	struct fence_cb base;
+	struct dma_fence_cb base;
 	void *task;
 };
 
 static void
-radeon_fence_wait_cb(struct fence *fence, struct fence_cb *cb)
+radeon_fence_wait_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
 	struct radeon_wait_cb *wait =
 		container_of(cb, struct radeon_wait_cb, base);
+
 	wake_up_process(wait->task);
 }
 
-static signed long radeon_fence_default_wait(struct fence *f, bool intr,
+static signed long radeon_fence_default_wait(struct dma_fence *f, bool intr,
 					     signed long t)
 {
 	struct radeon_fence *fence = to_radeon_fence(f);
@@ -1047,7 +1079,7 @@ static signed long radeon_fence_default_wait(struct fence *f, bool intr,
 
 	cb.task = curproc;
 
-	if (fence_add_callback(f, &cb.base, radeon_fence_wait_cb))
+	if (dma_fence_add_callback(f, &cb.base, radeon_fence_wait_cb))
 		return t;
 
 	while (t > 0) {
@@ -1076,12 +1108,12 @@ static signed long radeon_fence_default_wait(struct fence *f, bool intr,
 	}
 
 	__set_current_state(TASK_RUNNING);
-	fence_remove_callback(f, &cb.base);
+	dma_fence_remove_callback(f, &cb.base);
 
 	return t;
 }
 
-const struct fence_ops radeon_fence_ops = {
+const struct dma_fence_ops radeon_fence_ops = {
 	.get_driver_name = radeon_fence_get_driver_name,
 	.get_timeline_name = radeon_fence_get_timeline_name,
 	.enable_signaling = radeon_fence_enable_signaling,
diff --git a/sys/dev/pci/drm/radeon/radeon_gart.c b/sys/dev/pci/drm/radeon/radeon_gart.c
index 4ab155bd0c9..3c1a6191d72 100644
--- a/sys/dev/pci/drm/radeon/radeon_gart.c
+++ b/sys/dev/pci/drm/radeon/radeon_gart.c
@@ -25,8 +25,11 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
+#ifdef CONFIG_X86
+#include <asm/set_memory.h>
+#endif
 #include "radeon.h"
 
 /*
@@ -344,13 +347,14 @@ int radeon_gart_init(struct radeon_device *rdev)
 	DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n",
 		 rdev->gart.num_cpu_pages, rdev->gart.num_gpu_pages);
 	/* Allocate pages table */
-	rdev->gart.pages = vzalloc(sizeof(void *) * rdev->gart.num_cpu_pages);
+	rdev->gart.pages = vzalloc(array_size(sizeof(void *),
+				   rdev->gart.num_cpu_pages));
 	if (rdev->gart.pages == NULL) {
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
 	}
-	rdev->gart.pages_entry = vmalloc(sizeof(uint64_t) *
-					 rdev->gart.num_gpu_pages);
+	rdev->gart.pages_entry = vmalloc(array_size(sizeof(uint64_t),
+						    rdev->gart.num_gpu_pages));
 	if (rdev->gart.pages_entry == NULL) {
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
diff --git a/sys/dev/pci/drm/radeon/radeon_gem.c b/sys/dev/pci/drm/radeon/radeon_gem.c
index c9a4478b254..0bc4c919299 100644
--- a/sys/dev/pci/drm/radeon/radeon_gem.c
+++ b/sys/dev/pci/drm/radeon/radeon_gem.c
@@ -25,8 +25,8 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 void radeon_gem_object_free(struct drm_gem_object *gobj)
@@ -34,10 +34,6 @@ void radeon_gem_object_free(struct drm_gem_object *gobj)
 	struct radeon_bo *robj = gem_to_radeon_bo(gobj);
 
 	if (robj) {
-#ifdef notyet
-		if (robj->gem_base.import_attach)
-			drm_prime_gem_destroy(&robj->gem_base, robj->tbo.sg);
-#endif
 		radeon_mn_unregister(robj);
 		radeon_bo_unref(&robj);
 	}
@@ -108,7 +104,7 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 	}
 	if (!domain) {
 		/* Do nothings */
-		printk(KERN_WARNING "Set domain without domain !\n");
+		pr_warn("Set domain without domain !\n");
 		return 0;
 	}
 	if (domain == RADEON_GEM_DOMAIN_CPU) {
@@ -118,10 +114,14 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 			r = -EBUSY;
 
 		if (r < 0 && r != -EINTR) {
-			printk(KERN_ERR "Failed to wait for object: %li\n", r);
+			pr_err("Failed to wait for object: %li\n", r);
 			return r;
 		}
 	}
+	if (domain == RADEON_GEM_DOMAIN_VRAM && robj->prime_shared_count) {
+		/* A BO that is associated with a dma-buf cannot be sensibly migrated to VRAM */
+		return -EINVAL;
+	}
 	return 0;
 }
 
@@ -222,8 +222,8 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data,
 
 	man = &rdev->mman.bdev.man[TTM_PL_VRAM];
 
-	args->vram_size = rdev->mc.real_vram_size;
-	args->vram_visible = (u64)man->size << PAGE_SHIFT;
+	args->vram_size = (u64)man->size << PAGE_SHIFT;
+	args->vram_visible = rdev->mc.visible_vram_size;
 	args->vram_visible -= rdev->vram_pin_size;
 	args->gart_size = rdev->mc.gtt_size;
 	args->gart_size -= rdev->gart_pin_size;
@@ -269,7 +269,7 @@ int radeon_gem_create_ioctl(struct drm_device *dev, void *data,
 	}
 	r = drm_gem_handle_create(filp, gobj, &handle);
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	if (r) {
 		up_read(&rdev->exclusive_lock);
 		r = radeon_gem_handle_lockup(rdev, r);
@@ -285,6 +285,7 @@ int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
 {
 	return -ENOSYS;
 #ifdef notyet
+	struct ttm_operation_ctx ctx = { true, false };
 	struct radeon_device *rdev = dev->dev_private;
 	struct drm_radeon_gem_userptr *args = data;
 	struct drm_gem_object *gobj;
@@ -343,7 +344,7 @@ int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
 		}
 
 		radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_GTT);
-		r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 		radeon_bo_unreserve(bo);
 		up_read(&current->mm->mmap_sem);
 		if (r)
@@ -352,7 +353,7 @@ int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
 
 	r = drm_gem_handle_create(filp, gobj, &handle);
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	if (r)
 		goto handle_lockup;
 
@@ -361,7 +362,7 @@ int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data,
 	return 0;
 
 release_object:
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 
 handle_lockup:
 	up_read(&rdev->exclusive_lock);
@@ -387,7 +388,7 @@ int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	down_read(&rdev->exclusive_lock);
 
 	/* just do a BO wait for now */
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL) {
 		up_read(&rdev->exclusive_lock);
 		return -ENOENT;
@@ -396,7 +397,7 @@ int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 
 	r = radeon_gem_set_domain(gobj, args->read_domains, args->write_domain);
 
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	up_read(&rdev->exclusive_lock);
 	r = radeon_gem_handle_lockup(robj->rdev, r);
 	return r;
@@ -409,17 +410,17 @@ int radeon_mode_dumb_mmap(struct drm_file *filp,
 	struct drm_gem_object *gobj;
 	struct radeon_bo *robj;
 
-	gobj = drm_gem_object_lookup(dev, filp, handle);
+	gobj = drm_gem_object_lookup(filp, handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
 	robj = gem_to_radeon_bo(gobj);
 	if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) {
-		drm_gem_object_unreference_unlocked(gobj);
+		drm_gem_object_put_unlocked(gobj);
 		return -EPERM;
 	}
 	*offset_p = radeon_bo_mmap_offset(robj);
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return 0;
 }
 
@@ -440,7 +441,7 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data,
 	int r;
 	uint32_t cur_placement = 0;
 
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
@@ -452,9 +453,9 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data,
 	else
 		r = 0;
 
-	cur_placement = ACCESS_ONCE(robj->tbo.mem.mem_type);
+	cur_placement = READ_ONCE(robj->tbo.mem.mem_type);
 	args->domain = radeon_mem_type_to_domain(cur_placement);
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return r;
 }
 
@@ -469,7 +470,7 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 	uint32_t cur_placement = 0;
 	long ret;
 
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
@@ -482,11 +483,11 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 		r = ret;
 
 	/* Flush HDP cache via MMIO if necessary */
-	cur_placement = ACCESS_ONCE(robj->tbo.mem.mem_type);
+	cur_placement = READ_ONCE(robj->tbo.mem.mem_type);
 	if (rdev->asic->mmio_hdp_flush &&
 	    radeon_mem_type_to_domain(cur_placement) == RADEON_GEM_DOMAIN_VRAM)
 		robj->rdev->asic->mmio_hdp_flush(rdev);
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	r = radeon_gem_handle_lockup(rdev, r);
 	return r;
 }
@@ -500,12 +501,12 @@ int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
 	int r = 0;
 
 	DRM_DEBUG("%d \n", args->handle);
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL)
 		return -ENOENT;
 	robj = gem_to_radeon_bo(gobj);
 	r = radeon_bo_set_tiling_flags(robj, args->tiling_flags, args->pitch);
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return r;
 }
 
@@ -518,7 +519,7 @@ int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
 	int r = 0;
 
 	DRM_DEBUG("\n");
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL)
 		return -ENOENT;
 	rbo = gem_to_radeon_bo(gobj);
@@ -528,7 +529,7 @@ int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data,
 	radeon_bo_get_tiling_flags(rbo, &args->tiling_flags, &args->pitch);
 	radeon_bo_unreserve(rbo);
 out:
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return r;
 }
 
@@ -588,7 +589,7 @@ error_unreserve:
 	ttm_eu_backoff_reservation(&ticket, &list);
 
 error_free:
-	drm_free_large(vm_bos);
+	kvfree(vm_bos);
 
 	if (r && r != -ERESTARTSYS)
 		DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
@@ -653,7 +654,7 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL) {
 		args->operation = RADEON_VA_RESULT_ERROR;
 		return -ENOENT;
@@ -662,13 +663,14 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 	r = radeon_bo_reserve(rbo, false);
 	if (r) {
 		args->operation = RADEON_VA_RESULT_ERROR;
-		drm_gem_object_unreference_unlocked(gobj);
+		drm_gem_object_put_unlocked(gobj);
 		return r;
 	}
 	bo_va = radeon_vm_bo_find(&fpriv->vm, rbo);
 	if (!bo_va) {
 		args->operation = RADEON_VA_RESULT_ERROR;
-		drm_gem_object_unreference_unlocked(gobj);
+		radeon_bo_unreserve(rbo);
+		drm_gem_object_put_unlocked(gobj);
 		return -ENOENT;
 	}
 
@@ -695,7 +697,7 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 		args->operation = RADEON_VA_RESULT_ERROR;
 	}
 out:
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return r;
 }
 
@@ -707,7 +709,7 @@ int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
 	struct radeon_bo *robj;
 	int r;
 
-	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	gobj = drm_gem_object_lookup(filp, args->handle);
 	if (gobj == NULL) {
 		return -ENOENT;
 	}
@@ -736,7 +738,7 @@ int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
 
 	radeon_bo_unreserve(robj);
 out:
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	return r;
 }
 
@@ -749,7 +751,8 @@ int radeon_mode_dumb_create(struct drm_file *file_priv,
 	uint32_t handle;
 	int r;
 
-	args->pitch = radeon_align_pitch(rdev, args->width, args->bpp, 0) * ((args->bpp + 1) / 8);
+	args->pitch = radeon_align_pitch(rdev, args->width,
+					 DIV_ROUND_UP(args->bpp, 8), 0);
 	args->size = args->pitch * args->height;
 	args->size = roundup2(args->size, PAGE_SIZE);
 
@@ -761,7 +764,7 @@ int radeon_mode_dumb_create(struct drm_file *file_priv,
 
 	r = drm_gem_handle_create(file_priv, gobj, &handle);
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_unreference_unlocked(gobj);
+	drm_gem_object_put_unlocked(gobj);
 	if (r) {
 		return r;
 	}
diff --git a/sys/dev/pci/drm/radeon/radeon_i2c.c b/sys/dev/pci/drm/radeon/radeon_i2c.c
index 2794f28b9d0..82933ebf6aa 100644
--- a/sys/dev/pci/drm/radeon/radeon_i2c.c
+++ b/sys/dev/pci/drm/radeon/radeon_i2c.c
@@ -23,10 +23,11 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
+#include <linux/export.h>
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_edid.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_edid.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "atom.h"
 
diff --git a/sys/dev/pci/drm/radeon/radeon_ib.c b/sys/dev/pci/drm/radeon/radeon_ib.c
index af089965ac1..92ce0e533bc 100644
--- a/sys/dev/pci/drm/radeon/radeon_ib.c
+++ b/sys/dev/pci/drm/radeon/radeon_ib.c
@@ -26,7 +26,7 @@
  *          Jerome Glisse
  *          Christian König
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 
 /*
@@ -274,7 +274,7 @@ int radeon_ib_ring_tests(struct radeon_device *rdev)
 			if (i == RADEON_RING_TYPE_GFX_INDEX) {
 				/* oh, oh, that's really bad */
 				DRM_ERROR("radeon: failed testing IB on GFX ring (%d).\n", r);
-		                rdev->accel_working = false;
+				rdev->accel_working = false;
 				return r;
 
 			} else {
@@ -304,7 +304,7 @@ static int radeon_debugfs_sa_info(struct seq_file *m, void *data)
 }
 
 static struct drm_info_list radeon_debugfs_sa_list[] = {
-        {"radeon_sa_info", &radeon_debugfs_sa_info, 0, NULL},
+	{"radeon_sa_info", &radeon_debugfs_sa_info, 0, NULL},
 };
 
 #endif
diff --git a/sys/dev/pci/drm/radeon/radeon_ioc32.c b/sys/dev/pci/drm/radeon/radeon_ioc32.c
index 5c92eff554e..98492552c8f 100644
--- a/sys/dev/pci/drm/radeon/radeon_ioc32.c
+++ b/sys/dev/pci/drm/radeon/radeon_ioc32.c
@@ -28,8 +28,8 @@
  * IN THE SOFTWARE.
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_drv.h"
 
 typedef struct drm_radeon_init32 {
diff --git a/sys/dev/pci/drm/radeon/radeon_irq_kms.c b/sys/dev/pci/drm/radeon/radeon_irq_kms.c
index f75204ddb25..fb8f85124ce 100644
--- a/sys/dev/pci/drm/radeon/radeon_irq_kms.c
+++ b/sys/dev/pci/drm/radeon/radeon_irq_kms.c
@@ -25,13 +25,14 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "atom.h"
 
+#include <linux/pm_runtime.h>
 
 #define RADEON_WAIT_IDLE_TIMEOUT 200
 
@@ -87,10 +88,8 @@ static void radeon_hotplug_work_func(struct work_struct *work)
 		return;
 
 	mutex_lock(&mode_config->mutex);
-	if (mode_config->num_connector) {
-		list_for_each_entry(connector, &mode_config->connector_list, head)
-			radeon_connector_hotplug(connector);
-	}
+	list_for_each_entry(connector, &mode_config->connector_list, head)
+		radeon_connector_hotplug(connector);
 	mutex_unlock(&mode_config->mutex);
 	/* Just fire off a uevent and let userspace tell us what to do */
 	drm_helper_hpd_irq_event(dev);
@@ -105,10 +104,8 @@ static void radeon_dp_work_func(struct work_struct *work)
 	struct drm_connector *connector;
 
 	/* this should take a mutex */
-	if (mode_config->num_connector) {
-		list_for_each_entry(connector, &mode_config->connector_list, head)
-			radeon_connector_hotplug(connector);
-	}
+	list_for_each_entry(connector, &mode_config->connector_list, head)
+		radeon_connector_hotplug(connector);
 }
 /**
  * radeon_driver_irq_preinstall_kms - drm irq preinstall callback
@@ -219,12 +216,10 @@ bool radeon_msi_ok(struct radeon_device *rdev)
 	 * of address for "64-bit" MSIs which breaks on some platforms, notably
 	 * IBM POWER servers, so we limit them
 	 */
-#ifdef notyet
 	if (rdev->family < CHIP_BONAIRE) {
 		dev_info(rdev->dev, "radeon: MSI limited to 32-bit\n");
 		rdev->pdev->no_64bit_msi = 1;
 	}
-#endif
 
 	/* force MSI on */
 	if (radeon_msi == 1)
@@ -291,10 +286,15 @@ int radeon_irq_kms_init(struct radeon_device *rdev)
 	int r = 0;
 
 	mtx_init(&rdev->irq.lock, IPL_TTY);
+
+	/* Disable vblank irqs aggressively for power-saving */
+	rdev->ddev->vblank_disable_immediate = true;
+
 	r = drm_vblank_init(rdev->ddev, rdev->num_crtc);
 	if (r) {
 		return r;
 	}
+
 #ifdef notyet
 	/* enable msi */
 	rdev->msi_enabled = 0;
@@ -333,7 +333,6 @@ int radeon_irq_kms_init(struct radeon_device *rdev)
  */
 void radeon_irq_kms_fini(struct radeon_device *rdev)
 {
-	drm_vblank_cleanup(rdev->ddev);
 	if (rdev->irq.installed) {
 		drm_irq_uninstall(rdev->ddev);
 		rdev->irq.installed = false;
@@ -547,3 +546,38 @@ void radeon_irq_kms_disable_hpd(struct radeon_device *rdev, unsigned hpd_mask)
 	spin_unlock_irqrestore(&rdev->irq.lock, irqflags);
 }
 
+/**
+ * radeon_irq_kms_update_int_n - helper for updating interrupt enable registers
+ *
+ * @rdev: radeon device pointer
+ * @reg: the register to write to enable/disable interrupts
+ * @mask: the mask that enables the interrupts
+ * @enable: whether to enable or disable the interrupt register
+ * @name: the name of the interrupt register to print to the kernel log
+ * @num: the number of the interrupt register to print to the kernel log
+ *
+ * Helper for updating the enable state of interrupt registers. Checks whether
+ * or not the interrupt matches the enable state we want. If it doesn't, then
+ * we update it and print a debugging message to the kernel log indicating the
+ * new state of the interrupt register.
+ *
+ * Used for updating sequences of interrupts registers like HPD1, HPD2, etc.
+ */
+void radeon_irq_kms_set_irq_n_enabled(struct radeon_device *rdev,
+				      u32 reg, u32 mask,
+				      bool enable, const char *name, unsigned n)
+{
+	u32 tmp = RREG32(reg);
+
+	/* Interrupt state didn't change */
+	if (!!(tmp & mask) == enable)
+		return;
+
+	if (enable) {
+		DRM_DEBUG("%s%d interrupts enabled\n", name, n);
+		WREG32(reg, tmp |= mask);
+	} else {
+		DRM_DEBUG("%s%d interrupts disabled\n", name, n);
+		WREG32(reg, tmp & ~mask);
+	}
+}
diff --git a/sys/dev/pci/drm/radeon/radeon_kfd.c b/sys/dev/pci/drm/radeon/radeon_kfd.c
index 101f7b3302e..b07d87369a1 100644
--- a/sys/dev/pci/drm/radeon/radeon_kfd.c
+++ b/sys/dev/pci/drm/radeon/radeon_kfd.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "cik_reg.h"
diff --git a/sys/dev/pci/drm/radeon/radeon_kms.c b/sys/dev/pci/drm/radeon/radeon_kms.c
index f9e72e449cd..21ef9aae2c2 100644
--- a/sys/dev/pci/drm/radeon/radeon_kms.c
+++ b/sys/dev/pci/drm/radeon/radeon_kms.c
@@ -25,12 +25,15 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_fb_helper.h>
+#include <drm/drmP.h>
+#include <drm/drm_fb_helper.h>
 #include "radeon.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "radeon_asic.h"
 
+#include <linux/vga_switcheroo.h>
+#include <linux/slab.h>
+#include <linux/pm_runtime.h>
 #include "radeon_kfd.h"
 
 #if defined(CONFIG_VGA_SWITCHEROO)
@@ -316,18 +319,20 @@ radeondrm_setcolor(void *v, u_int index, u_int8_t r, u_int8_t g, u_int8_t b)
 {
 	struct sunfb *sf = v;
 	struct radeon_device *rdev = sf->sf_ro.ri_hw;
-	struct drm_fb_helper *fb_helper = (void *)rdev->mode_info.rfbdev;
 	u_int16_t red, green, blue;
-	struct drm_crtc *crtc;
-	int i;
+	struct radeon_crtc *radeon_crtc;
+	int i, crtc;
 
-	for (i = 0; i < fb_helper->crtc_count; i++) {
-		crtc = fb_helper->crtc_info[i].mode_set.crtc;
+	for (crtc = 0; i < rdev->num_crtc; crtc++) {
+		radeon_crtc = rdev->mode_info.crtcs[crtc];
 
 		red = (r << 8) | r;
 		green = (g << 8) | g;
 		blue = (b << 8) | b;
-		fb_helper->funcs->gamma_set(crtc, red, green, blue, index);
+
+		radeon_crtc->lut_r[index] = red >> 6;
+		radeon_crtc->lut_g[index] = green >> 6;
+		radeon_crtc->lut_b[index] = blue >> 6;
 	}
 }
 #endif
@@ -351,6 +356,31 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 	struct radeon_device *rdev;
 	int r, acpi_status;
 
+	if (!radeon_si_support) {
+		switch (flags & RADEON_FAMILY_MASK) {
+		case CHIP_TAHITI:
+		case CHIP_PITCAIRN:
+		case CHIP_VERDE:
+		case CHIP_OLAND:
+		case CHIP_HAINAN:
+			dev_info(dev->dev,
+				 "SI support disabled by module param\n");
+			return -ENODEV;
+		}
+	}
+	if (!radeon_cik_support) {
+		switch (flags & RADEON_FAMILY_MASK) {
+		case CHIP_KAVERI:
+		case CHIP_BONAIRE:
+		case CHIP_HAWAII:
+		case CHIP_KABINI:
+		case CHIP_MULLINS:
+			dev_info(dev->dev,
+				 "CIK support disabled by module param\n");
+			return -ENODEV;
+		}
+	}
+
 	rdev = kzalloc(sizeof(struct radeon_device), GFP_KERNEL);
 	if (rdev == NULL) {
 		return -ENOMEM;
@@ -368,7 +398,8 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 
 	if ((radeon_runtime_pm != 0) &&
 	    radeon_has_atpx() &&
-	    ((flags & RADEON_IS_IGP) == 0))
+	    ((flags & RADEON_IS_IGP) == 0) &&
+	    !pci_is_thunderbolt_attached(dev->pdev))
 		flags |= RADEON_IS_PX;
 
 	/* radeon_device_init should report only fatal error
@@ -401,12 +432,8 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 				"Error during ACPI methods call\n");
 	}
 
-#ifdef notyet
-	radeon_kfd_device_probe(rdev);
-	radeon_kfd_device_init(rdev);
-#endif
-
 	if (radeon_is_px(dev)) {
+		dev_pm_set_driver_flags(dev->dev, DPM_FLAG_NEVER_SKIP);
 		pm_runtime_use_autosuspend(dev->dev);
 		pm_runtime_set_autosuspend_delay(dev->dev, 5000);
 		pm_runtime_set_active(dev->dev);
@@ -601,8 +628,8 @@ radeondrm_attach_kms(struct device *parent, struct device *self, void *aux)
 	kms_driver.num_ioctls = radeon_max_kms_ioctl;
 	kms_driver.driver_features |= DRIVER_MODESET;
 
-	dev = (struct drm_device *)drm_attach_pci(&kms_driver, pa, is_agp,
-	    rdev->console, self);
+	dev = drm_attach_pci(&kms_driver, pa, is_agp, rdev->console,
+	    self, NULL);
 	rdev->ddev = dev;
 	rdev->pdev = dev->pdev;
 
@@ -805,6 +832,11 @@ radeondrm_attachhook(struct device *self)
 
 	config_found_sm(&rdev->self, &aa, wsemuldisplaydevprint,
 	    wsemuldisplaydevsubmatch);
+
+	/*
+	 * in linux via radeon_pci_probe -> drm_get_pci_dev -> drm_dev_register
+	 */
+	drm_dev_register(rdev->ddev, rdev->flags);
 }
 }
 
@@ -820,7 +852,7 @@ radeondrm_activate_kms(struct device *self, int act)
 	switch (act) {
 	case DVACT_QUIESCE:
 		rv = config_activate_children(self, act);
-		radeon_suspend_kms(rdev->ddev, true, true);
+		radeon_suspend_kms(rdev->ddev, true, true, false);
 		break;
 	case DVACT_SUSPEND:
 		break;
@@ -1278,13 +1310,11 @@ static int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file
  */
 void radeon_driver_lastclose_kms(struct drm_device *dev)
 {
-	struct radeon_device *rdev = dev->dev_private;
-
 #ifdef __sparc64__
+	struct radeon_device *rdev = dev->dev_private;
 	fbwscons_setcolormap(&rdev->sf, radeondrm_setcolor);
 #endif
-	if (rdev->mode_info.mode_config_initialized)
-		radeon_fbdev_restore_mode(rdev);
+	drm_fb_helper_lastclose(dev);
 	vga_switcheroo_process_delayed_switch();
 }
 
@@ -1312,11 +1342,11 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 	if (rdev->family >= CHIP_CAYMAN) {
 		struct radeon_fpriv *fpriv;
 		struct radeon_vm *vm;
-		int r;
 
 		fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
 		if (unlikely(!fpriv)) {
-			return -ENOMEM;
+			r = -ENOMEM;
+			goto out_suspend;
 		}
 
 		if (rdev->accel_working) {
@@ -1324,14 +1354,14 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 			r = radeon_vm_init(rdev, vm);
 			if (r) {
 				kfree(fpriv);
-				return r;
+				goto out_suspend;
 			}
 
 			r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
 			if (r) {
 				radeon_vm_fini(rdev, vm);
 				kfree(fpriv);
-				return r;
+				goto out_suspend;
 			}
 
 			/* map the ib pool buffer read only into
@@ -1345,15 +1375,16 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 			if (r) {
 				radeon_vm_fini(rdev, vm);
 				kfree(fpriv);
-				return r;
+				goto out_suspend;
 			}
 		}
 		file_priv->driver_priv = fpriv;
 	}
 
+out_suspend:
 	pm_runtime_mark_last_busy(dev->dev);
 	pm_runtime_put_autosuspend(dev->dev);
-	return 0;
+	return r;
 }
 
 /**
@@ -1362,13 +1393,26 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
  * @dev: drm dev pointer
  * @file_priv: drm file
  *
- * On device post close, tear down vm on cayman+ (all asics).
+ * On device close, tear down hyperz and cmask filps on r1xx-r5xx
+ * (all asics).  And tear down vm on cayman+ (all asics).
  */
 void radeon_driver_postclose_kms(struct drm_device *dev,
 				 struct drm_file *file_priv)
 {
 	struct radeon_device *rdev = dev->dev_private;
 
+	pm_runtime_get_sync(dev->dev);
+
+	mutex_lock(&rdev->gem.mutex);
+	if (rdev->hyperz_filp == file_priv)
+		rdev->hyperz_filp = NULL;
+	if (rdev->cmask_filp == file_priv)
+		rdev->cmask_filp = NULL;
+	mutex_unlock(&rdev->gem.mutex);
+
+	radeon_uvd_free_handles(rdev, file_priv);
+	radeon_vce_free_handles(rdev, file_priv);
+
 	/* new gpu have virtual address space support */
 	if (rdev->family >= CHIP_CAYMAN && file_priv->driver_priv) {
 		struct radeon_fpriv *fpriv = file_priv->driver_priv;
@@ -1388,31 +1432,8 @@ void radeon_driver_postclose_kms(struct drm_device *dev,
 		kfree(fpriv);
 		file_priv->driver_priv = NULL;
 	}
-}
-
-/**
- * radeon_driver_preclose_kms - drm callback for pre close
- *
- * @dev: drm dev pointer
- * @file_priv: drm file
- *
- * On device pre close, tear down hyperz and cmask filps on r1xx-r5xx
- * (all asics).
- */
-void radeon_driver_preclose_kms(struct drm_device *dev,
-				struct drm_file *file_priv)
-{
-	struct radeon_device *rdev = dev->dev_private;
-
-	mutex_lock(&rdev->gem.mutex);
-	if (rdev->hyperz_filp == file_priv)
-		rdev->hyperz_filp = NULL;
-	if (rdev->cmask_filp == file_priv)
-		rdev->cmask_filp = NULL;
-	mutex_unlock(&rdev->gem.mutex);
-
-	radeon_uvd_free_handles(rdev, file_priv);
-	radeon_vce_free_handles(rdev, file_priv);
+	pm_runtime_mark_last_busy(dev->dev);
+	pm_runtime_put_autosuspend(dev->dev);
 }
 
 /*
@@ -1422,19 +1443,19 @@ void radeon_driver_preclose_kms(struct drm_device *dev,
  * radeon_get_vblank_counter_kms - get frame count
  *
  * @dev: drm dev pointer
- * @crtc: crtc to get the frame count from
+ * @pipe: crtc to get the frame count from
  *
  * Gets the frame count on the requested crtc (all asics).
  * Returns frame count on success, -EINVAL on failure.
  */
-u32 radeon_get_vblank_counter_kms(struct drm_device *dev, int crtc)
+u32 radeon_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe)
 {
 	int vpos, hpos, stat;
 	u32 count;
 	struct radeon_device *rdev = dev->dev_private;
 
-	if (crtc < 0 || crtc >= rdev->num_crtc) {
-		DRM_ERROR("Invalid crtc %d\n", crtc);
+	if (pipe >= rdev->num_crtc) {
+		DRM_ERROR("Invalid crtc %u\n", pipe);
 		return -EINVAL;
 	}
 
@@ -1446,29 +1467,29 @@ u32 radeon_get_vblank_counter_kms(struct drm_device *dev, int crtc)
 	 * and start of vsync, so vpos >= 0 means to bump the hw frame counter
 	 * result by 1 to give the proper appearance to caller.
 	 */
-	if (rdev->mode_info.crtcs[crtc]) {
+	if (rdev->mode_info.crtcs[pipe]) {
 		/* Repeat readout if needed to provide stable result if
 		 * we cross start of vsync during the queries.
 		 */
 		do {
-			count = radeon_get_vblank_counter(rdev, crtc);
+			count = radeon_get_vblank_counter(rdev, pipe);
 			/* Ask radeon_get_crtc_scanoutpos to return vpos as
 			 * distance to start of vblank, instead of regular
 			 * vertical scanout pos.
 			 */
 			stat = radeon_get_crtc_scanoutpos(
-				dev, crtc, GET_DISTANCE_TO_VBLANKSTART,
+				dev, pipe, GET_DISTANCE_TO_VBLANKSTART,
 				&vpos, &hpos, NULL, NULL,
-				&rdev->mode_info.crtcs[crtc]->base.hwmode);
-		} while (count != radeon_get_vblank_counter(rdev, crtc));
+				&rdev->mode_info.crtcs[pipe]->base.hwmode);
+		} while (count != radeon_get_vblank_counter(rdev, pipe));
 
 		if (((stat & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE)) !=
 		    (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE))) {
 			DRM_DEBUG_VBL("Query failed! stat %d\n", stat);
 		}
 		else {
-			DRM_DEBUG_VBL("crtc %d: dist from vblank start %d\n",
-				      crtc, vpos);
+			DRM_DEBUG_VBL("crtc %u: dist from vblank start %d\n",
+				      pipe, vpos);
 
 			/* Bump counter if we are at >= leading edge of vblank,
 			 * but before vsync where vpos would turn negative and
@@ -1480,7 +1501,7 @@ u32 radeon_get_vblank_counter_kms(struct drm_device *dev, int crtc)
 	}
 	else {
 	    /* Fallback to use value as is. */
-	    count = radeon_get_vblank_counter(rdev, crtc);
+	    count = radeon_get_vblank_counter(rdev, pipe);
 	    DRM_DEBUG_VBL("NULL mode info! Returned count may be wrong.\n");
 	}
 
@@ -1538,43 +1559,6 @@ void radeon_disable_vblank_kms(struct drm_device *dev, int crtc)
 	spin_unlock_irqrestore(&rdev->irq.lock, irqflags);
 }
 
-/**
- * radeon_get_vblank_timestamp_kms - get vblank timestamp
- *
- * @dev: drm dev pointer
- * @crtc: crtc to get the timestamp for
- * @max_error: max error
- * @vblank_time: time value
- * @flags: flags passed to the driver
- *
- * Gets the timestamp on the requested crtc based on the
- * scanout position.  (all asics).
- * Returns postive status flags on success, negative error on failure.
- */
-int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags)
-{
-	struct drm_crtc *drmcrtc;
-	struct radeon_device *rdev = dev->dev_private;
-
-	if (crtc < 0 || crtc >= dev->num_crtcs) {
-		DRM_ERROR("Invalid crtc %d\n", crtc);
-		return -EINVAL;
-	}
-
-	/* Get associated drm_crtc: */
-	drmcrtc = &rdev->mode_info.crtcs[crtc]->base;
-	if (!drmcrtc)
-		return -EINVAL;
-
-	/* Helper routine in DRM core does all the work: */
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error,
-						     vblank_time, flags,
-						     &drmcrtc->hwmode);
-}
-
 const struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_CP_INIT, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
 	DRM_IOCTL_DEF_DRV(RADEON_CP_START, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
diff --git a/sys/dev/pci/drm/radeon/radeon_legacy_crtc.c b/sys/dev/pci/drm/radeon/radeon_legacy_crtc.c
index 5dc0dc09277..35a205ae431 100644
--- a/sys/dev/pci/drm/radeon/radeon_legacy_crtc.c
+++ b/sys/dev/pci/drm/radeon/radeon_legacy_crtc.c
@@ -23,10 +23,11 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
-#include <dev/pci/drm/drm_fixed.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/radeon_drm.h>
+#include <drm/drm_fixed.h>
 #include "radeon.h"
 #include "atom.h"
 
@@ -330,15 +331,15 @@ static void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
 									 RADEON_CRTC_DISP_REQ_EN_B));
 			WREG32_P(RADEON_CRTC_EXT_CNTL, crtc_ext_cntl, ~(mask | crtc_ext_cntl));
 		}
-		drm_vblank_post_modeset(dev, radeon_crtc->crtc_id);
-		/* Make sure vblank interrupt is still enabled if needed */
-		radeon_irq_set(rdev);
+		if (dev->num_crtcs > radeon_crtc->crtc_id)
+			drm_crtc_vblank_on(crtc);
 		radeon_crtc_load_lut(crtc);
 		break;
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 	case DRM_MODE_DPMS_OFF:
-		drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id);
+		if (dev->num_crtcs > radeon_crtc->crtc_id)
+			drm_crtc_vblank_off(crtc);
 		if (radeon_crtc->crtc_id)
 			WREG32_P(RADEON_CRTC2_GEN_CNTL, mask, ~(RADEON_CRTC2_EN | mask));
 		else {
@@ -373,7 +374,6 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -392,16 +392,12 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
-	switch (target_fb->bits_per_pixel) {
+	switch (target_fb->format->cpp[0] * 8) {
 	case 8:
 		format = 2;
 		break;
@@ -422,7 +418,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	}
 
 	/* Pin framebuffer & get tilling informations */
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 retry:
 	r = radeon_bo_reserve(rbo, false);
@@ -450,7 +446,7 @@ retry:
 			struct radeon_bo *old_rbo;
 			unsigned long nsize, osize;
 
-			old_rbo = gem_to_radeon_bo(to_radeon_framebuffer(fb)->obj);
+			old_rbo = gem_to_radeon_bo(fb->obj[0]);
 			osize = radeon_bo_size(old_rbo);
 			nsize = radeon_bo_size(rbo);
 			if (nsize <= osize && !radeon_bo_reserve(old_rbo, false)) {
@@ -475,10 +471,9 @@ retry:
 
 	crtc_offset_cntl = 0;
 
-	pitch_pixels = target_fb->pitches[0] / (target_fb->bits_per_pixel / 8);
-	crtc_pitch  = (((pitch_pixels * target_fb->bits_per_pixel) +
-			((target_fb->bits_per_pixel * 8) - 1)) /
-		       (target_fb->bits_per_pixel * 8));
+	pitch_pixels = target_fb->pitches[0] / target_fb->format->cpp[0];
+	crtc_pitch = DIV_ROUND_UP(pitch_pixels * target_fb->format->cpp[0] * 8,
+				  target_fb->format->cpp[0] * 8 * 8);
 	crtc_pitch |= crtc_pitch << 16;
 
 	crtc_offset_cntl |= RADEON_CRTC_GUI_TRIG_OFFSET_LEFT_EN;
@@ -503,14 +498,14 @@ retry:
 			crtc_tile_x0_y0 = x | (y << 16);
 			base &= ~0x7ff;
 		} else {
-			int byteshift = target_fb->bits_per_pixel >> 4;
+			int byteshift = target_fb->format->cpp[0] * 8 >> 4;
 			int tile_addr = (((y >> 3) * pitch_pixels +  x) >> (8 - byteshift)) << 11;
 			base += tile_addr + ((x << byteshift) % 256) + ((y % 8) << 8);
 			crtc_offset_cntl |= (y % 16);
 		}
 	} else {
 		int offset = y * pitch_pixels + x;
-		switch (target_fb->bits_per_pixel) {
+		switch (target_fb->format->cpp[0] * 8) {
 		case 8:
 			offset *= 1;
 			break;
@@ -558,8 +553,7 @@ retry:
 	WREG32(RADEON_CRTC_PITCH + radeon_crtc->crtc_offset, crtc_pitch);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -578,6 +572,7 @@ static bool radeon_set_crtc_timing(struct drm_crtc *crtc, struct drm_display_mod
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
+	const struct drm_framebuffer *fb = crtc->primary->fb;
 	struct drm_encoder *encoder;
 	int format;
 	int hsync_start;
@@ -601,7 +596,7 @@ static bool radeon_set_crtc_timing(struct drm_crtc *crtc, struct drm_display_mod
 		}
 	}
 
-	switch (crtc->primary->fb->bits_per_pixel) {
+	switch (fb->format->cpp[0] * 8) {
 	case 8:
 		format = 2;
 		break;
@@ -1092,11 +1087,9 @@ static void radeon_crtc_disable(struct drm_crtc *crtc)
 	radeon_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
@@ -1115,7 +1108,6 @@ static const struct drm_crtc_helper_funcs legacy_helper_funcs = {
 	.mode_set_base_atomic = radeon_crtc_set_base_atomic,
 	.prepare = radeon_crtc_prepare,
 	.commit = radeon_crtc_commit,
-	.load_lut = radeon_crtc_load_lut,
 	.disable = radeon_crtc_disable
 };
 
diff --git a/sys/dev/pci/drm/radeon/radeon_legacy_encoders.c b/sys/dev/pci/drm/radeon/radeon_legacy_encoders.c
index 46a8aa14ff8..572951f45cf 100644
--- a/sys/dev/pci/drm/radeon/radeon_legacy_encoders.c
+++ b/sys/dev/pci/drm/radeon/radeon_legacy_encoders.c
@@ -23,11 +23,13 @@
  * Authors: Dave Airlie
  *          Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
+#include "radeon_asic.h"
 #include "atom.h"
+#include <linux/backlight.h>
 #ifdef CONFIG_PMAC_BACKLIGHT
 #include <asm/backlight.h>
 #endif
@@ -822,52 +824,52 @@ static void radeon_legacy_tmds_int_mode_set(struct drm_encoder *encoder,
 	tmds_transmitter_cntl = RREG32(RADEON_TMDS_TRANSMITTER_CNTL) &
 		~(RADEON_TMDS_TRANSMITTER_PLLRST);
 
-    if (rdev->family == CHIP_R200 ||
-	rdev->family == CHIP_R100 ||
-	ASIC_IS_R300(rdev))
-	    tmds_transmitter_cntl &= ~(RADEON_TMDS_TRANSMITTER_PLLEN);
-    else /* RV chips got this bit reversed */
-	    tmds_transmitter_cntl |= RADEON_TMDS_TRANSMITTER_PLLEN;
-
-    fp_gen_cntl = (RREG32(RADEON_FP_GEN_CNTL) |
-		   (RADEON_FP_CRTC_DONT_SHADOW_VPAR |
-		    RADEON_FP_CRTC_DONT_SHADOW_HEND));
-
-    fp_gen_cntl &= ~(RADEON_FP_FPON | RADEON_FP_TMDS_EN);
-
-    fp_gen_cntl &= ~(RADEON_FP_RMX_HVSYNC_CONTROL_EN |
-		     RADEON_FP_DFP_SYNC_SEL |
-		     RADEON_FP_CRT_SYNC_SEL |
-		     RADEON_FP_CRTC_LOCK_8DOT |
-		     RADEON_FP_USE_SHADOW_EN |
-		     RADEON_FP_CRTC_USE_SHADOW_VEND |
-		     RADEON_FP_CRT_SYNC_ALT);
-
-    if (1) /*  FIXME rgbBits == 8 */
-	    fp_gen_cntl |= RADEON_FP_PANEL_FORMAT;  /* 24 bit format */
-    else
-	    fp_gen_cntl &= ~RADEON_FP_PANEL_FORMAT;/* 18 bit format */
-
-    if (radeon_crtc->crtc_id == 0) {
-	    if (ASIC_IS_R300(rdev) || rdev->family == CHIP_R200) {
-		    fp_gen_cntl &= ~R200_FP_SOURCE_SEL_MASK;
-		    if (radeon_encoder->rmx_type != RMX_OFF)
-			    fp_gen_cntl |= R200_FP_SOURCE_SEL_RMX;
-		    else
-			    fp_gen_cntl |= R200_FP_SOURCE_SEL_CRTC1;
-	    } else
-		    fp_gen_cntl &= ~RADEON_FP_SEL_CRTC2;
-    } else {
-	    if (ASIC_IS_R300(rdev) || rdev->family == CHIP_R200) {
-		    fp_gen_cntl &= ~R200_FP_SOURCE_SEL_MASK;
-		    fp_gen_cntl |= R200_FP_SOURCE_SEL_CRTC2;
-	    } else
-		    fp_gen_cntl |= RADEON_FP_SEL_CRTC2;
-    }
-
-    WREG32(RADEON_TMDS_PLL_CNTL, tmds_pll_cntl);
-    WREG32(RADEON_TMDS_TRANSMITTER_CNTL, tmds_transmitter_cntl);
-    WREG32(RADEON_FP_GEN_CNTL, fp_gen_cntl);
+	if (rdev->family == CHIP_R200 ||
+	    rdev->family == CHIP_R100 ||
+	    ASIC_IS_R300(rdev))
+		tmds_transmitter_cntl &= ~(RADEON_TMDS_TRANSMITTER_PLLEN);
+	else /* RV chips got this bit reversed */
+		tmds_transmitter_cntl |= RADEON_TMDS_TRANSMITTER_PLLEN;
+
+	fp_gen_cntl = (RREG32(RADEON_FP_GEN_CNTL) |
+		      (RADEON_FP_CRTC_DONT_SHADOW_VPAR |
+		       RADEON_FP_CRTC_DONT_SHADOW_HEND));
+
+	fp_gen_cntl &= ~(RADEON_FP_FPON | RADEON_FP_TMDS_EN);
+
+	fp_gen_cntl &= ~(RADEON_FP_RMX_HVSYNC_CONTROL_EN |
+			 RADEON_FP_DFP_SYNC_SEL |
+			 RADEON_FP_CRT_SYNC_SEL |
+			 RADEON_FP_CRTC_LOCK_8DOT |
+			 RADEON_FP_USE_SHADOW_EN |
+			 RADEON_FP_CRTC_USE_SHADOW_VEND |
+			 RADEON_FP_CRT_SYNC_ALT);
+
+	if (1) /*  FIXME rgbBits == 8 */
+		fp_gen_cntl |= RADEON_FP_PANEL_FORMAT;  /* 24 bit format */
+	else
+		fp_gen_cntl &= ~RADEON_FP_PANEL_FORMAT;/* 18 bit format */
+
+	if (radeon_crtc->crtc_id == 0) {
+		if (ASIC_IS_R300(rdev) || rdev->family == CHIP_R200) {
+			fp_gen_cntl &= ~R200_FP_SOURCE_SEL_MASK;
+			if (radeon_encoder->rmx_type != RMX_OFF)
+				fp_gen_cntl |= R200_FP_SOURCE_SEL_RMX;
+			else
+				fp_gen_cntl |= R200_FP_SOURCE_SEL_CRTC1;
+		} else
+			fp_gen_cntl &= ~RADEON_FP_SEL_CRTC2;
+	} else {
+		if (ASIC_IS_R300(rdev) || rdev->family == CHIP_R200) {
+			fp_gen_cntl &= ~R200_FP_SOURCE_SEL_MASK;
+			fp_gen_cntl |= R200_FP_SOURCE_SEL_CRTC2;
+		} else
+			fp_gen_cntl |= RADEON_FP_SEL_CRTC2;
+	}
+
+	WREG32(RADEON_TMDS_PLL_CNTL, tmds_pll_cntl);
+	WREG32(RADEON_TMDS_TRANSMITTER_CNTL, tmds_transmitter_cntl);
+	WREG32(RADEON_FP_GEN_CNTL, fp_gen_cntl);
 
 	if (rdev->is_atom_bios)
 		radeon_atombios_encoder_crtc_scratch_regs(encoder, radeon_crtc->crtc_id);
@@ -1776,7 +1778,8 @@ radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_
 	switch (radeon_encoder->encoder_id) {
 	case ENCODER_OBJECT_ID_INTERNAL_LVDS:
 		encoder->possible_crtcs = 0x1;
-		drm_encoder_init(dev, encoder, &radeon_legacy_lvds_enc_funcs, DRM_MODE_ENCODER_LVDS);
+		drm_encoder_init(dev, encoder, &radeon_legacy_lvds_enc_funcs,
+				 DRM_MODE_ENCODER_LVDS, NULL);
 		drm_encoder_helper_add(encoder, &radeon_legacy_lvds_helper_funcs);
 		if (rdev->is_atom_bios)
 			radeon_encoder->enc_priv = radeon_atombios_get_lvds_info(radeon_encoder);
@@ -1785,12 +1788,14 @@ radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_
 		radeon_encoder->rmx_type = RMX_FULL;
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_TMDS1:
-		drm_encoder_init(dev, encoder, &radeon_legacy_tmds_int_enc_funcs, DRM_MODE_ENCODER_TMDS);
+		drm_encoder_init(dev, encoder, &radeon_legacy_tmds_int_enc_funcs,
+				 DRM_MODE_ENCODER_TMDS, NULL);
 		drm_encoder_helper_add(encoder, &radeon_legacy_tmds_int_helper_funcs);
 		radeon_encoder->enc_priv = radeon_legacy_get_tmds_info(radeon_encoder);
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_DAC1:
-		drm_encoder_init(dev, encoder, &radeon_legacy_primary_dac_enc_funcs, DRM_MODE_ENCODER_DAC);
+		drm_encoder_init(dev, encoder, &radeon_legacy_primary_dac_enc_funcs,
+				 DRM_MODE_ENCODER_DAC, NULL);
 		drm_encoder_helper_add(encoder, &radeon_legacy_primary_dac_helper_funcs);
 		if (rdev->is_atom_bios)
 			radeon_encoder->enc_priv = radeon_atombios_get_primary_dac_info(radeon_encoder);
@@ -1798,7 +1803,8 @@ radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_
 			radeon_encoder->enc_priv = radeon_combios_get_primary_dac_info(radeon_encoder);
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_DAC2:
-		drm_encoder_init(dev, encoder, &radeon_legacy_tv_dac_enc_funcs, DRM_MODE_ENCODER_TVDAC);
+		drm_encoder_init(dev, encoder, &radeon_legacy_tv_dac_enc_funcs,
+				 DRM_MODE_ENCODER_TVDAC, NULL);
 		drm_encoder_helper_add(encoder, &radeon_legacy_tv_dac_helper_funcs);
 		if (rdev->is_atom_bios)
 			radeon_encoder->enc_priv = radeon_atombios_get_tv_dac_info(radeon_encoder);
@@ -1806,7 +1812,8 @@ radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_
 			radeon_encoder->enc_priv = radeon_combios_get_tv_dac_info(radeon_encoder);
 		break;
 	case ENCODER_OBJECT_ID_INTERNAL_DVO1:
-		drm_encoder_init(dev, encoder, &radeon_legacy_tmds_ext_enc_funcs, DRM_MODE_ENCODER_TMDS);
+		drm_encoder_init(dev, encoder, &radeon_legacy_tmds_ext_enc_funcs,
+				 DRM_MODE_ENCODER_TMDS, NULL);
 		drm_encoder_helper_add(encoder, &radeon_legacy_tmds_ext_helper_funcs);
 		if (!rdev->is_atom_bios)
 			radeon_encoder->enc_priv = radeon_legacy_get_ext_tmds_info(radeon_encoder);
diff --git a/sys/dev/pci/drm/radeon/radeon_legacy_tv.c b/sys/dev/pci/drm/radeon/radeon_legacy_tv.c
index 424cd2387e9..4278272e319 100644
--- a/sys/dev/pci/drm/radeon/radeon_legacy_tv.c
+++ b/sys/dev/pci/drm/radeon/radeon_legacy_tv.c
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: MIT
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
 #include "radeon.h"
 
 /*
diff --git a/sys/dev/pci/drm/radeon/radeon_mn.c b/sys/dev/pci/drm/radeon/radeon_mn.c
index 460a6a40a69..f8b35df44c6 100644
--- a/sys/dev/pci/drm/radeon/radeon_mn.c
+++ b/sys/dev/pci/drm/radeon/radeon_mn.c
@@ -28,8 +28,11 @@
  *    Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <linux/mmu_notifier.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
 
 #include "radeon.h"
 
@@ -46,8 +49,8 @@ struct radeon_mn {
 	struct hlist_node	node;
 
 	/* objects protected by lock */
-	struct rwlock		lock;
-	struct rb_root		objects;
+	struct mutex		lock;
+	struct rb_root_cached	objects;
 };
 
 struct radeon_mn_node {
@@ -72,8 +75,8 @@ static void radeon_mn_destroy(struct work_struct *work)
 	mutex_lock(&rdev->mn_lock);
 	mutex_lock(&rmn->lock);
 	hash_del(&rmn->node);
-	rbtree_postorder_for_each_entry_safe(node, next_node, &rmn->objects,
-					     it.rb) {
+	rbtree_postorder_for_each_entry_safe(node, next_node,
+					     &rmn->objects.rb_root, it.rb) {
 
 		interval_tree_remove(&node->it, &rmn->objects);
 		list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) {
@@ -115,18 +118,27 @@ static void radeon_mn_release(struct mmu_notifier *mn,
  * We block for all BOs between start and end to be idle and
  * unmap them by move them into system domain again.
  */
-static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
+static int radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
 					     struct mm_struct *mm,
 					     unsigned long start,
-					     unsigned long end)
+					     unsigned long end,
+					     bool blockable)
 {
 	struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
+	struct ttm_operation_ctx ctx = { false, false };
 	struct interval_tree_node *it;
+	int ret = 0;
 
 	/* notification is exclusive, but interval is inclusive */
 	end -= 1;
 
-	mutex_lock(&rmn->lock);
+	/* TODO we should be able to split locking for interval tree and
+	 * the tear down.
+	 */
+	if (blockable)
+		mutex_lock(&rmn->lock);
+	else if (!mutex_trylock(&rmn->lock))
+		return -EAGAIN;
 
 	it = interval_tree_iter_first(&rmn->objects, start, end);
 	while (it) {
@@ -134,6 +146,11 @@ static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
 		struct radeon_bo *bo;
 		long r;
 
+		if (!blockable) {
+			ret = -EAGAIN;
+			goto out_unlock;
+		}
+
 		node = container_of(it, struct radeon_mn_node, it);
 		it = interval_tree_iter_next(it, start, end);
 
@@ -154,7 +171,7 @@ static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
 				DRM_ERROR("(%ld) failed to wait for user bo\n", r);
 
 			radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_CPU);
-			r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+			r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 			if (r)
 				DRM_ERROR("(%ld) failed to validate user bo\n", r);
 
@@ -162,7 +179,10 @@ static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn,
 		}
 	}
 	
+out_unlock:
 	mutex_unlock(&rmn->lock);
+
+	return ret;
 }
 
 static const struct mmu_notifier_ops radeon_mn_ops = {
@@ -183,7 +203,9 @@ static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev)
 	struct radeon_mn *rmn;
 	int r;
 
-	down_write(&mm->mmap_sem);
+	if (down_write_killable(&mm->mmap_sem))
+		return ERR_PTR(-EINTR);
+
 	mutex_lock(&rdev->mn_lock);
 
 	hash_for_each_possible(rdev->mn_hash, rmn, node, (unsigned long)mm)
@@ -199,8 +221,8 @@ static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev)
 	rmn->rdev = rdev;
 	rmn->mm = mm;
 	rmn->mn.ops = &radeon_mn_ops;
-	rw_init(&rmn->lock, "rmn");
-	rmn->objects = RB_ROOT;
+	mutex_init(&rmn->lock);
+	rmn->objects = RB_ROOT_CACHED;
 	
 	r = __mmu_notifier_register(&rmn->mn, mm);
 	if (r)
diff --git a/sys/dev/pci/drm/radeon/radeon_mode.h b/sys/dev/pci/drm/radeon/radeon_mode.h
index 81a6df8913e..fc9f959141a 100644
--- a/sys/dev/pci/drm/radeon/radeon_mode.h
+++ b/sys/dev/pci/drm/radeon/radeon_mode.h
@@ -30,12 +30,15 @@
 #ifndef RADEON_MODE_H
 #define RADEON_MODE_H
 
-#include <dev/pci/drm/drm_crtc.h>
-#include <dev/pci/drm/drm_edid.h>
-#include <dev/pci/drm/drm_dp_helper.h>
-#include <dev/pci/drm/drm_dp_mst_helper.h>
-#include <dev/pci/drm/drm_fixed.h>
-#include <dev/pci/drm/drm_crtc_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_encoder.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_dp_mst_helper.h>
+#include <drm/drm_fixed.h>
+#include <drm/drm_crtc_helper.h>
+#include <linux/i2c.h>
+#include <linux/i2c-algo-bit.h>
 
 struct radeon_bo;
 struct radeon_device;
@@ -43,7 +46,6 @@ struct radeon_device;
 #define to_radeon_crtc(x) container_of(x, struct radeon_crtc, base)
 #define to_radeon_connector(x) container_of(x, struct radeon_connector, base)
 #define to_radeon_encoder(x) container_of(x, struct radeon_encoder, base)
-#define to_radeon_framebuffer(x) container_of(x, struct radeon_framebuffer, base)
 
 #define RADEON_MAX_HPD_PINS 7
 #define RADEON_MAX_CRTCS 6
@@ -571,11 +573,6 @@ struct radeon_connector {
 	int enabled_attribs;
 };
 
-struct radeon_framebuffer {
-	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
-};
-
 #define ENCODER_MODE_IS_DP(em) (((em) == ATOM_ENCODER_MODE_DP) || \
 				((em) == ATOM_ENCODER_MODE_DP_MST))
 
@@ -688,6 +685,9 @@ struct atom_voltage_table
 };
 
 /* Driver internal use only flags of radeon_get_crtc_scanoutpos() */
+#define DRM_SCANOUTPOS_VALID        (1 << 0)
+#define DRM_SCANOUTPOS_IN_VBLANK    (1 << 1)
+#define DRM_SCANOUTPOS_ACCURATE     (1 << 2)
 #define USE_REAL_VBLANKSTART 		(1 << 30)
 #define GET_DISTANCE_TO_VBLANKSTART	(1 << 31)
 
@@ -756,10 +756,6 @@ extern u8 radeon_dp_getsinktype(struct radeon_connector *radeon_connector);
 extern bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector);
 extern int radeon_dp_get_panel_mode(struct drm_encoder *encoder,
 				    struct drm_connector *connector);
-extern int radeon_dp_get_dp_link_config(struct drm_connector *connector,
-					const u8 *dpcd,
-					unsigned pix_clock,
-					unsigned *dp_lanes, unsigned *dp_rate);
 extern void radeon_dp_set_rx_power_state(struct drm_connector *connector,
 					 u8 power_state);
 extern void radeon_dp_aux_init(struct radeon_connector *radeon_connector);
@@ -929,13 +925,9 @@ extern void
 radeon_combios_encoder_crtc_scratch_regs(struct drm_encoder *encoder, int crtc);
 extern void
 radeon_combios_encoder_dpms_scratch_regs(struct drm_encoder *encoder, bool on);
-extern void radeon_crtc_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green,
-				     u16 blue, int regno);
-extern void radeon_crtc_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green,
-				     u16 *blue, int regno);
 int radeon_framebuffer_init(struct drm_device *dev,
-			     struct radeon_framebuffer *rfb,
-			     struct drm_mode_fb_cmd2 *mode_cmd,
+			     struct drm_framebuffer *rfb,
+			     const struct drm_mode_fb_cmd2 *mode_cmd,
 			     struct drm_gem_object *obj);
 
 int radeonfb_remove(struct drm_device *dev, struct drm_framebuffer *fb);
@@ -986,9 +978,6 @@ int radeon_fbdev_init(struct radeon_device *rdev);
 void radeon_fbdev_fini(struct radeon_device *rdev);
 void radeon_fbdev_set_suspend(struct radeon_device *rdev, int state);
 bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj);
-void radeon_fbdev_restore_mode(struct radeon_device *rdev);
-
-void radeon_fb_output_poll_changed(struct radeon_device *rdev);
 
 void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id);
 
diff --git a/sys/dev/pci/drm/radeon/radeon_object.c b/sys/dev/pci/drm/radeon/radeon_object.c
index d8006d66ccf..35948e87328 100644
--- a/sys/dev/pci/drm/radeon/radeon_object.c
+++ b/sys/dev/pci/drm/radeon/radeon_object.c
@@ -29,9 +29,11 @@
  *    Thomas Hellstrom <thomas-at-tungstengraphics-dot-com>
  *    Dave Airlie
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
-#include <dev/pci/drm/drm_cache.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
+#include <drm/drm_cache.h>
 #include "radeon.h"
 #include "radeon_trace.h"
 
@@ -79,7 +81,11 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	list_del_init(&bo->list);
 	mutex_unlock(&bo->rdev->gem.mutex);
 	radeon_bo_clear_surface_reg(bo);
-	WARN_ON(!list_empty(&bo->va));
+	WARN_ON_ONCE(!list_empty(&bo->va));
+#ifdef notyet
+	if (bo->gem_base.import_attach)
+		drm_prime_gem_destroy(&bo->gem_base, bo->tbo.sg);
+#endif
 	drm_gem_object_release(&bo->gem_base);
 	pool_put(&bo->rdev->ddev->objpl, bo);
 }
@@ -202,18 +208,14 @@ int radeon_bo_create(struct radeon_device *rdev,
 	bo = pool_get(&rdev->ddev->objpl, PR_WAITOK | PR_ZERO);
 	if (bo == NULL)
 		return -ENOMEM;
-	r = drm_gem_object_init(rdev->ddev, &bo->gem_base, size);
-	if (unlikely(r)) {
-		pool_put(&rdev->ddev->objpl, bo);
-		return r;
-	}
+	drm_gem_private_object_init(rdev->ddev, &bo->gem_base, size);
 	bo->rdev = rdev;
 	bo->surface_reg = -1;
 	INIT_LIST_HEAD(&bo->list);
 	INIT_LIST_HEAD(&bo->va);
 	bo->initial_domain = domain & (RADEON_GEM_DOMAIN_VRAM |
-	                               RADEON_GEM_DOMAIN_GTT |
-	                               RADEON_GEM_DOMAIN_CPU);
+				       RADEON_GEM_DOMAIN_GTT |
+				       RADEON_GEM_DOMAIN_CPU);
 
 	bo->flags = flags;
 	/* PCI GART is always snooped */
@@ -257,8 +259,8 @@ int radeon_bo_create(struct radeon_device *rdev,
 	/* Kernel allocation are uninterruptible */
 	down_read(&rdev->pm.mclk_lock);
 	r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type,
-			&bo->placement, page_align, !kernel, NULL,
-			acc_size, sg, resv, &radeon_ttm_bo_destroy);
+			&bo->placement, page_align, !kernel, acc_size,
+			sg, resv, &radeon_ttm_bo_destroy);
 	up_read(&rdev->pm.mclk_lock);
 	if (unlikely(r != 0)) {
 		return r;
@@ -307,7 +309,7 @@ struct radeon_bo *radeon_bo_ref(struct radeon_bo *bo)
 	if (bo == NULL)
 		return NULL;
 
-	ttm_bo_reference(&bo->tbo);
+	ttm_bo_get(&bo->tbo);
 	return bo;
 }
 
@@ -320,14 +322,14 @@ void radeon_bo_unref(struct radeon_bo **bo)
 		return;
 	rdev = (*bo)->rdev;
 	tbo = &((*bo)->tbo);
-	ttm_bo_unref(&tbo);
-	if (tbo == NULL)
-		*bo = NULL;
+	ttm_bo_put(tbo);
+	*bo = NULL;
 }
 
 int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 			     u64 *gpu_addr)
 {
+	struct ttm_operation_ctx ctx = { false, false };
 	int r, i;
 
 	if (radeon_ttm_tt_has_userptr(bo->tbo.ttm))
@@ -351,6 +353,11 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 
 		return 0;
 	}
+	if (bo->prime_shared_count && domain == RADEON_GEM_DOMAIN_VRAM) {
+		/* A BO shared as a dma-buf cannot be sensibly migrated to VRAM */
+		return -EINVAL;
+	}
+
 	radeon_ttm_placement_from_domain(bo, domain);
 	for (i = 0; i < bo->placement.num_placement; i++) {
 		/* force to pin into visible video ram */
@@ -365,7 +372,7 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset,
 		bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT;
 	}
 
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 	if (likely(r == 0)) {
 		bo->pin_count = 1;
 		if (gpu_addr != NULL)
@@ -387,6 +394,7 @@ int radeon_bo_pin(struct radeon_bo *bo, u32 domain, u64 *gpu_addr)
 
 int radeon_bo_unpin(struct radeon_bo *bo)
 {
+	struct ttm_operation_ctx ctx = { false, false };
 	int r, i;
 
 	if (!bo->pin_count) {
@@ -400,7 +408,7 @@ int radeon_bo_unpin(struct radeon_bo *bo)
 		bo->placements[i].lpfn = 0;
 		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	}
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 	if (likely(r == 0)) {
 		if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
 			bo->rdev->vram_pin_size -= radeon_bo_size(bo);
@@ -439,7 +447,7 @@ void radeon_bo_force_delete(struct radeon_device *rdev)
 		list_del_init(&bo->list);
 		mutex_unlock(&bo->rdev->gem.mutex);
 		/* this should unref the ttm bo */
-		drm_gem_object_unreference_unlocked(&bo->gem_base);
+		drm_gem_object_put_unlocked(&bo->gem_base);
 	}
 }
 
@@ -544,6 +552,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 			    struct ww_acquire_ctx *ticket,
 			    struct list_head *head, int ring)
 {
+	struct ttm_operation_ctx ctx = { true, false };
 	struct radeon_bo_list *lobj;
 	struct list_head duplicates;
 	int r;
@@ -559,7 +568,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 	list_for_each_entry(lobj, head, tv.head) {
 		struct radeon_bo *bo = lobj->robj;
 		if (!bo->pin_count) {
-			u32 domain = lobj->prefered_domains;
+			u32 domain = lobj->preferred_domains;
 			u32 allowed = lobj->allowed_domains;
 			u32 current_domain =
 				radeon_mem_type_to_domain(bo->tbo.mem.mem_type);
@@ -585,7 +594,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 				radeon_uvd_force_into_uvd_segment(bo, allowed);
 
 			initial_bytes_moved = atomic64_read(&rdev->num_bytes_moved);
-			r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+			r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 			bytes_moved += atomic64_read(&rdev->num_bytes_moved) -
 				       initial_bytes_moved;
 
@@ -789,6 +798,7 @@ int radeon_bo_check_tiling(struct radeon_bo *bo, bool has_moved,
 }
 
 void radeon_bo_move_notify(struct ttm_buffer_object *bo,
+			   bool evict,
 			   struct ttm_mem_reg *new_mem)
 {
 	struct radeon_bo *rbo;
@@ -810,6 +820,7 @@ void radeon_bo_move_notify(struct ttm_buffer_object *bo,
 
 int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 {
+	struct ttm_operation_ctx ctx = { false, false };
 	struct radeon_device *rdev;
 	struct radeon_bo *rbo;
 	unsigned long offset, size, lpfn;
@@ -828,6 +839,10 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 	if ((offset + size) <= rdev->mc.visible_vram_size)
 		return 0;
 
+	/* Can't move a pinned BO to visible VRAM */
+	if (rbo->pin_count > 0)
+		return -EINVAL;
+
 	/* hurrah the memory is not visible ! */
 	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
 	lpfn =	rdev->mc.visible_vram_size >> PAGE_SHIFT;
@@ -837,10 +852,10 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 		    (!rbo->placements[i].lpfn || rbo->placements[i].lpfn > lpfn))
 			rbo->placements[i].lpfn = lpfn;
 	}
-	r = ttm_bo_validate(bo, &rbo->placement, false, false);
+	r = ttm_bo_validate(bo, &rbo->placement, &ctx);
 	if (unlikely(r == -ENOMEM)) {
 		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
-		return ttm_bo_validate(bo, &rbo->placement, false, false);
+		return ttm_bo_validate(bo, &rbo->placement, &ctx);
 	} else if (unlikely(r != 0)) {
 		return r;
 	}
@@ -857,13 +872,13 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait)
 {
 	int r;
 
-	r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, NULL);
+	r = ttm_bo_reserve(&bo->tbo, true, no_wait, NULL);
 	if (unlikely(r != 0))
 		return r;
 	if (mem_type)
 		*mem_type = bo->tbo.mem.mem_type;
 
-	r = ttm_bo_wait(&bo->tbo, true, true, no_wait);
+	r = ttm_bo_wait(&bo->tbo, true, no_wait);
 	ttm_bo_unreserve(&bo->tbo);
 	return r;
 }
@@ -877,7 +892,7 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait)
  *
  */
 void radeon_bo_fence(struct radeon_bo *bo, struct radeon_fence *fence,
-                     bool shared)
+		     bool shared)
 {
 	struct reservation_object *resv = bo->tbo.resv;
 
diff --git a/sys/dev/pci/drm/radeon/radeon_object.h b/sys/dev/pci/drm/radeon/radeon_object.h
index 7b07e8f32ce..9ffd8215d38 100644
--- a/sys/dev/pci/drm/radeon/radeon_object.h
+++ b/sys/dev/pci/drm/radeon/radeon_object.h
@@ -28,7 +28,7 @@
 #ifndef __RADEON_OBJECT_H__
 #define __RADEON_OBJECT_H__
 
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 /**
@@ -65,7 +65,7 @@ static inline int radeon_bo_reserve(struct radeon_bo *bo, bool no_intr)
 {
 	int r;
 
-	r = ttm_bo_reserve(&bo->tbo, !no_intr, false, false, 0);
+	r = ttm_bo_reserve(&bo->tbo, !no_intr, false, NULL);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			dev_err(bo->rdev->dev, "%p reserve failed\n", bo);
@@ -150,6 +150,7 @@ extern void radeon_bo_get_tiling_flags(struct radeon_bo *bo,
 extern int radeon_bo_check_tiling(struct radeon_bo *bo, bool has_moved,
 				bool force_drop);
 extern void radeon_bo_move_notify(struct ttm_buffer_object *bo,
+				  bool evict,
 				  struct ttm_mem_reg *new_mem);
 extern int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
 extern int radeon_bo_get_surface_reg(struct radeon_bo *bo);
diff --git a/sys/dev/pci/drm/radeon/radeon_pm.c b/sys/dev/pci/drm/radeon/radeon_pm.c
index 08829521923..82d3e4a66d0 100644
--- a/sys/dev/pci/drm/radeon/radeon_pm.c
+++ b/sys/dev/pci/drm/radeon/radeon_pm.c
@@ -20,11 +20,14 @@
  * Authors: Rafał Miłecki <zajec5@gmail.com>
  *          Alex Deucher <alexdeucher@gmail.com>
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "avivod.h"
 #include "atom.h"
 #include "r600_dpm.h"
+#include <linux/power_supply.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
 
 #define RADEON_IDLE_LOOP_MS 100
 #define RADEON_RECLOCK_DELAY_MS 200
@@ -78,7 +81,7 @@ void radeon_pm_acpi_event_handler(struct radeon_device *rdev)
 				radeon_dpm_enable_bapm(rdev, rdev->pm.dpm.ac_power);
 		}
 		mutex_unlock(&rdev->pm.mutex);
-        } else if (rdev->pm.pm_method == PM_METHOD_PROFILE) {
+	} else if (rdev->pm.pm_method == PM_METHOD_PROFILE) {
 		if (rdev->pm.profile == PM_PROFILE_AUTO) {
 			mutex_lock(&rdev->pm.mutex);
 			radeon_pm_update_profile(rdev);
@@ -245,6 +248,7 @@ static void radeon_set_power_state(struct radeon_device *rdev)
 
 static void radeon_pm_set_clocks(struct radeon_device *rdev)
 {
+	struct drm_crtc *crtc;
 	int i, r;
 
 	/* no need to take locks, etc. if nothing's going to change */
@@ -273,22 +277,30 @@ static void radeon_pm_set_clocks(struct radeon_device *rdev)
 	radeon_unmap_vram_bos(rdev);
 
 	if (rdev->irq.installed) {
-		for (i = 0; i < rdev->num_crtc; i++) {
+		i = 0;
+		drm_for_each_crtc(crtc, rdev->ddev) {
 			if (rdev->pm.active_crtcs & (1 << i)) {
-				rdev->pm.req_vblank |= (1 << i);
-				drm_vblank_get(rdev->ddev, i);
+				/* This can fail if a modeset is in progress */
+				if (drm_crtc_vblank_get(crtc) == 0)
+					rdev->pm.req_vblank |= (1 << i);
+				else
+					DRM_DEBUG_DRIVER("crtc %d no vblank, can glitch\n",
+							 i);
 			}
+			i++;
 		}
 	}
 
 	radeon_set_power_state(rdev);
 
 	if (rdev->irq.installed) {
-		for (i = 0; i < rdev->num_crtc; i++) {
+		i = 0;
+		drm_for_each_crtc(crtc, rdev->ddev) {
 			if (rdev->pm.req_vblank & (1 << i)) {
 				rdev->pm.req_vblank &= ~(1 << i);
-				drm_vblank_put(rdev->ddev, i);
+				drm_crtc_vblank_put(crtc);
 			}
+			i++;
 		}
 	}
 
@@ -725,7 +737,7 @@ static struct attribute *hwmon_attributes[] = {
 static umode_t hwmon_attributes_visible(struct kobject *kobj,
 					struct attribute *attr, int index)
 {
-	struct device *dev = container_of(kobj, struct device, kobj);
+	struct device *dev = kobj_to_dev(kobj);
 	struct radeon_device *rdev = dev_get_drvdata(dev);
 	umode_t effective_mode = attr->mode;
 
diff --git a/sys/dev/pci/drm/radeon/radeon_prime.c b/sys/dev/pci/drm/radeon/radeon_prime.c
index 1a44476e5a1..be4c4f1b531 100644
--- a/sys/dev/pci/drm/radeon/radeon_prime.c
+++ b/sys/dev/pci/drm/radeon/radeon_prime.c
@@ -23,10 +23,11 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 
 #include "radeon.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
+#include <linux/dma-buf.h>
 
 #ifdef notyet
 
@@ -78,6 +79,7 @@ struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev,
 	list_add_tail(&bo->list, &rdev->gem.objects);
 	mutex_unlock(&rdev->gem.mutex);
 
+	bo->prime_shared_count = 1;
 	return &bo->gem_base;
 }
 
@@ -92,6 +94,9 @@ int radeon_gem_prime_pin(struct drm_gem_object *obj)
 
 	/* pin buffer into GTT */
 	ret = radeon_bo_pin(bo, RADEON_GEM_DOMAIN_GTT, NULL);
+	if (likely(ret == 0))
+		bo->prime_shared_count++;
+
 	radeon_bo_unreserve(bo);
 	return ret;
 }
@@ -106,9 +111,12 @@ void radeon_gem_prime_unpin(struct drm_gem_object *obj)
 		return;
 
 	radeon_bo_unpin(bo);
+	if (bo->prime_shared_count)
+		bo->prime_shared_count--;
 	radeon_bo_unreserve(bo);
 }
 
+#endif
 
 struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *obj)
 {
@@ -117,8 +125,6 @@ struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *obj)
 	return bo->tbo.resv;
 }
 
-#endif
-
 struct dma_buf *radeon_gem_prime_export(struct drm_device *dev,
 					struct drm_gem_object *gobj,
 					int flags)
diff --git a/sys/dev/pci/drm/radeon/radeon_ring.c b/sys/dev/pci/drm/radeon/radeon_ring.c
index 54d2e7ef9c2..84802b201be 100644
--- a/sys/dev/pci/drm/radeon/radeon_ring.c
+++ b/sys/dev/pci/drm/radeon/radeon_ring.c
@@ -26,7 +26,7 @@
  *          Jerome Glisse
  *          Christian König
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 
 /*
@@ -314,7 +314,7 @@ unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring
 	}
 
 	/* and then save the content of the ring */
-	*data = drm_malloc_ab(size, sizeof(uint32_t));
+	*data = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
 	if (!*data) {
 		mutex_unlock(&rdev->ring_lock);
 		return 0;
@@ -356,7 +356,7 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	drm_free_large(data);
+	kvfree(data);
 	return 0;
 }
 
diff --git a/sys/dev/pci/drm/radeon/radeon_sa.c b/sys/dev/pci/drm/radeon/radeon_sa.c
index a9e3a2308c8..197b157b73d 100644
--- a/sys/dev/pci/drm/radeon/radeon_sa.c
+++ b/sys/dev/pci/drm/radeon/radeon_sa.c
@@ -41,7 +41,7 @@
  * If we are asked to block we wait on all the oldest fence of all
  * rings. We just wait for any of those fence to complete.
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 
 static void radeon_sa_bo_remove_locked(struct radeon_sa_bo *sa_bo);
diff --git a/sys/dev/pci/drm/radeon/radeon_semaphore.c b/sys/dev/pci/drm/radeon/radeon_semaphore.c
index e6d2f7bdcb1..b0eb28e8fb7 100644
--- a/sys/dev/pci/drm/radeon/radeon_semaphore.c
+++ b/sys/dev/pci/drm/radeon/radeon_semaphore.c
@@ -27,7 +27,7 @@
  * Authors:
  *    Christian König <deathsimple@vodafone.de>
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_trace.h"
 
@@ -56,7 +56,7 @@ int radeon_semaphore_create(struct radeon_device *rdev,
 }
 
 bool radeon_semaphore_emit_signal(struct radeon_device *rdev, int ridx,
-			          struct radeon_semaphore *semaphore)
+				  struct radeon_semaphore *semaphore)
 {
 	struct radeon_ring *ring = &rdev->ring[ridx];
 
@@ -73,7 +73,7 @@ bool radeon_semaphore_emit_signal(struct radeon_device *rdev, int ridx,
 }
 
 bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ridx,
-			        struct radeon_semaphore *semaphore)
+				struct radeon_semaphore *semaphore)
 {
 	struct radeon_ring *ring = &rdev->ring[ridx];
 
diff --git a/sys/dev/pci/drm/radeon/radeon_sync.c b/sys/dev/pci/drm/radeon/radeon_sync.c
index 668b45c0d9b..be5d7a38d3a 100644
--- a/sys/dev/pci/drm/radeon/radeon_sync.c
+++ b/sys/dev/pci/drm/radeon/radeon_sync.c
@@ -28,7 +28,7 @@
  *    Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_trace.h"
 
@@ -92,7 +92,7 @@ int radeon_sync_resv(struct radeon_device *rdev,
 		     bool shared)
 {
 	struct reservation_object_list *flist;
-	struct fence *f;
+	struct dma_fence *f;
 	struct radeon_fence *fence;
 	unsigned i;
 	int r = 0;
@@ -103,7 +103,7 @@ int radeon_sync_resv(struct radeon_device *rdev,
 	if (fence && fence->rdev == rdev)
 		radeon_sync_fence(sync, fence);
 	else if (f)
-		r = fence_wait(f, true);
+		r = dma_fence_wait(f, true);
 
 	flist = reservation_object_get_list(resv);
 	if (shared || !flist || r)
@@ -116,7 +116,7 @@ int radeon_sync_resv(struct radeon_device *rdev,
 		if (fence && fence->rdev == rdev)
 			radeon_sync_fence(sync, fence);
 		else
-			r = fence_wait(f, true);
+			r = dma_fence_wait(f, true);
 
 		if (r)
 			break;
diff --git a/sys/dev/pci/drm/radeon/radeon_test.c b/sys/dev/pci/drm/radeon/radeon_test.c
index 4ab69640929..0c7f228db6e 100644
--- a/sys/dev/pci/drm/radeon/radeon_test.c
+++ b/sys/dev/pci/drm/radeon/radeon_test.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
  * Copyright 2009 VMware, Inc.
  *
@@ -21,8 +22,8 @@
  *
  * Authors: Michel Dänzer
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 
@@ -59,7 +60,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 	n = rdev->mc.gtt_size - rdev->gart_pin_size;
 	n /= size;
 
-	gtt_obj = kzalloc(n * sizeof(*gtt_obj), GFP_KERNEL);
+	gtt_obj = kcalloc(n, sizeof(*gtt_obj), GFP_KERNEL);
 	if (!gtt_obj) {
 		DRM_ERROR("Failed to allocate %d pointers\n", n);
 		r = 1;
@@ -246,7 +247,7 @@ out_unref:
 out_cleanup:
 	kfree(gtt_obj);
 	if (r) {
-		printk(KERN_WARNING "Error while testing BO move.\n");
+		pr_warn("Error while testing BO move\n");
 	}
 }
 
@@ -298,7 +299,12 @@ static int radeon_test_create_and_emit_fence(struct radeon_device *rdev,
 			DRM_ERROR("Failed to lock ring A %d\n", ring->idx);
 			return r;
 		}
-		radeon_fence_emit(rdev, fence, ring->idx);
+		r = radeon_fence_emit(rdev, fence, ring->idx);
+		if (r) {
+			DRM_ERROR("Failed to emit fence\n");
+			radeon_ring_unlock_undo(rdev, ring);
+			return r;
+		}
 		radeon_ring_unlock_commit(rdev, ring, false);
 	}
 	return 0;
@@ -394,7 +400,7 @@ out_cleanup:
 		radeon_fence_unref(&fence2);
 
 	if (r)
-		printk(KERN_WARNING "Error while testing ring sync (%d).\n", r);
+		pr_warn("Error while testing ring sync (%d)\n", r);
 }
 
 static void radeon_test_ring_sync2(struct radeon_device *rdev,
@@ -504,7 +510,7 @@ out_cleanup:
 		radeon_fence_unref(&fenceB);
 
 	if (r)
-		printk(KERN_WARNING "Error while testing ring sync (%d).\n", r);
+		pr_warn("Error while testing ring sync (%d)\n", r);
 }
 
 static bool radeon_test_sync_possible(struct radeon_ring *ringA,
diff --git a/sys/dev/pci/drm/radeon/radeon_trace.h b/sys/dev/pci/drm/radeon/radeon_trace.h
index a7c4ba700a8..86ee54969d8 100644
--- a/sys/dev/pci/drm/radeon/radeon_trace.h
+++ b/sys/dev/pci/drm/radeon/radeon_trace.h
@@ -2,7 +2,11 @@
 #if !defined(_RADEON_TRACE_H) || defined(TRACE_HEADER_MULTI_READ)
 #define _RADEON_TRACE_H_
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <linux/tracepoint.h>
+
+#include <drm/drmP.h>
 
 #undef TRACE_SYSTEM
 #define TRACE_SYSTEM radeon
diff --git a/sys/dev/pci/drm/radeon/radeon_trace_points.c b/sys/dev/pci/drm/radeon/radeon_trace_points.c
index 246bdb7f513..65e92302f97 100644
--- a/sys/dev/pci/drm/radeon/radeon_trace_points.c
+++ b/sys/dev/pci/drm/radeon/radeon_trace_points.c
@@ -2,8 +2,8 @@
 /* Copyright Red Hat Inc 2010.
  * Author : Dave Airlie <airlied@redhat.com>
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 
 #define CREATE_TRACE_POINTS
diff --git a/sys/dev/pci/drm/radeon/radeon_ttm.c b/sys/dev/pci/drm/radeon/radeon_ttm.c
index faf9f8310e8..d9fb2661c3d 100644
--- a/sys/dev/pci/drm/radeon/radeon_ttm.c
+++ b/sys/dev/pci/drm/radeon/radeon_ttm.c
@@ -29,13 +29,19 @@
  *    Thomas Hellstrom <thomas-at-tungstengraphics-dot-com>
  *    Dave Airlie
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/ttm/ttm_bo_api.h>
-#include <dev/pci/drm/ttm/ttm_bo_driver.h>
-#include <dev/pci/drm/ttm/ttm_placement.h>
-#include <dev/pci/drm/ttm/ttm_module.h>
-#include <dev/pci/drm/ttm/ttm_page_alloc.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_page_alloc.h>
+#include <drm/radeon_drm.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <linux/swiotlb.h>
+#include <linux/swap.h>
+#include <linux/pagemap.h>
+#include <linux/debugfs.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 
@@ -146,7 +152,7 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->available_caching = TTM_PL_MASK_CACHING;
 		man->default_caching = TTM_PL_FLAG_CACHED;
 		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE | TTM_MEMTYPE_FLAG_CMA;
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 		if (rdev->flags & RADEON_IS_AGP) {
 			if (!rdev->ddev->agp) {
 				DRM_ERROR("AGP is not enabled for memory type %u\n",
@@ -180,7 +186,7 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 static void radeon_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
-	static struct ttm_place placements = {
+	static const struct ttm_place placements = {
 		.fpfn = 0,
 		.lpfn = 0,
 		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
@@ -302,8 +308,7 @@ static int radeon_move_blit(struct ttm_buffer_object *bo,
 	if (IS_ERR(fence))
 		return PTR_ERR(fence);
 
-	r = ttm_bo_move_accel_cleanup(bo, &fence->base,
-				      evict, no_wait_gpu, new_mem);
+	r = ttm_bo_move_accel_cleanup(bo, &fence->base, evict, new_mem);
 	radeon_fence_unref(&fence);
 	return r;
 }
@@ -313,6 +318,7 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 				bool no_wait_gpu,
 				struct ttm_mem_reg *new_mem)
 {
+	struct ttm_operation_ctx ctx = { interruptible, no_wait_gpu };
 	struct radeon_device *rdev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
@@ -330,8 +336,7 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 	placements.fpfn = 0;
 	placements.lpfn = 0;
 	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
-	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
-			     interruptible, no_wait_gpu);
+	r = ttm_bo_mem_space(bo, &placement, &tmp_mem, &ctx);
 	if (unlikely(r)) {
 		return r;
 	}
@@ -341,7 +346,7 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 		goto out_cleanup;
 	}
 
-	r = ttm_tt_bind(bo->ttm, &tmp_mem);
+	r = ttm_tt_bind(bo->ttm, &tmp_mem, &ctx);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
@@ -349,7 +354,7 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo,
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
-	r = ttm_bo_move_ttm(bo, true, no_wait_gpu, new_mem);
+	r = ttm_bo_move_ttm(bo, &ctx, new_mem);
 out_cleanup:
 	ttm_bo_mem_put(bo, &tmp_mem);
 	return r;
@@ -360,6 +365,7 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo,
 				bool no_wait_gpu,
 				struct ttm_mem_reg *new_mem)
 {
+	struct ttm_operation_ctx ctx = { interruptible, no_wait_gpu };
 	struct radeon_device *rdev;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	struct ttm_mem_reg tmp_mem;
@@ -377,12 +383,11 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo,
 	placements.fpfn = 0;
 	placements.lpfn = 0;
 	placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
-	r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
-			     interruptible, no_wait_gpu);
+	r = ttm_bo_mem_space(bo, &placement, &tmp_mem, &ctx);
 	if (unlikely(r)) {
 		return r;
 	}
-	r = ttm_bo_move_ttm(bo, true, no_wait_gpu, &tmp_mem);
+	r = ttm_bo_move_ttm(bo, &ctx, &tmp_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
@@ -395,15 +400,24 @@ out_cleanup:
 	return r;
 }
 
-static int radeon_bo_move(struct ttm_buffer_object *bo,
-			bool evict, bool interruptible,
-			bool no_wait_gpu,
-			struct ttm_mem_reg *new_mem)
+static int radeon_bo_move(struct ttm_buffer_object *bo, bool evict,
+			  struct ttm_operation_ctx *ctx,
+			  struct ttm_mem_reg *new_mem)
 {
 	struct radeon_device *rdev;
+	struct radeon_bo *rbo;
 	struct ttm_mem_reg *old_mem = &bo->mem;
 	int r;
 
+	r = ttm_bo_wait(bo, ctx->interruptible, ctx->no_wait_gpu);
+	if (r)
+		return r;
+
+	/* Can't move a pinned BO */
+	rbo = container_of(bo, struct radeon_bo, tbo);
+	if (WARN_ON_ONCE(rbo->pin_count > 0))
+		return -EINVAL;
+
 	rdev = radeon_get_rdev(bo->bdev);
 	if (old_mem->mem_type == TTM_PL_SYSTEM && bo->ttm == NULL) {
 		radeon_move_null(bo, new_mem);
@@ -425,19 +439,20 @@ static int radeon_bo_move(struct ttm_buffer_object *bo,
 
 	if (old_mem->mem_type == TTM_PL_VRAM &&
 	    new_mem->mem_type == TTM_PL_SYSTEM) {
-		r = radeon_move_vram_ram(bo, evict, interruptible,
-					no_wait_gpu, new_mem);
+		r = radeon_move_vram_ram(bo, evict, ctx->interruptible,
+					ctx->no_wait_gpu, new_mem);
 	} else if (old_mem->mem_type == TTM_PL_SYSTEM &&
 		   new_mem->mem_type == TTM_PL_VRAM) {
-		r = radeon_move_ram_vram(bo, evict, interruptible,
-					    no_wait_gpu, new_mem);
+		r = radeon_move_ram_vram(bo, evict, ctx->interruptible,
+					    ctx->no_wait_gpu, new_mem);
 	} else {
-		r = radeon_move_blit(bo, evict, no_wait_gpu, new_mem, old_mem);
+		r = radeon_move_blit(bo, evict, ctx->no_wait_gpu,
+				     new_mem, old_mem);
 	}
 
 	if (r) {
 memcpy:
-		r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+		r = ttm_bo_move_memcpy(bo, ctx, new_mem);
 		if (r) {
 			return r;
 		}
@@ -465,7 +480,7 @@ static int radeon_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_
 		/* system memory */
 		return 0;
 	case TTM_PL_TT:
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 		if (rdev->flags & RADEON_IS_AGP) {
 			/* RADEON_IS_AGP is set only if AGP is active */
 			mem->bus.offset = mem->start << PAGE_SHIFT;
@@ -494,6 +509,8 @@ static int radeon_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_
 			mem->bus.addr =
 				ioremap_nocache(mem->bus.base + mem->bus.offset,
 						mem->bus.size);
+		if (!mem->bus.addr)
+			return -ENOMEM;
 
 		/*
 		 * Alpha: Use just the bus offset plus
@@ -564,8 +581,8 @@ static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 		uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE;
 		struct vm_page **pages = ttm->pages + pinned;
 
-		r = get_user_pages(current, current->mm, userptr, num_pages,
-				   write, 0, pages, NULL);
+		r = get_user_pages(userptr, num_pages, write ? FOLL_WRITE : 0,
+				   pages, NULL);
 		if (r < 0)
 			goto release_pages;
 
@@ -593,7 +610,7 @@ release_sg:
 	kfree(ttm->sg);
 
 release_pages:
-	release_pages(ttm->pages, pinned, 0);
+	release_pages(ttm->pages, pinned);
 	return r;
 #endif
 }
@@ -623,7 +640,7 @@ static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 			set_page_dirty(page);
 
 		mark_page_accessed(page);
-		page_cache_release(page);
+		put_page(page);
 	}
 
 	sg_free_table(ttm->sg);
@@ -688,18 +705,18 @@ static struct ttm_backend_func radeon_backend_func = {
 	.destroy = &radeon_ttm_backend_destroy,
 };
 
-static struct ttm_tt *radeon_ttm_tt_create(struct ttm_bo_device *bdev,
-				    unsigned long size, uint32_t page_flags,
-				    struct vm_page *dummy_read_page)
+static struct ttm_tt *radeon_ttm_tt_create(struct ttm_buffer_object *bo,
+					   uint32_t page_flags)
 {
 	struct radeon_device *rdev;
 	struct radeon_ttm_tt *gtt;
+	unsigned long size = bo->num_pages << PAGE_SHIFT;
 
-	rdev = radeon_get_rdev(bdev);
-#if __OS_HAS_AGP
+	rdev = radeon_get_rdev(bo->bdev);
+#if IS_ENABLED(CONFIG_AGP)
 	if (rdev->flags & RADEON_IS_AGP) {
-		return ttm_agp_tt_create(bdev, rdev->ddev->agp,
-					 size, page_flags, dummy_read_page);
+		return ttm_agp_tt_create(bo, rdev->ddev->agp,
+					 page_flags);
 	}
 #endif
 
@@ -709,7 +726,7 @@ static struct ttm_tt *radeon_ttm_tt_create(struct ttm_bo_device *bdev,
 	}
 	gtt->ttm.ttm.func = &radeon_backend_func;
 	gtt->rdev = rdev;
-	if (ttm_dma_tt_init(&gtt->ttm, bdev, size, page_flags, dummy_read_page)) {
+	if (ttm_dma_tt_init(&gtt->ttm, bo, page_flags)) {
 		kfree(gtt);
 		return NULL;
 	}
@@ -735,18 +752,13 @@ static struct radeon_ttm_tt *radeon_ttm_tt_to_gtt(struct ttm_tt *ttm)
 	return (struct radeon_ttm_tt *)ttm;
 }
 
-static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
+static int radeon_ttm_tt_populate(struct ttm_tt *ttm,
+			struct ttm_operation_ctx *ctx)
 {
 	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 	struct radeon_device *rdev;
-	unsigned i;
-	int r;
-	int seg;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
-	if (ttm->state != tt_unpopulated)
-		return 0;
-
 	if (gtt && gtt->userptr) {
 		ttm->sg = kzalloc(sizeof(struct sg_table), GFP_KERNEL);
 		if (!ttm->sg)
@@ -767,71 +779,25 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm)
 	}
 
 	rdev = radeon_get_rdev(ttm->bdev);
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 	if (rdev->flags & RADEON_IS_AGP) {
-		return ttm_agp_tt_populate(ttm);
+		return ttm_agp_tt_populate(ttm, ctx);
 	}
 #endif
 
 #ifdef CONFIG_SWIOTLB
-	if (swiotlb_nr_tbl()) {
-		return ttm_dma_populate(&gtt->ttm, rdev->dev);
-	}
-#endif
-
-	r = ttm_pool_populate(ttm);
-	if (r) {
-		return r;
-	}
-
-#ifdef __linux__
-	for (i = 0; i < ttm->num_pages; i++) {
-		gtt->ttm.dma_address[i] = pci_map_page(rdev->pdev, ttm->pages[i],
-						       0, PAGE_SIZE,
-						       PCI_DMA_BIDIRECTIONAL);
-		if (pci_dma_mapping_error(rdev->pdev, gtt->ttm.dma_address[i])) {
-			while (i--) {
-				pci_unmap_page(rdev->pdev, gtt->ttm.dma_address[i],
-					       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-				gtt->ttm.dma_address[i] = 0;
-			}
-			ttm_pool_unpopulate(ttm);
-			return -EFAULT;
-		}
-	}
-#else
-	for (i = 0; i < ttm->num_pages; i++) {
-		gtt->segs[i].ds_addr = VM_PAGE_TO_PHYS(ttm->pages[i]);
-		gtt->segs[i].ds_len = PAGE_SIZE;
-	}
-
-	if (bus_dmamap_load_raw(rdev->dmat, gtt->map, gtt->segs,
-				ttm->num_pages,
-				ttm->num_pages * PAGE_SIZE, 0)) {
-		ttm_pool_unpopulate(ttm);
-		return -EFAULT;
+	if (rdev->need_swiotlb && swiotlb_nr_tbl()) {
+		return ttm_dma_populate(&gtt->ttm, rdev->dev, ctx);
 	}
 #endif
 
-	for (seg = 0, i = 0; seg < gtt->map->dm_nsegs; seg++) {
-		bus_addr_t addr = gtt->map->dm_segs[seg].ds_addr;
-		bus_size_t len = gtt->map->dm_segs[seg].ds_len;
-
-		while (len > 0) {
-			gtt->ttm.dma_address[i++] = addr;
-			addr += PAGE_SIZE;
-			len -= PAGE_SIZE;
-		}
-	}
-
-	return 0;
+	return ttm_populate_and_map_pages(rdev->dev, &gtt->ttm, ctx);
 }
 
 static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct radeon_device *rdev;
 	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
-	unsigned i;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
 	if (gtt && gtt->userptr) {
@@ -844,7 +810,7 @@ static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 		return;
 
 	rdev = radeon_get_rdev(ttm->bdev);
-#if __OS_HAS_AGP
+#if IS_ENABLED(CONFIG_AGP)
 	if (rdev->flags & RADEON_IS_AGP) {
 		ttm_agp_tt_unpopulate(ttm);
 		return;
@@ -852,27 +818,13 @@ static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm)
 #endif
 
 #ifdef CONFIG_SWIOTLB
-	if (swiotlb_nr_tbl()) {
+	if (rdev->need_swiotlb && swiotlb_nr_tbl()) {
 		ttm_dma_unpopulate(&gtt->ttm, rdev->dev);
 		return;
 	}
 #endif
 
-#ifdef __linux__
-	for (i = 0; i < ttm->num_pages; i++) {
-		if (gtt->ttm.dma_address[i]) {
-			pci_unmap_page(rdev->pdev, gtt->ttm.dma_address[i],
-				       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-		}
-	}
-#else
-	bus_dmamap_unload(rdev->dmat, gtt->map);
-	for (i = 0; i < ttm->num_pages; i++)
-		gtt->ttm.dma_address[i] = 0;
-
-#endif
-
-	ttm_pool_unpopulate(ttm);
+	ttm_unmap_and_unpopulate_pages(rdev->dev, &gtt->ttm);
 }
 
 int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
@@ -880,7 +832,7 @@ int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 {
 	STUB();
 	return -ENOSYS;
-#ifdef notyet
+#if 0
 	struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm);
 
 	if (gtt == NULL)
@@ -919,6 +871,7 @@ static struct ttm_bo_driver radeon_bo_driver = {
 	.ttm_tt_unpopulate = &radeon_ttm_tt_unpopulate,
 	.invalidate_caches = &radeon_invalidate_caches,
 	.init_mem_type = &radeon_init_mem_type,
+	.eviction_valuable = ttm_bo_eviction_valuable,
 	.evict_flags = &radeon_evict_flags,
 	.move = &radeon_bo_move,
 	.verify_access = &radeon_verify_access,
@@ -979,22 +932,22 @@ int radeon_ttm_init(struct radeon_device *rdev)
 #ifdef __sparc64__
 	r = radeon_bo_create(rdev, rdev->fb_offset, PAGE_SIZE, true,
 			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-			     NULL, &rdev->stollen_vga_memory);
+			     NULL, &rdev->stolen_vga_memory);
 #else
 	r = radeon_bo_create(rdev, stolen_size, PAGE_SIZE, true,
 			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
-			     NULL, &rdev->stollen_vga_memory);
+			     NULL, &rdev->stolen_vga_memory);
 #endif
 	if (r) {
 		return r;
 	}
-	r = radeon_bo_reserve(rdev->stollen_vga_memory, false);
+	r = radeon_bo_reserve(rdev->stolen_vga_memory, false);
 	if (r)
 		return r;
-	r = radeon_bo_pin(rdev->stollen_vga_memory, RADEON_GEM_DOMAIN_VRAM, NULL);
-	radeon_bo_unreserve(rdev->stollen_vga_memory);
+	r = radeon_bo_pin(rdev->stolen_vga_memory, RADEON_GEM_DOMAIN_VRAM, NULL);
+	radeon_bo_unreserve(rdev->stolen_vga_memory);
 	if (r) {
-		radeon_bo_unref(&rdev->stollen_vga_memory);
+		radeon_bo_unref(&rdev->stolen_vga_memory);
 		return r;
 	}
 	DRM_INFO("radeon: %uM of VRAM memory ready\n",
@@ -1023,13 +976,13 @@ void radeon_ttm_fini(struct radeon_device *rdev)
 	if (!rdev->mman.initialized)
 		return;
 	radeon_ttm_debugfs_fini(rdev);
-	if (rdev->stollen_vga_memory) {
-		r = radeon_bo_reserve(rdev->stollen_vga_memory, false);
+	if (rdev->stolen_vga_memory) {
+		r = radeon_bo_reserve(rdev->stolen_vga_memory, false);
 		if (r == 0) {
-			radeon_bo_unpin(rdev->stollen_vga_memory);
-			radeon_bo_unreserve(rdev->stollen_vga_memory);
+			radeon_bo_unpin(rdev->stolen_vga_memory);
+			radeon_bo_unreserve(rdev->stolen_vga_memory);
 		}
-		radeon_bo_unref(&rdev->stollen_vga_memory);
+		radeon_bo_unref(&rdev->stolen_vga_memory);
 	}
 	ttm_bo_clean_mm(&rdev->mman.bdev, TTM_PL_VRAM);
 	ttm_bo_clean_mm(&rdev->mman.bdev, TTM_PL_TT);
@@ -1058,21 +1011,21 @@ void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size)
 static struct vm_operations_struct radeon_ttm_vm_ops;
 static const struct vm_operations_struct *ttm_vm_ops = NULL;
 
-static int radeon_ttm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
+static vm_fault_t radeon_ttm_fault(struct vm_fault *vmf)
 {
 	struct ttm_buffer_object *bo;
 	struct radeon_device *rdev;
-	int r;
+	vm_fault_t ret;
 
-	bo = (struct ttm_buffer_object *)vma->vm_private_data;	
+	bo = (struct ttm_buffer_object *)vmf->vma->vm_private_data;
 	if (bo == NULL) {
 		return VM_FAULT_NOPAGE;
 	}
 	rdev = radeon_get_rdev(bo->bdev);
 	down_read(&rdev->pm.mclk_lock);
-	r = ttm_vm_ops->fault(vma, vmf);
+	ret = ttm_vm_ops->fault(vmf);
 	up_read(&rdev->pm.mclk_lock);
-	return r;
+	return ret;
 }
 
 int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
@@ -1160,19 +1113,17 @@ radeon_mmap(struct drm_device *dev, voff_t off, vsize_t size)
 static int radeon_mm_dump_table(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = (struct drm_info_node *)m->private;
-	unsigned ttm_pl = *(int *)node->info_ent->data;
+	unsigned ttm_pl = *(int*)node->info_ent->data;
 	struct drm_device *dev = node->minor->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct drm_mm *mm = (struct drm_mm *)rdev->mman.bdev.man[ttm_pl].priv;
-	int ret;
-	struct ttm_bo_global *glob = rdev->mman.bdev.glob;
+	struct ttm_mem_type_manager *man = &rdev->mman.bdev.man[ttm_pl];
+	struct drm_printer p = drm_seq_file_printer(m);
 
-	spin_lock(&glob->lru_lock);
-	ret = drm_mm_dump_table(m, mm);
-	spin_unlock(&glob->lru_lock);
-	return ret;
+	man->func->debug(man, &p);
+	return 0;
 }
 
+
 static int ttm_pl_vram = TTM_PL_VRAM;
 static int ttm_pl_tt = TTM_PL_TT;
 
@@ -1268,7 +1219,7 @@ static ssize_t radeon_ttm_gtt_read(struct file *f, char __user *buf,
 			ptr += off;
 
 			r = copy_to_user(buf, ptr, cur_size);
-			kunmap(rdev->gart.pages[p]);
+			kunmap(ptr);
 		} else
 			r = clear_user(buf, cur_size);
 
@@ -1316,7 +1267,7 @@ static int radeon_ttm_debugfs_init(struct radeon_device *rdev)
 	count = ARRAY_SIZE(radeon_ttm_debugfs_list);
 
 #ifdef CONFIG_SWIOTLB
-	if (!swiotlb_nr_tbl())
+	if (!(rdev->need_swiotlb && swiotlb_nr_tbl()))
 		--count;
 #endif
 
diff --git a/sys/dev/pci/drm/radeon/radeon_ucode.c b/sys/dev/pci/drm/radeon/radeon_ucode.c
index ea8fa92f88e..0790bdbe323 100644
--- a/sys/dev/pci/drm/radeon/radeon_ucode.c
+++ b/sys/dev/pci/drm/radeon/radeon_ucode.c
@@ -21,7 +21,10 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_ucode.h"
 
diff --git a/sys/dev/pci/drm/radeon/radeon_uvd.c b/sys/dev/pci/drm/radeon/radeon_uvd.c
index 04cbe5dd8d8..6d56b720c12 100644
--- a/sys/dev/pci/drm/radeon/radeon_uvd.c
+++ b/sys/dev/pci/drm/radeon/radeon_uvd.c
@@ -28,10 +28,13 @@
  *    Christian König <deathsimple@vodafone.de>
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
 
 #include "radeon.h"
+#include "radeon_ucode.h"
 #include "r600d.h"
 
 /* 1 second timeout */
@@ -45,7 +48,8 @@
 #define FIRMWARE_CYPRESS	"radeon/CYPRESS_uvd.bin"
 #define FIRMWARE_SUMO		"radeon/SUMO_uvd.bin"
 #define FIRMWARE_TAHITI		"radeon/TAHITI_uvd.bin"
-#define FIRMWARE_BONAIRE	"radeon/BONAIRE_uvd.bin"
+#define FIRMWARE_BONAIRE_LEGACY	"radeon/BONAIRE_uvd.bin"
+#define FIRMWARE_BONAIRE	"radeon/bonaire_uvd.bin"
 
 MODULE_FIRMWARE(FIRMWARE_R600);
 MODULE_FIRMWARE(FIRMWARE_RS780);
@@ -54,6 +58,7 @@ MODULE_FIRMWARE(FIRMWARE_RV710);
 MODULE_FIRMWARE(FIRMWARE_CYPRESS);
 MODULE_FIRMWARE(FIRMWARE_SUMO);
 MODULE_FIRMWARE(FIRMWARE_TAHITI);
+MODULE_FIRMWARE(FIRMWARE_BONAIRE_LEGACY);
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
 
 static void radeon_uvd_idle_work_handler(struct work_struct *work);
@@ -61,7 +66,7 @@ static void radeon_uvd_idle_work_handler(struct work_struct *work);
 int radeon_uvd_init(struct radeon_device *rdev)
 {
 	unsigned long bo_size;
-	const char *fw_name;
+	const char *fw_name = NULL, *legacy_fw_name = NULL;
 	int i, r;
 
 	INIT_DELAYED_WORK(&rdev->uvd.idle_work, radeon_uvd_idle_work_handler);
@@ -72,22 +77,22 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_RV670:
 	case CHIP_RV620:
 	case CHIP_RV635:
-		fw_name = FIRMWARE_R600;
+		legacy_fw_name = FIRMWARE_R600;
 		break;
 
 	case CHIP_RS780:
 	case CHIP_RS880:
-		fw_name = FIRMWARE_RS780;
+		legacy_fw_name = FIRMWARE_RS780;
 		break;
 
 	case CHIP_RV770:
-		fw_name = FIRMWARE_RV770;
+		legacy_fw_name = FIRMWARE_RV770;
 		break;
 
 	case CHIP_RV710:
 	case CHIP_RV730:
 	case CHIP_RV740:
-		fw_name = FIRMWARE_RV710;
+		legacy_fw_name = FIRMWARE_RV710;
 		break;
 
 	case CHIP_CYPRESS:
@@ -95,7 +100,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_JUNIPER:
 	case CHIP_REDWOOD:
 	case CHIP_CEDAR:
-		fw_name = FIRMWARE_CYPRESS;
+		legacy_fw_name = FIRMWARE_CYPRESS;
 		break;
 
 	case CHIP_SUMO:
@@ -105,7 +110,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_BARTS:
 	case CHIP_TURKS:
 	case CHIP_CAICOS:
-		fw_name = FIRMWARE_SUMO;
+		legacy_fw_name = FIRMWARE_SUMO;
 		break;
 
 	case CHIP_TAHITI:
@@ -113,7 +118,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_PITCAIRN:
 	case CHIP_ARUBA:
 	case CHIP_OLAND:
-		fw_name = FIRMWARE_TAHITI;
+		legacy_fw_name = FIRMWARE_TAHITI;
 		break;
 
 	case CHIP_BONAIRE:
@@ -121,6 +126,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_KAVERI:
 	case CHIP_HAWAII:
 	case CHIP_MULLINS:
+		legacy_fw_name = FIRMWARE_BONAIRE_LEGACY;
 		fw_name = FIRMWARE_BONAIRE;
 		break;
 
@@ -128,16 +134,56 @@ int radeon_uvd_init(struct radeon_device *rdev)
 		return -EINVAL;
 	}
 
-	r = request_firmware(&rdev->uvd_fw, fw_name, rdev->dev);
-	if (r) {
-		dev_err(rdev->dev, "radeon_uvd: Can't load firmware \"%s\"\n",
-			fw_name);
-		return r;
+	rdev->uvd.fw_header_present = false;
+	rdev->uvd.max_handles = RADEON_DEFAULT_UVD_HANDLES;
+	if (fw_name) {
+		/* Let's try to load the newer firmware first */
+		r = request_firmware(&rdev->uvd_fw, fw_name, rdev->dev);
+		if (r) {
+			dev_err(rdev->dev, "radeon_uvd: Can't load firmware \"%s\"\n",
+				fw_name);
+		} else {
+			struct common_firmware_header *hdr = (void *)rdev->uvd_fw->data;
+			unsigned version_major, version_minor, family_id;
+
+			r = radeon_ucode_validate(rdev->uvd_fw);
+			if (r)
+				return r;
+
+			rdev->uvd.fw_header_present = true;
+
+			family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
+			version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
+			version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
+			DRM_INFO("Found UVD firmware Version: %u.%u Family ID: %u\n",
+				 version_major, version_minor, family_id);
+
+			/*
+			 * Limit the number of UVD handles depending on
+			 * microcode major and minor versions.
+			 */
+			if ((version_major >= 0x01) && (version_minor >= 0x37))
+				rdev->uvd.max_handles = RADEON_MAX_UVD_HANDLES;
+		}
+	}
+
+	/*
+	 * In case there is only legacy firmware, or we encounter an error
+	 * while loading the new firmware, we fall back to loading the legacy
+	 * firmware now.
+	 */
+	if (!fw_name || r) {
+		r = request_firmware(&rdev->uvd_fw, legacy_fw_name, rdev->dev);
+		if (r) {
+			dev_err(rdev->dev, "radeon_uvd: Can't load firmware \"%s\"\n",
+				legacy_fw_name);
+			return r;
+		}
 	}
 
 	bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) +
 		  RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE +
-		  RADEON_GPU_PAGE_SIZE;
+		  RADEON_UVD_SESSION_SIZE * rdev->uvd.max_handles;
 	r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
 			     RADEON_GEM_DOMAIN_VRAM, 0, NULL,
 			     NULL, &rdev->uvd.vcpu_bo);
@@ -170,7 +216,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 
 	radeon_bo_unreserve(rdev->uvd.vcpu_bo);
 
-	for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+	for (i = 0; i < rdev->uvd.max_handles; ++i) {
 		atomic_set(&rdev->uvd.handles[i], 0);
 		rdev->uvd.filp[i] = NULL;
 		rdev->uvd.img_size[i] = 0;
@@ -207,7 +253,7 @@ int radeon_uvd_suspend(struct radeon_device *rdev)
 	if (rdev->uvd.vcpu_bo == NULL)
 		return 0;
 
-	for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+	for (i = 0; i < rdev->uvd.max_handles; ++i) {
 		uint32_t handle = atomic_read(&rdev->uvd.handles[i]);
 		if (handle != 0) {
 			struct radeon_fence *fence;
@@ -282,7 +328,7 @@ void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo,
 void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp)
 {
 	int i, r;
-	for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+	for (i = 0; i < rdev->uvd.max_handles; ++i) {
 		uint32_t handle = atomic_read(&rdev->uvd.handles[i]);
 		if (handle != 0 && rdev->uvd.filp[i] == filp) {
 			struct radeon_fence *fence;
@@ -421,7 +467,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 {
 	int32_t *msg, msg_type, handle;
 	unsigned img_size = 0;
-	struct fence *f;
+	struct dma_fence *f;
 	void *ptr;
 
 	int i, r;
@@ -467,7 +513,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 			return r;
 
 		/* try to alloc a new handle */
-		for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+		for (i = 0; i < p->rdev->uvd.max_handles; ++i) {
 			if (atomic_read(&p->rdev->uvd.handles[i]) == handle) {
 				DRM_ERROR("Handle 0x%x already in use!\n", handle);
 				return -EINVAL;
@@ -493,7 +539,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 			return r;
 
 		/* validate the handle */
-		for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+		for (i = 0; i < p->rdev->uvd.max_handles; ++i) {
 			if (atomic_read(&p->rdev->uvd.handles[i]) == handle) {
 				if (p->rdev->uvd.filp[i] != p->filp) {
 					DRM_ERROR("UVD handle collision detected!\n");
@@ -508,7 +554,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
 
 	case 2:
 		/* it's a destroy msg, free the handle */
-		for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i)
+		for (i = 0; i < p->rdev->uvd.max_handles; ++i)
 			atomic_cmpxchg(&p->rdev->uvd.handles[i], handle, 0);
 		radeon_bo_kunmap(bo);
 		return 0;
@@ -623,6 +669,7 @@ static int radeon_uvd_cs_reg(struct radeon_cs_parser *p,
 				return r;
 			break;
 		case UVD_ENGINE_CNTL:
+		case UVD_NO_OP:
 			break;
 		default:
 			DRM_ERROR("Invalid reg 0x%X!\n",
@@ -707,8 +754,10 @@ static int radeon_uvd_send_msg(struct radeon_device *rdev,
 	ib.ptr[3] = addr >> 32;
 	ib.ptr[4] = PACKET0(UVD_GPCOM_VCPU_CMD, 0);
 	ib.ptr[5] = 0;
-	for (i = 6; i < 16; ++i)
-		ib.ptr[i] = PACKET2(0);
+	for (i = 6; i < 16; i += 2) {
+		ib.ptr[i] = PACKET0(UVD_NO_OP, 0);
+		ib.ptr[i+1] = 0;
+	}
 	ib.length_dw = 16;
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
@@ -720,9 +769,11 @@ static int radeon_uvd_send_msg(struct radeon_device *rdev,
 	return r;
 }
 
-/* multiple fence commands without any stream commands in between can
-   crash the vcpu so just try to emmit a dummy create/destroy msg to
-   avoid this */
+/*
+ * multiple fence commands without any stream commands in between can
+ * crash the vcpu so just try to emmit a dummy create/destroy msg to
+ * avoid this
+ */
 int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
 			      uint32_t handle, struct radeon_fence **fence)
 {
@@ -805,7 +856,7 @@ static void radeon_uvd_count_handles(struct radeon_device *rdev,
 	*sd = 0;
 	*hd = 0;
 
-	for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
+	for (i = 0; i < rdev->uvd.max_handles; ++i) {
 		if (!atomic_read(&rdev->uvd.handles[i]))
 			continue;
 
diff --git a/sys/dev/pci/drm/radeon/radeon_vce.c b/sys/dev/pci/drm/radeon/radeon_vce.c
index 94eb667044a..2e7c209e8b0 100644
--- a/sys/dev/pci/drm/radeon/radeon_vce.c
+++ b/sys/dev/pci/drm/radeon/radeon_vce.c
@@ -25,8 +25,10 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/drm.h>
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
 
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -164,7 +166,7 @@ int radeon_vce_init(struct radeon_device *rdev)
 	rdev->vce.fb_version = (10 * (c[0] - '0')) + (c[1] - '0');
 #endif
 
-	DRM_INFO("Found VCE firmware/feedback version %hhd.%hhd.%hhd / %d!\n",
+	DRM_INFO("Found VCE firmware/feedback version %d.%d.%d / %d!\n",
 		 start, mid, end, rdev->vce.fb_version);
 
 	rdev->vce.fw_version = (start << 24) | (mid << 16) | (end << 8);
@@ -208,7 +210,7 @@ int radeon_vce_init(struct radeon_device *rdev)
 	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
 		atomic_set(&rdev->vce.handles[i], 0);
 		rdev->vce.filp[i] = NULL;
-        }
+	}
 
 	return 0;
 }
@@ -431,7 +433,7 @@ int radeon_vce_get_create_msg(struct radeon_device *rdev, int ring,
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
 	if (r) {
-	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 	}
 
 	if (fence)
@@ -488,7 +490,7 @@ int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
 
 	r = radeon_ib_schedule(rdev, &ib, NULL, false);
 	if (r) {
-	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+		DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
 	}
 
 	if (fence)
@@ -811,18 +813,18 @@ int radeon_vce_ring_test(struct radeon_device *rdev, struct radeon_ring *ring)
 	radeon_ring_unlock_commit(rdev, ring, false);
 
 	for (i = 0; i < rdev->usec_timeout; i++) {
-	        if (vce_v1_0_get_rptr(rdev, ring) != rptr)
-	                break;
-	        DRM_UDELAY(1);
+		if (vce_v1_0_get_rptr(rdev, ring) != rptr)
+			break;
+		DRM_UDELAY(1);
 	}
 
 	if (i < rdev->usec_timeout) {
-	        DRM_INFO("ring test on %d succeeded in %d usecs\n",
-	                 ring->idx, i);
+		DRM_INFO("ring test on %d succeeded in %d usecs\n",
+			 ring->idx, i);
 	} else {
-	        DRM_ERROR("radeon: ring %d test failed\n",
-	                  ring->idx);
-	        r = -ETIMEDOUT;
+		DRM_ERROR("radeon: ring %d test failed\n",
+			 ring->idx);
+		r = -ETIMEDOUT;
 	}
 
 	return r;
@@ -852,11 +854,16 @@ int radeon_vce_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		goto error;
 	}
 
-	r = radeon_fence_wait(fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		r = -ETIMEDOUT;
 	} else {
-	        DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+		DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+		r = 0;
 	}
 error:
 	radeon_fence_unref(&fence);
diff --git a/sys/dev/pci/drm/radeon/radeon_vm.c b/sys/dev/pci/drm/radeon/radeon_vm.c
index 6629c071b7f..862fd898196 100644
--- a/sys/dev/pci/drm/radeon/radeon_vm.c
+++ b/sys/dev/pci/drm/radeon/radeon_vm.c
@@ -25,8 +25,8 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
 #include "radeon.h"
 #include "radeon_trace.h"
 
@@ -132,14 +132,14 @@ struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
 	struct radeon_bo_list *list;
 	unsigned i, idx;
 
-	list = drm_malloc_ab(vm->max_pde_used + 2,
-			     sizeof(struct radeon_bo_list));
+	list = kvmalloc_array(vm->max_pde_used + 2,
+			     sizeof(struct radeon_bo_list), GFP_KERNEL);
 	if (!list)
 		return NULL;
 
 	/* add the vm page table to the list */
 	list[0].robj = vm->page_directory;
-	list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
+	list[0].preferred_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 	list[0].tv.bo = &vm->page_directory->tbo;
 	list[0].tv.shared = true;
@@ -151,7 +151,7 @@ struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
 			continue;
 
 		list[idx].robj = vm->page_tables[i].bo;
-		list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
+		list[idx].preferred_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
 		list[idx].tv.bo = &list[idx].robj->tbo;
 		list[idx].tv.shared = true;
@@ -387,6 +387,7 @@ static void radeon_vm_set_pages(struct radeon_device *rdev,
 static int radeon_vm_clear_bo(struct radeon_device *rdev,
 			      struct radeon_bo *bo)
 {
+	struct ttm_operation_ctx ctx = { true, false };
 	struct radeon_ib ib;
 	unsigned entries;
 	uint64_t addr;
@@ -396,7 +397,7 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev,
 	if (r)
 		return r;
 
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 	if (r)
 		goto error_unreserve;
 
@@ -611,15 +612,16 @@ uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr)
  */
 static uint32_t radeon_vm_page_flags(uint32_t flags)
 {
-        uint32_t hw_flags = 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_VALID) ? R600_PTE_VALID : 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_READABLE) ? R600_PTE_READABLE : 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_WRITEABLE) ? R600_PTE_WRITEABLE : 0;
-        if (flags & RADEON_VM_PAGE_SYSTEM) {
-                hw_flags |= R600_PTE_SYSTEM;
-                hw_flags |= (flags & RADEON_VM_PAGE_SNOOPED) ? R600_PTE_SNOOPED : 0;
-        }
-        return hw_flags;
+	uint32_t hw_flags = 0;
+
+	hw_flags |= (flags & RADEON_VM_PAGE_VALID) ? R600_PTE_VALID : 0;
+	hw_flags |= (flags & RADEON_VM_PAGE_READABLE) ? R600_PTE_READABLE : 0;
+	hw_flags |= (flags & RADEON_VM_PAGE_WRITEABLE) ? R600_PTE_WRITEABLE : 0;
+	if (flags & RADEON_VM_PAGE_SYSTEM) {
+		hw_flags |= R600_PTE_SYSTEM;
+		hw_flags |= (flags & RADEON_VM_PAGE_SNOOPED) ? R600_PTE_SNOOPED : 0;
+	}
+	return hw_flags;
 }
 
 /**
@@ -1234,7 +1236,8 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
 	if (!RB_EMPTY_ROOT(&vm->va)) {
 		dev_err(rdev->dev, "still active bo inside vm\n");
 	}
-	rbtree_postorder_for_each_entry_safe(bo_va, tmp, &vm->va, it.rb) {
+	rbtree_postorder_for_each_entry_safe(bo_va, tmp,
+					     &vm->va, it.rb) {
 		interval_tree_remove(&bo_va->it, &vm->va);
 		r = radeon_bo_reserve(bo_va->bo, false);
 		if (!r) {
diff --git a/sys/dev/pci/drm/radeon/rs400.c b/sys/dev/pci/drm/radeon/rs400.c
index 6e2eb342d07..1cebeb945b4 100644
--- a/sys/dev/pci/drm/radeon/rs400.c
+++ b/sys/dev/pci/drm/radeon/rs400.c
@@ -25,7 +25,9 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rs400d.h"
@@ -253,8 +255,8 @@ static void rs400_gpu_init(struct radeon_device *rdev)
 	/* FIXME: is this correct ? */
 	r420_pipes_init(rdev);
 	if (rs400_mc_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "rs400: Failed to wait MC idle while "
-		       "programming pipes. Bad things might happen. %08x\n", RREG32(RADEON_MC_STATUS));
+		pr_warn("rs400: Failed to wait MC idle while programming pipes. Bad things might happen. %08x\n",
+			RREG32(RADEON_MC_STATUS));
 	}
 }
 
diff --git a/sys/dev/pci/drm/radeon/rs600.c b/sys/dev/pci/drm/radeon/rs600.c
index c6c7e2213a8..5bb1b8e4566 100644
--- a/sys/dev/pci/drm/radeon/rs600.c
+++ b/sys/dev/pci/drm/radeon/rs600.c
@@ -35,7 +35,7 @@
  * close to the one of the R600 family (R600 likely being an evolution
  * of the RS600 GART block).
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -110,7 +110,7 @@ void avivo_wait_for_vblank(struct radeon_device *rdev, int crtc)
 	}
 }
 
-void rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base, bool async)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset);
@@ -121,6 +121,8 @@ void rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	WREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset, tmp);
 
 	/* update the scanout addresses */
+	WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset,
+	       async ? AVIVO_D1GRPH_SURFACE_UPDATE_H_RETRACE_EN : 0);
 	WREG32(AVIVO_D1GRPH_SECONDARY_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
 	       (u32)crtc_base);
 	WREG32(AVIVO_D1GRPH_PRIMARY_SURFACE_ADDRESS + radeon_crtc->crtc_offset,
@@ -413,7 +415,8 @@ void rs600_hpd_init(struct radeon_device *rdev)
 		default:
 			break;
 		}
-		enable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			enable |= 1 << radeon_connector->hpd.hpd;
 		radeon_hpd_set_polarity(rdev, radeon_connector->hpd.hpd);
 	}
 	radeon_irq_kms_enable_hpd(rdev, enable);
@@ -439,12 +442,13 @@ void rs600_hpd_fini(struct radeon_device *rdev)
 		default:
 			break;
 		}
-		disable |= 1 << radeon_connector->hpd.hpd;
+		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
+			disable |= 1 << radeon_connector->hpd.hpd;
 	}
 	radeon_irq_kms_disable_hpd(rdev, disable);
 }
 
-int rs600_asic_reset(struct radeon_device *rdev)
+int rs600_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	struct rv515_mc_save save;
 	u32 status, tmp;
diff --git a/sys/dev/pci/drm/radeon/rs690.c b/sys/dev/pci/drm/radeon/rs690.c
index 488341c2021..d772b919fa3 100644
--- a/sys/dev/pci/drm/radeon/rs690.c
+++ b/sys/dev/pci/drm/radeon/rs690.c
@@ -25,7 +25,7 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
@@ -52,8 +52,7 @@ static void rs690_gpu_init(struct radeon_device *rdev)
 	/* FIXME: is this correct ? */
 	r420_pipes_init(rdev);
 	if (rs690_mc_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait MC idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait MC idle while programming pipes. Bad things might happen.\n");
 	}
 }
 
diff --git a/sys/dev/pci/drm/radeon/rs780_dpm.c b/sys/dev/pci/drm/radeon/rs780_dpm.c
index c2a6ddf5434..694b7b3e979 100644
--- a/sys/dev/pci/drm/radeon/rs780_dpm.c
+++ b/sys/dev/pci/drm/radeon/rs780_dpm.c
@@ -22,13 +22,14 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rs780d.h"
 #include "r600_dpm.h"
 #include "rs780_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 static struct igp_ps *rs780_get_ps(struct radeon_ps *rps)
 {
@@ -794,7 +795,7 @@ static int rs780_parse_power_table(struct radeon_device *rdev)
 	union pplib_clock_info *clock_info;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	struct igp_ps *ps;
 
@@ -803,8 +804,9 @@ static int rs780_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/sys/dev/pci/drm/radeon/rv515.c b/sys/dev/pci/drm/radeon/rv515.c
index 995e01221db..ffbd2c006f6 100644
--- a/sys/dev/pci/drm/radeon/rv515.c
+++ b/sys/dev/pci/drm/radeon/rv515.c
@@ -25,7 +25,9 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
 #include "rv515d.h"
 #include "radeon.h"
 #include "radeon_asic.h"
@@ -152,8 +154,7 @@ static void rv515_gpu_init(struct radeon_device *rdev)
 	unsigned pipe_select_current, gb_pipe_select, tmp;
 
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "resetting GPU. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while resetting GPU. Bad things might happen.\n");
 	}
 	rv515_vga_render_disable(rdev);
 	r420_pipes_init(rdev);
@@ -164,12 +165,10 @@ static void rv515_gpu_init(struct radeon_device *rdev)
 	      (((gb_pipe_select >> 8) & 0xF) << 4);
 	WREG32_PLL(0x000D, tmp);
 	if (r100_gui_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait GUI idle while "
-		       "resetting GPU. Bad things might happen.\n");
+		pr_warn("Failed to wait GUI idle while resetting GPU. Bad things might happen.\n");
 	}
 	if (rv515_mc_wait_for_idle(rdev)) {
-		printk(KERN_WARNING "Failed to wait MC idle while "
-		       "programming pipes. Bad things might happen.\n");
+		pr_warn("Failed to wait MC idle while programming pipes. Bad things might happen.\n");
 	}
 }
 
diff --git a/sys/dev/pci/drm/radeon/rv6xx_dpm.c b/sys/dev/pci/drm/radeon/rv6xx_dpm.c
index 6278967ee77..009eaf61a9f 100644
--- a/sys/dev/pci/drm/radeon/rv6xx_dpm.c
+++ b/sys/dev/pci/drm/radeon/rv6xx_dpm.c
@@ -22,13 +22,14 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv6xxd.h"
 #include "r600_dpm.h"
 #include "rv6xx_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 static u32 rv6xx_scale_count_given_unit(struct radeon_device *rdev,
 					u32 unscaled_count, u32 unit);
@@ -208,7 +209,7 @@ static struct rv6xx_sclk_stepping rv6xx_next_vco_step(struct radeon_device *rdev
 
 static bool rv6xx_can_step_post_div(struct radeon_device *rdev,
 				    struct rv6xx_sclk_stepping *cur,
-                                    struct rv6xx_sclk_stepping *target)
+				    struct rv6xx_sclk_stepping *target)
 {
 	return (cur->post_divider > target->post_divider) &&
 		((cur->vco_frequency * target->post_divider) <=
@@ -238,7 +239,7 @@ static bool rv6xx_reached_stepping_target(struct radeon_device *rdev,
 
 static void rv6xx_generate_steps(struct radeon_device *rdev,
 				 u32 low, u32 high,
-                                 u32 start_index, u8 *end_index)
+				 u32 start_index, u8 *end_index)
 {
 	struct rv6xx_sclk_stepping cur;
 	struct rv6xx_sclk_stepping target;
@@ -1355,23 +1356,23 @@ static void rv6xx_set_dpm_event_sources(struct radeon_device *rdev, u32 sources)
 	enum radeon_dpm_event_src dpm_event_src;
 
 	switch (sources) {
-        case 0:
-        default:
+	case 0:
+	default:
 		want_thermal_protection = false;
 		break;
-        case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL):
+	case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_DIGITAL;
 		break;
 
-        case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL):
+	case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_EXTERNAL;
 		break;
 
-        case ((1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL) |
+	case ((1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL) |
 	      (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL)):
-		want_thermal_protection = true;
+			want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL;
 		break;
 	}
@@ -1878,7 +1879,7 @@ static int rv6xx_parse_power_table(struct radeon_device *rdev)
 	union pplib_clock_info *clock_info;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	struct rv6xx_ps *ps;
 
@@ -1887,8 +1888,9 @@ static int rv6xx_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/sys/dev/pci/drm/radeon/rv730_dpm.c b/sys/dev/pci/drm/radeon/rv730_dpm.c
index 6c07a0aa300..38fdb4152e2 100644
--- a/sys/dev/pci/drm/radeon/rv730_dpm.c
+++ b/sys/dev/pci/drm/radeon/rv730_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv730d.h"
 #include "r600_dpm.h"
diff --git a/sys/dev/pci/drm/radeon/rv740_dpm.c b/sys/dev/pci/drm/radeon/rv740_dpm.c
index ace9f804a61..afd597ec508 100644
--- a/sys/dev/pci/drm/radeon/rv740_dpm.c
+++ b/sys/dev/pci/drm/radeon/rv740_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv740d.h"
 #include "r600_dpm.h"
@@ -36,28 +36,28 @@ u32 rv740_get_decoded_reference_divider(u32 encoded_ref)
 	u32 ref = 0;
 
 	switch (encoded_ref) {
-        case 0:
+	case 0:
 		ref = 1;
 		break;
-        case 16:
+	case 16:
 		ref = 2;
 		break;
-        case 17:
+	case 17:
 		ref = 3;
 		break;
-        case 18:
+	case 18:
 		ref = 2;
 		break;
-        case 19:
+	case 19:
 		ref = 3;
 		break;
-        case 20:
+	case 20:
 		ref = 4;
 		break;
-        case 21:
+	case 21:
 		ref = 5;
 		break;
-        default:
+	default:
 		DRM_ERROR("Invalid encoded Reference Divider\n");
 		ref = 0;
 		break;
diff --git a/sys/dev/pci/drm/radeon/rv770.c b/sys/dev/pci/drm/radeon/rv770.c
index f9abd846a25..dccfd5f14a7 100644
--- a/sys/dev/pci/drm/radeon/rv770.c
+++ b/sys/dev/pci/drm/radeon/rv770.c
@@ -25,11 +25,13 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "rv770d.h"
 #include "atom.h"
 #include "avivod.h"
@@ -799,7 +801,7 @@ u32 rv770_get_xclk(struct radeon_device *rdev)
 	return reference_clock;
 }
 
-void rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base, bool async)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset);
@@ -810,6 +812,8 @@ void rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	WREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset, tmp);
 
 	/* update the scanout addresses */
+	WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset,
+	       async ? AVIVO_D1GRPH_SURFACE_UPDATE_H_RETRACE_EN : 0);
 	if (radeon_crtc->crtc_id) {
 		WREG32(D2GRPH_SECONDARY_SURFACE_ADDRESS_HIGH, upper_32_bits(crtc_base));
 		WREG32(D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH, upper_32_bits(crtc_base));
@@ -1679,6 +1683,73 @@ static int rv770_mc_init(struct radeon_device *rdev)
 	return 0;
 }
 
+static void rv770_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails uvd_v2_2_resume() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void rv770_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = uvd_v2_2_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void rv770_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
 static int rv770_startup(struct radeon_device *rdev)
 {
 	struct radeon_ring *ring;
@@ -1721,16 +1792,7 @@ static int rv770_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = uvd_v2_2_resume(rdev);
-	if (!r) {
-		r = radeon_fence_driver_start_ring(rdev,
-						   R600_RING_TYPE_UVD_INDEX);
-		if (r)
-			dev_err(rdev->dev, "UVD fences init error (%d).\n", r);
-	}
-
-	if (r)
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+	rv770_uvd_start(rdev);
 
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
@@ -1770,16 +1832,7 @@ static int rv770_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-	if (ring->ring_size) {
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     RADEON_CP_PACKET2);
-		if (!r)
-			r = uvd_v1_0_init(rdev);
-
-		if (r)
-			DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
-	}
+	rv770_uvd_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -1829,8 +1882,10 @@ int rv770_suspend(struct radeon_device *rdev)
 {
 	radeon_pm_suspend(rdev);
 	radeon_audio_fini(rdev);
-	uvd_v1_0_fini(rdev);
-	radeon_uvd_suspend(rdev);
+	if (rdev->has_uvd) {
+		uvd_v1_0_fini(rdev);
+		radeon_uvd_suspend(rdev);
+	}
 	r700_cp_stop(rdev);
 	r600_dma_stop(rdev);
 	r600_irq_suspend(rdev);
@@ -1915,12 +1970,7 @@ int rv770_init(struct radeon_device *rdev)
 	rdev->ring[R600_RING_TYPE_DMA_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX], 64 * 1024);
 
-	r = radeon_uvd_init(rdev);
-	if (!r) {
-		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
-		r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX],
-			       4096);
-	}
+	rv770_uvd_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -1972,7 +2022,7 @@ static void rv770_pcie_gen2_enable(struct radeon_device *rdev)
 {
 	u32 link_width_cntl, lanes, speed_cntl, tmp;
 	u16 link_cntl2;
-	u32 mask;
+	enum pci_bus_speed max_bus_speed;
 
 	if (radeon_pcie_gen2 == 0)
 		return;
@@ -1987,10 +2037,9 @@ static void rv770_pcie_gen2_enable(struct radeon_device *rdev)
 	if (ASIC_IS_X2(rdev))
 		return;
 
-	if (drm_pcie_get_speed_cap_mask(rdev->ddev, &mask))
-		return;
-
-	if (!(mask & (DRM_PCIE_SPEED_50|DRM_PCIE_SPEED_80)))
+	max_bus_speed = pcie_get_speed_cap(rdev->pdev->bus->self);
+	if ((max_bus_speed != PCIE_SPEED_5_0GT) &&
+		(max_bus_speed != PCIE_SPEED_8_0GT))
 		return;
 
 	DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
diff --git a/sys/dev/pci/drm/radeon/rv770_dma.c b/sys/dev/pci/drm/radeon/rv770_dma.c
index c13bf2cb08e..acff6e09cc4 100644
--- a/sys/dev/pci/drm/radeon/rv770_dma.c
+++ b/sys/dev/pci/drm/radeon/rv770_dma.c
@@ -21,7 +21,7 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv770d.h"
diff --git a/sys/dev/pci/drm/radeon/rv770_dpm.c b/sys/dev/pci/drm/radeon/rv770_dpm.c
index 502d5da13ce..c765ae7ea80 100644
--- a/sys/dev/pci/drm/radeon/rv770_dpm.c
+++ b/sys/dev/pci/drm/radeon/rv770_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv770d.h"
@@ -30,6 +30,7 @@
 #include "rv770_dpm.h"
 #include "cypress_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -344,27 +345,27 @@ static int rv770_encode_yclk_post_div(u32 postdiv, u32 *encoded_postdiv)
 	int ret = 0;
 
 	switch (postdiv) {
-        case 1:
+	case 1:
 		*encoded_postdiv = 0;
 		break;
-        case 2:
+	case 2:
 		*encoded_postdiv = 1;
 		break;
-        case 4:
+	case 4:
 		*encoded_postdiv = 2;
 		break;
-        case 8:
+	case 8:
 		*encoded_postdiv = 3;
 		break;
-        case 16:
+	case 16:
 		*encoded_postdiv = 4;
 		break;
-        default:
+	default:
 		ret = -EINVAL;
 		break;
 	}
 
-    return ret;
+	return ret;
 }
 
 u32 rv770_map_clkf_to_ibias(struct radeon_device *rdev, u32 clkf)
@@ -1174,15 +1175,15 @@ static int rv770_init_smc_table(struct radeon_device *rdev,
 	rv770_populate_smc_mvdd_table(rdev, table);
 
 	switch (rdev->pm.int_thermal_type) {
-        case THERMAL_TYPE_RV770:
-        case THERMAL_TYPE_ADT7473_WITH_INTERNAL:
+	case THERMAL_TYPE_RV770:
+	case THERMAL_TYPE_ADT7473_WITH_INTERNAL:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_INTERNAL;
 		break;
-        case THERMAL_TYPE_NONE:
+	case THERMAL_TYPE_NONE:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_NONE;
 		break;
-        case THERMAL_TYPE_EXTERNAL_GPIO:
-        default:
+	case THERMAL_TYPE_EXTERNAL_GPIO:
+	default:
 		table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_EXTERNAL;
 		break;
 	}
@@ -1566,18 +1567,18 @@ void rv770_reset_smio_status(struct radeon_device *rdev)
 	sw_smio_index =
 		(RREG32(GENERAL_PWRMGT) & SW_SMIO_INDEX_MASK) >> SW_SMIO_INDEX_SHIFT;
 	switch (sw_smio_index) {
-        case 3:
+	case 3:
 		vid_smio_cntl = RREG32(S3_VID_LOWER_SMIO_CNTL);
 		break;
-        case 2:
+	case 2:
 		vid_smio_cntl = RREG32(S2_VID_LOWER_SMIO_CNTL);
 		break;
-        case 1:
+	case 1:
 		vid_smio_cntl = RREG32(S1_VID_LOWER_SMIO_CNTL);
 		break;
-        case 0:
+	case 0:
 		return;
-        default:
+	default:
 		vid_smio_cntl = pi->s0_vid_lower_smio_cntl;
 		break;
 	}
@@ -1816,21 +1817,21 @@ static void rv770_set_dpm_event_sources(struct radeon_device *rdev, u32 sources)
 	enum radeon_dpm_event_src dpm_event_src;
 
 	switch (sources) {
-        case 0:
-        default:
+	case 0:
+	default:
 		want_thermal_protection = false;
 		break;
-        case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL):
+	case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_DIGITAL;
 		break;
 
-        case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL):
+	case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_EXTERNAL;
 		break;
 
-        case ((1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL) |
+	case ((1 << RADEON_DPM_AUTO_THROTTLE_SRC_EXTERNAL) |
 	      (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL)):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL;
@@ -2272,7 +2273,7 @@ int rv7xx_parse_power_table(struct radeon_device *rdev)
 	union pplib_clock_info *clock_info;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	struct rv7xx_ps *ps;
 
@@ -2281,8 +2282,9 @@ int rv7xx_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/sys/dev/pci/drm/radeon/rv770_smc.c b/sys/dev/pci/drm/radeon/rv770_smc.c
index 7ac88f8389f..2b7ddee3984 100644
--- a/sys/dev/pci/drm/radeon/rv770_smc.c
+++ b/sys/dev/pci/drm/radeon/rv770_smc.c
@@ -22,7 +22,8 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv770d.h"
 #include "rv770_dpm.h"
diff --git a/sys/dev/pci/drm/radeon/rv770d.h b/sys/dev/pci/drm/radeon/rv770d.h
index 9ef2064b1c9..0271f4c559a 100644
--- a/sys/dev/pci/drm/radeon/rv770d.h
+++ b/sys/dev/pci/drm/radeon/rv770d.h
@@ -387,6 +387,7 @@
 #define UVD_UDEC_TILING_CONFIG                          0xef40
 #define UVD_UDEC_DB_TILING_CONFIG                       0xef44
 #define UVD_UDEC_DBW_TILING_CONFIG                      0xef48
+#define UVD_NO_OP					0xeffc
 
 #define	GC_USER_SHADER_PIPE_CONFIG			0x8954
 #define		INACTIVE_QD_PIPES(x)				((x) << 8)
diff --git a/sys/dev/pci/drm/radeon/si.c b/sys/dev/pci/drm/radeon/si.c
index f7e77e824b5..a66362d27ef 100644
--- a/sys/dev/pci/drm/radeon/si.c
+++ b/sys/dev/pci/drm/radeon/si.c
@@ -21,11 +21,14 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
-#include <dev/pci/drm/radeon_drm.h>
+#include <drm/radeon_drm.h>
 #include "sid.h"
 #include "atom.h"
 #include "si_blit_shaders.h"
@@ -62,6 +65,7 @@ MODULE_FIRMWARE("radeon/pitcairn_ce.bin");
 MODULE_FIRMWARE("radeon/pitcairn_mc.bin");
 MODULE_FIRMWARE("radeon/pitcairn_rlc.bin");
 MODULE_FIRMWARE("radeon/pitcairn_smc.bin");
+MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/VERDE_pfp.bin");
 MODULE_FIRMWARE("radeon/VERDE_me.bin");
@@ -77,6 +81,7 @@ MODULE_FIRMWARE("radeon/verde_ce.bin");
 MODULE_FIRMWARE("radeon/verde_mc.bin");
 MODULE_FIRMWARE("radeon/verde_rlc.bin");
 MODULE_FIRMWARE("radeon/verde_smc.bin");
+MODULE_FIRMWARE("radeon/verde_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/OLAND_pfp.bin");
 MODULE_FIRMWARE("radeon/OLAND_me.bin");
@@ -92,6 +97,7 @@ MODULE_FIRMWARE("radeon/oland_ce.bin");
 MODULE_FIRMWARE("radeon/oland_mc.bin");
 MODULE_FIRMWARE("radeon/oland_rlc.bin");
 MODULE_FIRMWARE("radeon/oland_smc.bin");
+MODULE_FIRMWARE("radeon/oland_k_smc.bin");
 
 MODULE_FIRMWARE("radeon/HAINAN_pfp.bin");
 MODULE_FIRMWARE("radeon/HAINAN_me.bin");
@@ -107,6 +113,10 @@ MODULE_FIRMWARE("radeon/hainan_ce.bin");
 MODULE_FIRMWARE("radeon/hainan_mc.bin");
 MODULE_FIRMWARE("radeon/hainan_rlc.bin");
 MODULE_FIRMWARE("radeon/hainan_smc.bin");
+MODULE_FIRMWARE("radeon/hainan_k_smc.bin");
+MODULE_FIRMWARE("radeon/banks_k_2_smc.bin");
+
+MODULE_FIRMWARE("radeon/si58_mc.bin");
 
 static u32 si_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh);
 static void si_pcie_gen3_enable(struct radeon_device *rdev);
@@ -129,6 +139,30 @@ static void si_fini_pg(struct radeon_device *rdev);
 static void si_fini_cg(struct radeon_device *rdev);
 static void si_rlc_stop(struct radeon_device *rdev);
 
+static const u32 crtc_offsets[] =
+{
+	EVERGREEN_CRTC0_REGISTER_OFFSET,
+	EVERGREEN_CRTC1_REGISTER_OFFSET,
+	EVERGREEN_CRTC2_REGISTER_OFFSET,
+	EVERGREEN_CRTC3_REGISTER_OFFSET,
+	EVERGREEN_CRTC4_REGISTER_OFFSET,
+	EVERGREEN_CRTC5_REGISTER_OFFSET
+};
+
+static const u32 si_disp_int_status[] =
+{
+	DISP_INTERRUPT_STATUS,
+	DISP_INTERRUPT_STATUS_CONTINUE,
+	DISP_INTERRUPT_STATUS_CONTINUE2,
+	DISP_INTERRUPT_STATUS_CONTINUE3,
+	DISP_INTERRUPT_STATUS_CONTINUE4,
+	DISP_INTERRUPT_STATUS_CONTINUE5
+};
+
+#define DC_HPDx_CONTROL(x)        (DC_HPD1_CONTROL     + (x * 0xc))
+#define DC_HPDx_INT_CONTROL(x)    (DC_HPD1_INT_CONTROL + (x * 0xc))
+#define DC_HPDx_INT_STATUS_REG(x) (DC_HPD1_INT_STATUS  + (x * 0xc))
+
 static const u32 verde_rlc_save_restore_register_list[] =
 {
 	(0x8000 << 16) | (0x98f4 >> 2),
@@ -1304,7 +1338,7 @@ int si_get_allowed_info_register(struct radeon_device *rdev,
  */
 u32 si_get_xclk(struct radeon_device *rdev)
 {
-        u32 reference_clock = rdev->clock.spll.reference_freq;
+	u32 reference_clock = rdev->clock.spll.reference_freq;
 	u32 tmp;
 
 	tmp = RREG32(CG_CLKPIN_CNTL_2);
@@ -1539,7 +1573,7 @@ int si_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data = NULL;
 	const __le32 *new_fw_data = NULL;
-	u32 running, blackout = 0;
+	u32 running;
 	u32 *io_mc_regs = NULL;
 	const __le32 *new_io_mc_regs = NULL;
 	int i, regs_size, ucode_size;
@@ -1590,11 +1624,6 @@ int si_mc_load_microcode(struct radeon_device *rdev)
 	running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK;
 
 	if (running == 0) {
-		if (running) {
-			blackout = RREG32(MC_SHARED_BLACKOUT_CNTL);
-			WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1);
-		}
-
 		/* reset the engine and set to writable */
 		WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
 		WREG32(MC_SEQ_SUP_CNTL, 0x00000010);
@@ -1633,9 +1662,6 @@ int si_mc_load_microcode(struct radeon_device *rdev)
 				break;
 			udelay(1);
 		}
-
-		if (running)
-			WREG32(MC_SHARED_BLACKOUT_CNTL, blackout);
 	}
 
 	return 0;
@@ -1650,6 +1676,9 @@ static int si_init_microcode(struct radeon_device *rdev)
 	char fw_name[30];
 	int err;
 	int new_fw = 0;
+	bool new_smc = false;
+	bool si58_fw = false;
+	bool banks2_fw = false;
 
 	DRM_DEBUG("\n");
 
@@ -1667,6 +1696,10 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_PITCAIRN:
 		chip_name = "PITCAIRN";
+		if ((rdev->pdev->revision == 0x81) &&
+		    ((rdev->pdev->device == 0x6810) ||
+		     (rdev->pdev->device == 0x6811)))
+			new_smc = true;
 		new_chip_name = "pitcairn";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1678,6 +1711,16 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_VERDE:
 		chip_name = "VERDE";
+		if (((rdev->pdev->device == 0x6820) &&
+		     ((rdev->pdev->revision == 0x81) ||
+		      (rdev->pdev->revision == 0x83))) ||
+		    ((rdev->pdev->device == 0x6821) &&
+		     ((rdev->pdev->revision == 0x83) ||
+		      (rdev->pdev->revision == 0x87))) ||
+		    ((rdev->pdev->revision == 0x87) &&
+		     ((rdev->pdev->device == 0x6823) ||
+		      (rdev->pdev->device == 0x682b))))
+			new_smc = true;
 		new_chip_name = "verde";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1689,6 +1732,14 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_OLAND:
 		chip_name = "OLAND";
+		if (((rdev->pdev->revision == 0x81) &&
+		     ((rdev->pdev->device == 0x6600) ||
+		      (rdev->pdev->device == 0x6604) ||
+		      (rdev->pdev->device == 0x6605) ||
+		      (rdev->pdev->device == 0x6610))) ||
+		    ((rdev->pdev->revision == 0x83) &&
+		     (rdev->pdev->device == 0x6610)))
+			new_smc = true;
 		new_chip_name = "oland";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1699,6 +1750,17 @@ static int si_init_microcode(struct radeon_device *rdev)
 		break;
 	case CHIP_HAINAN:
 		chip_name = "HAINAN";
+		if (((rdev->pdev->revision == 0x81) &&
+		     (rdev->pdev->device == 0x6660)) ||
+		    ((rdev->pdev->revision == 0x83) &&
+		     ((rdev->pdev->device == 0x6660) ||
+		      (rdev->pdev->device == 0x6663) ||
+		      (rdev->pdev->device == 0x6665) ||
+		      (rdev->pdev->device == 0x6667))))
+			new_smc = true;
+		else if ((rdev->pdev->revision == 0xc3) &&
+			 (rdev->pdev->device == 0x6665))
+			banks2_fw = true;
 		new_chip_name = "hainan";
 		pfp_req_size = SI_PFP_UCODE_SIZE * 4;
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
@@ -1710,6 +1772,10 @@ static int si_init_microcode(struct radeon_device *rdev)
 	default: BUG();
 	}
 
+	/* this memory configuration requires special firmware */
+	if (((RREG32(MC_SEQ_MISC0) & 0xff000000) >> 24) == 0x58)
+		si58_fw = true;
+
 	DRM_INFO("Loading %s Microcode\n", new_chip_name);
 
 	snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", new_chip_name);
@@ -1720,8 +1786,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->pfp_fw->size != pfp_req_size) {
-			printk(KERN_ERR
-			       "si_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->pfp_fw->size, fw_name);
 			err = -EINVAL;
 			goto out;
@@ -1729,8 +1794,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 	} else {
 		err = radeon_ucode_validate(rdev->pfp_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1746,16 +1810,14 @@ static int si_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->me_fw->size != me_req_size) {
-			printk(KERN_ERR
-			       "si_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->me_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->me_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1771,16 +1833,14 @@ static int si_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->ce_fw->size != ce_req_size) {
-			printk(KERN_ERR
-			       "si_cp: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_cp: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->ce_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->ce_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1796,16 +1856,14 @@ static int si_init_microcode(struct radeon_device *rdev)
 		if (err)
 			goto out;
 		if (rdev->rlc_fw->size != rlc_req_size) {
-			printk(KERN_ERR
-			       "si_rlc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_rlc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->rlc_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->rlc_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1813,7 +1871,10 @@ static int si_init_microcode(struct radeon_device *rdev)
 		}
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", new_chip_name);
+	if (si58_fw)
+		snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin");
+	else
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", new_chip_name);
 	err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
 	if (err) {
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc2.bin", chip_name);
@@ -1826,8 +1887,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		}
 		if ((rdev->mc_fw->size != mc_req_size) &&
 		    (rdev->mc_fw->size != mc2_req_size)) {
-			printk(KERN_ERR
-			       "si_mc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_mc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->mc_fw->size, fw_name);
 			err = -EINVAL;
 		}
@@ -1835,8 +1895,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 	} else {
 		err = radeon_ucode_validate(rdev->mc_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1844,29 +1903,30 @@ static int si_init_microcode(struct radeon_device *rdev)
 		}
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name);
+	if (banks2_fw)
+		snprintf(fw_name, sizeof(fw_name), "radeon/banks_k_2_smc.bin");
+	else if (new_smc)
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_k_smc.bin", new_chip_name);
+	else
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name);
 	err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 	if (err) {
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
 		err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
 		if (err) {
-			printk(KERN_ERR
-			       "smc: error loading firmware \"%s\"\n",
-			       fw_name);
+			pr_err("smc: error loading firmware \"%s\"\n", fw_name);
 			release_firmware(rdev->smc_fw);
 			rdev->smc_fw = NULL;
 			err = 0;
 		} else if (rdev->smc_fw->size != smc_req_size) {
-			printk(KERN_ERR
-			       "si_smc: Bogus length %zu in firmware \"%s\"\n",
+			pr_err("si_smc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->smc_fw->size, fw_name);
 			err = -EINVAL;
 		}
 	} else {
 		err = radeon_ucode_validate(rdev->smc_fw);
 		if (err) {
-			printk(KERN_ERR
-			       "si_cp: validation failed for firmware \"%s\"\n",
+			pr_err("si_cp: validation failed for firmware \"%s\"\n",
 			       fw_name);
 			goto out;
 		} else {
@@ -1877,7 +1937,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 	if (new_fw == 0) {
 		rdev->new_fw = false;
 	} else if (new_fw < 6) {
-		printk(KERN_ERR "si_fw: mixing new and old firmware!\n");
+		pr_err("si_fw: mixing new and old firmware!\n");
 		err = -EINVAL;
 	} else {
 		rdev->new_fw = true;
@@ -1885,8 +1945,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 out:
 	if (err) {
 		if (err != -EINVAL)
-			printk(KERN_ERR
-			       "si_cp: Failed to load firmware \"%s\"\n",
+			pr_err("si_cp: Failed to load firmware \"%s\"\n",
 			       fw_name);
 		release_firmware(rdev->pfp_fw);
 		rdev->pfp_fw = NULL;
@@ -2169,23 +2228,10 @@ static u32 dce6_latency_watermark(struct dce6_wm_params *wm)
 	a.full = dfixed_const(available_bandwidth);
 	b.full = dfixed_const(wm->num_heads);
 	a.full = dfixed_div(a, b);
+	tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512);
+	tmp = min(dfixed_trunc(a), tmp);
 
-	b.full = dfixed_const(mc_latency + 512);
-	c.full = dfixed_const(wm->disp_clk);
-	b.full = dfixed_div(b, c);
-
-	c.full = dfixed_const(dmif_size);
-	b.full = dfixed_div(c, b);
-
-	tmp = min(dfixed_trunc(a), dfixed_trunc(b));
-
-	b.full = dfixed_const(1000);
-	c.full = dfixed_const(wm->disp_clk);
-	b.full = dfixed_div(c, b);
-	c.full = dfixed_const(wm->bytes_per_pixel);
-	b.full = dfixed_mul(b, c);
-
-	lb_fill_bw = min(tmp, dfixed_trunc(b));
+	lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000);
 
 	a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel);
 	b.full = dfixed_const(1000);
@@ -2252,7 +2298,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev,
 	struct drm_display_mode *mode = &radeon_crtc->base.mode;
 	struct dce6_wm_params wm_low, wm_high;
 	u32 dram_channels;
-	u32 pixel_period;
+	u32 active_time;
 	u32 line_time = 0;
 	u32 latency_watermark_a = 0, latency_watermark_b = 0;
 	u32 priority_a_mark = 0, priority_b_mark = 0;
@@ -2262,8 +2308,11 @@ static void dce6_program_watermarks(struct radeon_device *rdev,
 	fixed20_12 a, b, c;
 
 	if (radeon_crtc->base.enabled && num_heads && mode) {
-		pixel_period = 1000000 / (u32)mode->clock;
-		line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535);
+		active_time = (u32) div_u64((u64)mode->crtc_hdisplay * 1000000,
+					    (u32)mode->clock);
+		line_time = (u32) div_u64((u64)mode->crtc_htotal * 1000000,
+					  (u32)mode->clock);
+		line_time = min(line_time, (u32)65535);
 		priority_a_cnt = 0;
 		priority_b_cnt = 0;
 
@@ -2285,7 +2334,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev,
 
 		wm_high.disp_clk = mode->clock;
 		wm_high.src_width = mode->crtc_hdisplay;
-		wm_high.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_high.active_time = active_time;
 		wm_high.blank_time = line_time - wm_high.active_time;
 		wm_high.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -2312,7 +2361,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev,
 
 		wm_low.disp_clk = mode->clock;
 		wm_low.src_width = mode->crtc_hdisplay;
-		wm_low.active_time = mode->crtc_hdisplay * pixel_period;
+		wm_low.active_time = active_time;
 		wm_low.blank_time = line_time - wm_low.active_time;
 		wm_low.interlaced = false;
 		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
@@ -2439,8 +2488,10 @@ void dce6_bandwidth_update(struct radeon_device *rdev)
  */
 static void si_tiling_mode_table_init(struct radeon_device *rdev)
 {
-	const u32 num_tile_mode_states = 32;
-	u32 reg_offset, gb_tile_moden, split_equal_to_row_size;
+	u32 *tile = rdev->config.si.tile_mode_array;
+	const u32 num_tile_mode_states =
+			ARRAY_SIZE(rdev->config.si.tile_mode_array);
+	u32 reg_offset, split_equal_to_row_size;
 
 	switch (rdev->config.si.mem_row_size_in_kb) {
 	case 1:
@@ -2455,491 +2506,442 @@ static void si_tiling_mode_table_init(struct radeon_device *rdev)
 		break;
 	}
 
-	if ((rdev->family == CHIP_TAHITI) ||
-	    (rdev->family == CHIP_PITCAIRN)) {
-		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:  /* non-AA compressed depth or any compressed stencil */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 1:  /* 2xAA/4xAA compressed depth only */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 2:  /* 8xAA compressed depth only */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 3:  /* 2xAA/4xAA compressed depth with stencil (for depth buffer) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 4:  /* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 5:  /* Uncompressed 16bpp depth - and stencil buffer allocated with it */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 6:  /* Uncompressed 32bpp depth - and stencil buffer allocated with it */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 7:  /* Uncompressed 8bpp stencil without depth (drivers typically do not use) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 8:  /* 1D and 1D Array Surfaces */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 9:  /* Displayable maps. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 10:  /* Display 8bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 11:  /* Display 16bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 12:  /* Display 32bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 13:  /* Thin. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 14:  /* Thin 8 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 15:  /* Thin 16 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 16:  /* Thin 32 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 17:  /* Thin 64 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			case 21:  /* 8 bpp PRT. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 22:  /* 16 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 23:  /* 32 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 24:  /* 64 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 25:  /* 128 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
-						 NUM_BANKS(ADDR_SURF_8_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.si.tile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-	} else if ((rdev->family == CHIP_VERDE) ||
-		   (rdev->family == CHIP_OLAND) ||
-		   (rdev->family == CHIP_HAINAN)) {
-		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
-			switch (reg_offset) {
-			case 0:  /* non-AA compressed depth or any compressed stencil */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 1:  /* 2xAA/4xAA compressed depth only */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 2:  /* 8xAA compressed depth only */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 3:  /* 2xAA/4xAA compressed depth with stencil (for depth buffer) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 4:  /* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 5:  /* Uncompressed 16bpp depth - and stencil buffer allocated with it */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 6:  /* Uncompressed 32bpp depth - and stencil buffer allocated with it */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 7:  /* Uncompressed 8bpp stencil without depth (drivers typically do not use) */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 8:  /* 1D and 1D Array Surfaces */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 9:  /* Displayable maps. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 10:  /* Display 8bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 11:  /* Display 16bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 12:  /* Display 32bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 13:  /* Thin. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 14:  /* Thin 8 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 15:  /* Thin 16 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 16:  /* Thin 32 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 17:  /* Thin 64 bpp. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
-						 TILE_SPLIT(split_equal_to_row_size) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 21:  /* 8 bpp PRT. */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 22:  /* 16 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
-				break;
-			case 23:  /* 32 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 24:  /* 64 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
-						 NUM_BANKS(ADDR_SURF_16_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
-				break;
-			case 25:  /* 128 bpp PRT */
-				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
-						 MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
-						 PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
-						 TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
-						 NUM_BANKS(ADDR_SURF_8_BANK) |
-						 BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
-						 BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
-						 MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
-				break;
-			default:
-				gb_tile_moden = 0;
-				break;
-			}
-			rdev->config.si.tile_mode_array[reg_offset] = gb_tile_moden;
-			WREG32(GB_TILE_MODE0 + (reg_offset * 4), gb_tile_moden);
-		}
-	} else
+	for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+		tile[reg_offset] = 0;
+
+	switch(rdev->family) {
+	case CHIP_TAHITI:
+	case CHIP_PITCAIRN:
+		/* non-AA compressed depth or any compressed stencil */
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 2xAA/4xAA compressed depth only */
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 8xAA compressed depth only */
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 2xAA/4xAA compressed depth with stencil (for depth buffer) */
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */
+		tile[4] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Uncompressed 16bpp depth - and stencil buffer allocated with it */
+		tile[5] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Uncompressed 32bpp depth - and stencil buffer allocated with it */
+		tile[6] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* Uncompressed 8bpp stencil without depth (drivers typically do not use) */
+		tile[7] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 1D and 1D Array Surfaces */
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Displayable maps. */
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Display 8bpp. */
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Display 16bpp. */
+		tile[11] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Display 32bpp. */
+		tile[12] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* Thin. */
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin 8 bpp. */
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* Thin 16 bpp. */
+		tile[15] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* Thin 32 bpp. */
+		tile[16] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* Thin 64 bpp. */
+		tile[17] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+		/* 8 bpp PRT. */
+		tile[21] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 16 bpp PRT */
+		tile[22] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 32 bpp PRT */
+		tile[23] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 64 bpp PRT */
+		tile[24] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 128 bpp PRT */
+		tile[25] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
+			   NUM_BANKS(ADDR_SURF_8_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		break;
+
+	case CHIP_VERDE:
+	case CHIP_OLAND:
+	case CHIP_HAINAN:
+		/* non-AA compressed depth or any compressed stencil */
+		tile[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 2xAA/4xAA compressed depth only */
+		tile[1] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 8xAA compressed depth only */
+		tile[2] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 2xAA/4xAA compressed depth with stencil (for depth buffer) */
+		tile[3] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */
+		tile[4] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Uncompressed 16bpp depth - and stencil buffer allocated with it */
+		tile[5] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Uncompressed 32bpp depth - and stencil buffer allocated with it */
+		tile[6] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Uncompressed 8bpp stencil without depth (drivers typically do not use) */
+		tile[7] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 1D and 1D Array Surfaces */
+		tile[8] = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Displayable maps. */
+		tile[9] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Display 8bpp. */
+		tile[10] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* Display 16bpp. */
+		tile[11] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Display 32bpp. */
+		tile[12] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin. */
+		tile[13] = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin 8 bpp. */
+		tile[14] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin 16 bpp. */
+		tile[15] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin 32 bpp. */
+		tile[16] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* Thin 64 bpp. */
+		tile[17] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+			   TILE_SPLIT(split_equal_to_row_size) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 8 bpp PRT. */
+		tile[21] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 16 bpp PRT */
+		tile[22] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+		/* 32 bpp PRT */
+		tile[23] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 64 bpp PRT */
+		tile[24] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+			   NUM_BANKS(ADDR_SURF_16_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+		/* 128 bpp PRT */
+		tile[25] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+			   MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+			   PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+			   TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
+			   NUM_BANKS(ADDR_SURF_8_BANK) |
+			   BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+			   BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+			   MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+
+		for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++)
+			WREG32(GB_TILE_MODE0 + (reg_offset * 4), tile[reg_offset]);
+		break;
+
+	default:
 		DRM_ERROR("unknown asic: 0x%x\n", rdev->family);
+	}
 }
 
 static void si_select_se_sh(struct radeon_device *rdev,
@@ -4078,10 +4080,15 @@ static void si_gpu_pci_config_reset(struct radeon_device *rdev)
 	}
 }
 
-int si_asic_reset(struct radeon_device *rdev)
+int si_asic_reset(struct radeon_device *rdev, bool hard)
 {
 	u32 reset_mask;
 
+	if (hard) {
+		si_gpu_pci_config_reset(rdev);
+		return 0;
+	}
+
 	reset_mask = si_gpu_check_soft_reset(rdev);
 
 	if (reset_mask)
@@ -4403,6 +4410,10 @@ static bool si_vm_reg_valid(u32 reg)
 	if (reg >= 0x28000)
 		return true;
 
+	/* shader regs are also fine */
+	if (reg >= 0xB000 && reg < 0xC000)
+		return true;
+
 	/* check config regs */
 	switch (reg) {
 	case GRBM_GFX_INDEX:
@@ -4431,6 +4442,7 @@ static bool si_vm_reg_valid(u32 reg)
 	case SPI_CONFIG_CNTL:
 	case SPI_CONFIG_CNTL_1:
 	case TA_CNTL_AUX:
+	case TA_CS_BC_BASE_ADDR:
 		return true;
 	default:
 		DRM_ERROR("Invalid register 0x%x in CS\n", reg);
@@ -5931,6 +5943,7 @@ static void si_disable_interrupts(struct radeon_device *rdev)
 
 static void si_disable_interrupt_state(struct radeon_device *rdev)
 {
+	int i;
 	u32 tmp;
 
 	tmp = RREG32(CP_INT_CNTL_RING0) &
@@ -5944,47 +5957,17 @@ static void si_disable_interrupt_state(struct radeon_device *rdev)
 	WREG32(DMA_CNTL + DMA1_REGISTER_OFFSET, tmp);
 	WREG32(GRBM_INT_CNTL, 0);
 	WREG32(SRBM_INT_CNTL, 0);
-	if (rdev->num_crtc >= 2) {
-		WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
-		WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 4) {
-		WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
-		WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
-		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
-	}
-
-	if (rdev->num_crtc >= 2) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
-	}
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(INT_MASK + crtc_offsets[i], 0);
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(GRPH_INT_CONTROL + crtc_offsets[i], 0);
 
 	if (!ASIC_IS_NODCE(rdev)) {
 		WREG32(DAC_AUTODETECT_INT_CONTROL, 0);
 
-		tmp = RREG32(DC_HPD1_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-		tmp = RREG32(DC_HPD2_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-		tmp = RREG32(DC_HPD3_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-		tmp = RREG32(DC_HPD4_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-		tmp = RREG32(DC_HPD5_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-		tmp = RREG32(DC_HPD6_INT_CONTROL) & DC_HPDx_INT_POLARITY;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
+		for (i = 0; i < 6; i++)
+			WREG32_AND(DC_HPDx_INT_CONTROL(i),
+				   DC_HPDx_INT_POLARITY);
 	}
 }
 
@@ -6059,12 +6042,12 @@ static int si_irq_init(struct radeon_device *rdev)
 	return ret;
 }
 
+/* The order we write back each register here is important */
 int si_irq_set(struct radeon_device *rdev)
 {
+	int i;
 	u32 cp_int_cntl;
 	u32 cp_int_cntl1 = 0, cp_int_cntl2 = 0;
-	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
-	u32 hpd1 = 0, hpd2 = 0, hpd3 = 0, hpd4 = 0, hpd5 = 0, hpd6 = 0;
 	u32 grbm_int_cntl = 0;
 	u32 dma_cntl, dma_cntl1;
 	u32 thermal_int = 0;
@@ -6084,15 +6067,6 @@ int si_irq_set(struct radeon_device *rdev)
 	cp_int_cntl = RREG32(CP_INT_CNTL_RING0) &
 		(CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE);
 
-	if (!ASIC_IS_NODCE(rdev)) {
-		hpd1 = RREG32(DC_HPD1_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-		hpd2 = RREG32(DC_HPD2_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-		hpd3 = RREG32(DC_HPD3_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-		hpd4 = RREG32(DC_HPD4_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-		hpd5 = RREG32(DC_HPD5_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-		hpd6 = RREG32(DC_HPD6_INT_CONTROL) & ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
-	}
-
 	dma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET) & ~TRAP_ENABLE;
 	dma_cntl1 = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET) & ~TRAP_ENABLE;
 
@@ -6121,60 +6095,6 @@ int si_irq_set(struct radeon_device *rdev)
 		DRM_DEBUG("si_irq_set: sw int dma1\n");
 		dma_cntl1 |= TRAP_ENABLE;
 	}
-	if (rdev->irq.crtc_vblank_int[0] ||
-	    atomic_read(&rdev->irq.pflip[0])) {
-		DRM_DEBUG("si_irq_set: vblank 0\n");
-		crtc1 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[1] ||
-	    atomic_read(&rdev->irq.pflip[1])) {
-		DRM_DEBUG("si_irq_set: vblank 1\n");
-		crtc2 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[2] ||
-	    atomic_read(&rdev->irq.pflip[2])) {
-		DRM_DEBUG("si_irq_set: vblank 2\n");
-		crtc3 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[3] ||
-	    atomic_read(&rdev->irq.pflip[3])) {
-		DRM_DEBUG("si_irq_set: vblank 3\n");
-		crtc4 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[4] ||
-	    atomic_read(&rdev->irq.pflip[4])) {
-		DRM_DEBUG("si_irq_set: vblank 4\n");
-		crtc5 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.crtc_vblank_int[5] ||
-	    atomic_read(&rdev->irq.pflip[5])) {
-		DRM_DEBUG("si_irq_set: vblank 5\n");
-		crtc6 |= VBLANK_INT_MASK;
-	}
-	if (rdev->irq.hpd[0]) {
-		DRM_DEBUG("si_irq_set: hpd 1\n");
-		hpd1 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[1]) {
-		DRM_DEBUG("si_irq_set: hpd 2\n");
-		hpd2 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[2]) {
-		DRM_DEBUG("si_irq_set: hpd 3\n");
-		hpd3 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[3]) {
-		DRM_DEBUG("si_irq_set: hpd 4\n");
-		hpd4 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[4]) {
-		DRM_DEBUG("si_irq_set: hpd 5\n");
-		hpd5 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
-	if (rdev->irq.hpd[5]) {
-		DRM_DEBUG("si_irq_set: hpd 6\n");
-		hpd6 |= DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN;
-	}
 
 	WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
 	WREG32(CP_INT_CNTL_RING1, cp_int_cntl1);
@@ -6190,45 +6110,23 @@ int si_irq_set(struct radeon_device *rdev)
 		thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW;
 	}
 
-	if (rdev->num_crtc >= 2) {
-		WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, crtc1);
-		WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, crtc2);
-	}
-	if (rdev->num_crtc >= 4) {
-		WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, crtc3);
-		WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, crtc4);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, crtc5);
-		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
+	for (i = 0; i < rdev->num_crtc; i++) {
+		radeon_irq_kms_set_irq_n_enabled(
+		    rdev, INT_MASK + crtc_offsets[i], VBLANK_INT_MASK,
+		    rdev->irq.crtc_vblank_int[i] ||
+		    atomic_read(&rdev->irq.pflip[i]), "vblank", i);
 	}
 
-	if (rdev->num_crtc >= 2) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-	}
-	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-	}
-	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
-		       GRPH_PFLIP_INT_MASK);
-	}
+	for (i = 0; i < rdev->num_crtc; i++)
+		WREG32(GRPH_INT_CONTROL + crtc_offsets[i], GRPH_PFLIP_INT_MASK);
 
 	if (!ASIC_IS_NODCE(rdev)) {
-		WREG32(DC_HPD1_INT_CONTROL, hpd1);
-		WREG32(DC_HPD2_INT_CONTROL, hpd2);
-		WREG32(DC_HPD3_INT_CONTROL, hpd3);
-		WREG32(DC_HPD4_INT_CONTROL, hpd4);
-		WREG32(DC_HPD5_INT_CONTROL, hpd5);
-		WREG32(DC_HPD6_INT_CONTROL, hpd6);
+		for (i = 0; i < 6; i++) {
+			radeon_irq_kms_set_irq_n_enabled(
+			    rdev, DC_HPDx_INT_CONTROL(i),
+			    DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN,
+			    rdev->irq.hpd[i], "HPD", i);
+		}
 	}
 
 	WREG32(CG_THERMAL_INT, thermal_int);
@@ -6239,133 +6137,48 @@ int si_irq_set(struct radeon_device *rdev)
 	return 0;
 }
 
+/* The order we write back each register here is important */
 static inline void si_irq_ack(struct radeon_device *rdev)
 {
-	u32 tmp;
+	int i, j;
+	u32 *disp_int = rdev->irq.stat_regs.evergreen.disp_int;
+	u32 *grph_int = rdev->irq.stat_regs.evergreen.grph_int;
 
 	if (ASIC_IS_NODCE(rdev))
 		return;
 
-	rdev->irq.stat_regs.evergreen.disp_int = RREG32(DISP_INTERRUPT_STATUS);
-	rdev->irq.stat_regs.evergreen.disp_int_cont = RREG32(DISP_INTERRUPT_STATUS_CONTINUE);
-	rdev->irq.stat_regs.evergreen.disp_int_cont2 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE2);
-	rdev->irq.stat_regs.evergreen.disp_int_cont3 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE3);
-	rdev->irq.stat_regs.evergreen.disp_int_cont4 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE4);
-	rdev->irq.stat_regs.evergreen.disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5);
-	rdev->irq.stat_regs.evergreen.d1grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET);
-	rdev->irq.stat_regs.evergreen.d2grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET);
-	if (rdev->num_crtc >= 4) {
-		rdev->irq.stat_regs.evergreen.d3grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET);
-		rdev->irq.stat_regs.evergreen.d4grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET);
-	}
-	if (rdev->num_crtc >= 6) {
-		rdev->irq.stat_regs.evergreen.d5grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET);
-		rdev->irq.stat_regs.evergreen.d6grph_int = RREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.d1grph_int & GRPH_PFLIP_INT_OCCURRED)
-		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-	if (rdev->irq.stat_regs.evergreen.d2grph_int & GRPH_PFLIP_INT_OCCURRED)
-		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-	if (rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VBLANK_INTERRUPT)
-		WREG32(VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VLINE_INTERRUPT)
-		WREG32(VLINE_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VLINE_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VBLANK_INTERRUPT)
-		WREG32(VBLANK_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VBLANK_ACK);
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VLINE_INTERRUPT)
-		WREG32(VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK);
-
-	if (rdev->num_crtc >= 4) {
-		if (rdev->irq.stat_regs.evergreen.d3grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.d4grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VLINE_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VLINE_ACK);
-	}
-
-	if (rdev->num_crtc >= 6) {
-		if (rdev->irq.stat_regs.evergreen.d5grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.d6grph_int & GRPH_PFLIP_INT_OCCURRED)
-			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, GRPH_PFLIP_INT_CLEAR);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VLINE_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VBLANK_INTERRUPT)
-			WREG32(VBLANK_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VBLANK_ACK);
-		if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VLINE_INTERRUPT)
-			WREG32(VLINE_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VLINE_ACK);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_INTERRUPT) {
-		tmp = RREG32(DC_HPD1_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_INTERRUPT) {
-		tmp = RREG32(DC_HPD2_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_INTERRUPT) {
-		tmp = RREG32(DC_HPD3_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_INTERRUPT) {
-		tmp = RREG32(DC_HPD4_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_INTERRUPT) {
-		tmp = RREG32(DC_HPD5_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_INTERRUPT) {
-		tmp = RREG32(DC_HPD6_INT_CONTROL);
-		tmp |= DC_HPDx_INT_ACK;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
-	}
-
-	if (rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD1_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD1_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD2_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD2_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD3_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD3_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD4_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD4_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD5_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD5_INT_CONTROL, tmp);
-	}
-	if (rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_RX_INTERRUPT) {
-		tmp = RREG32(DC_HPD6_INT_CONTROL);
-		tmp |= DC_HPDx_RX_INT_ACK;
-		WREG32(DC_HPD6_INT_CONTROL, tmp);
+	for (i = 0; i < 6; i++) {
+		disp_int[i] = RREG32(si_disp_int_status[i]);
+		if (i < rdev->num_crtc)
+			grph_int[i] = RREG32(GRPH_INT_STATUS + crtc_offsets[i]);
+	}
+
+	/* We write back each interrupt register in pairs of two */
+	for (i = 0; i < rdev->num_crtc; i += 2) {
+		for (j = i; j < (i + 2); j++) {
+			if (grph_int[j] & GRPH_PFLIP_INT_OCCURRED)
+				WREG32(GRPH_INT_STATUS + crtc_offsets[j],
+				       GRPH_PFLIP_INT_CLEAR);
+		}
+
+		for (j = i; j < (i + 2); j++) {
+			if (disp_int[j] & LB_D1_VBLANK_INTERRUPT)
+				WREG32(VBLANK_STATUS + crtc_offsets[j],
+				       VBLANK_ACK);
+			if (disp_int[j] & LB_D1_VLINE_INTERRUPT)
+				WREG32(VLINE_STATUS + crtc_offsets[j],
+				       VLINE_ACK);
+		}
+	}
+
+	for (i = 0; i < 6; i++) {
+		if (disp_int[i] & DC_HPD1_INTERRUPT)
+			WREG32_OR(DC_HPDx_INT_CONTROL(i), DC_HPDx_INT_ACK);
+	}
+
+	for (i = 0; i < 6; i++) {
+		if (disp_int[i] & DC_HPD1_RX_INTERRUPT)
+			WREG32_OR(DC_HPDx_INT_CONTROL(i), DC_HPDx_RX_INT_ACK);
 	}
 }
 
@@ -6427,6 +6240,9 @@ static inline u32 si_get_ih_wptr(struct radeon_device *rdev)
  */
 int si_irq_process(struct radeon_device *rdev)
 {
+	u32 *disp_int = rdev->irq.stat_regs.evergreen.disp_int;
+	u32 crtc_idx, hpd_idx;
+	u32 mask;
 	u32 wptr;
 	u32 rptr;
 	u32 src_id, src_data, ring_id;
@@ -6435,6 +6251,7 @@ int si_irq_process(struct radeon_device *rdev)
 	bool queue_dp = false;
 	bool queue_thermal = false;
 	u32 status, addr;
+	const char *event_name;
 
 	if (!rdev->ih.enabled || rdev->shutdown)
 		return IRQ_NONE;
@@ -6466,184 +6283,44 @@ restart_ih:
 
 		switch (src_id) {
 		case 1: /* D1 vblank/vline */
-			switch (src_data) {
-			case 0: /* D1 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[0]) {
-					drm_handle_vblank(rdev->ddev, 0);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[0]))
-					radeon_crtc_handle_vblank(rdev, 0);
-				rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D1 vblank\n");
-
-				break;
-			case 1: /* D1 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & LB_D1_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D1 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 2: /* D2 vblank/vline */
-			switch (src_data) {
-			case 0: /* D2 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[1]) {
-					drm_handle_vblank(rdev->ddev, 1);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[1]))
-					radeon_crtc_handle_vblank(rdev, 1);
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D2 vblank\n");
-
-				break;
-			case 1: /* D2 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & LB_D2_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D2 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 3: /* D3 vblank/vline */
-			switch (src_data) {
-			case 0: /* D3 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[2]) {
-					drm_handle_vblank(rdev->ddev, 2);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[2]))
-					radeon_crtc_handle_vblank(rdev, 2);
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D3 vblank\n");
-
-				break;
-			case 1: /* D3 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & LB_D3_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D3 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 4: /* D4 vblank/vline */
-			switch (src_data) {
-			case 0: /* D4 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[3]) {
-					drm_handle_vblank(rdev->ddev, 3);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[3]))
-					radeon_crtc_handle_vblank(rdev, 3);
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D4 vblank\n");
-
-				break;
-			case 1: /* D4 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & LB_D4_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D4 vline\n");
-
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
-			break;
 		case 5: /* D5 vblank/vline */
-			switch (src_data) {
-			case 0: /* D5 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+		case 6: /* D6 vblank/vline */
+			crtc_idx = src_id - 1;
 
-				if (rdev->irq.crtc_vblank_int[4]) {
-					drm_handle_vblank(rdev->ddev, 4);
+			if (src_data == 0) { /* vblank */
+				mask = LB_D1_VBLANK_INTERRUPT;
+				event_name = "vblank";
+
+				if (rdev->irq.crtc_vblank_int[crtc_idx]) {
+					drm_handle_vblank(rdev->ddev, crtc_idx);
 					rdev->pm.vblank_sync = true;
 					wake_up(&rdev->irq.vblank_queue);
 				}
-				if (atomic_read(&rdev->irq.pflip[4]))
-					radeon_crtc_handle_vblank(rdev, 4);
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D5 vblank\n");
-
-				break;
-			case 1: /* D5 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & LB_D5_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D5 vline\n");
+				if (atomic_read(&rdev->irq.pflip[crtc_idx])) {
+					radeon_crtc_handle_vblank(rdev,
+								  crtc_idx);
+				}
 
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
+			} else if (src_data == 1) { /* vline */
+				mask = LB_D1_VLINE_INTERRUPT;
+				event_name = "vline";
+			} else {
+				DRM_DEBUG("Unhandled interrupt: %d %d\n",
+					  src_id, src_data);
 				break;
 			}
-			break;
-		case 6: /* D6 vblank/vline */
-			switch (src_data) {
-			case 0: /* D6 vblank */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VBLANK_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				if (rdev->irq.crtc_vblank_int[5]) {
-					drm_handle_vblank(rdev->ddev, 5);
-					rdev->pm.vblank_sync = true;
-					wake_up(&rdev->irq.vblank_queue);
-				}
-				if (atomic_read(&rdev->irq.pflip[5]))
-					radeon_crtc_handle_vblank(rdev, 5);
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT;
-				DRM_DEBUG("IH: D6 vblank\n");
 
-				break;
-			case 1: /* D6 vline */
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & LB_D6_VLINE_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			if (!(disp_int[crtc_idx] & mask)) {
+				DRM_DEBUG("IH: D%d %s - IH event w/o asserted irq bit?\n",
+					  crtc_idx + 1, event_name);
+			}
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VLINE_INTERRUPT;
-				DRM_DEBUG("IH: D6 vline\n");
+			disp_int[crtc_idx] &= ~mask;
+			DRM_DEBUG("IH: D%d %s\n", crtc_idx + 1, event_name);
 
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
 			break;
 		case 8: /* D1 page flip */
 		case 10: /* D2 page flip */
@@ -6656,119 +6333,29 @@ restart_ih:
 				radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
 			break;
 		case 42: /* HPD hotplug */
-			switch (src_data) {
-			case 0:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~DC_HPD1_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD1\n");
-
-				break;
-			case 1:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~DC_HPD2_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD2\n");
-
-				break;
-			case 2:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~DC_HPD3_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD3\n");
-
-				break;
-			case 3:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~DC_HPD4_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD4\n");
-
-				break;
-			case 4:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~DC_HPD5_INTERRUPT;
-				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD5\n");
-
-				break;
-			case 5:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~DC_HPD6_INTERRUPT;
+			if (src_data <= 5) {
+				hpd_idx = src_data;
+				mask = DC_HPD1_INTERRUPT;
 				queue_hotplug = true;
-				DRM_DEBUG("IH: HPD6\n");
-
-				break;
-			case 6:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int & DC_HPD1_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int &= ~DC_HPD1_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 1\n");
-
-				break;
-			case 7:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont & DC_HPD2_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont &= ~DC_HPD2_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 2\n");
-
-				break;
-			case 8:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont2 & DC_HPD3_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~DC_HPD3_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 3\n");
-
-				break;
-			case 9:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont3 & DC_HPD4_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
-
-				rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~DC_HPD4_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 4\n");
-
-				break;
-			case 10:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont4 & DC_HPD5_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+				event_name = "HPD";
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~DC_HPD5_RX_INTERRUPT;
+			} else if (src_data <= 11) {
+				hpd_idx = src_data - 6;
+				mask = DC_HPD1_RX_INTERRUPT;
 				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 5\n");
+				event_name = "HPD_RX";
 
+			} else {
+				DRM_DEBUG("Unhandled interrupt: %d %d\n",
+					  src_id, src_data);
 				break;
-			case 11:
-				if (!(rdev->irq.stat_regs.evergreen.disp_int_cont5 & DC_HPD6_RX_INTERRUPT))
-					DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+			}
 
-				rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~DC_HPD6_RX_INTERRUPT;
-				queue_dp = true;
-				DRM_DEBUG("IH: HPD_RX 6\n");
+			if (!(disp_int[hpd_idx] & mask))
+				DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
 
-				break;
-			default:
-				DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data);
-				break;
-			}
+			disp_int[hpd_idx] &= ~mask;
+			DRM_DEBUG("IH: %s%d\n", event_name, hpd_idx + 1);
 			break;
 		case 96:
 			DRM_ERROR("SRBM_READ_ERROR: 0x%x\n", RREG32(SRBM_READ_ERROR));
@@ -6867,6 +6454,159 @@ restart_ih:
 /*
  * startup/shutdown callbacks
  */
+static void si_uvd_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = radeon_uvd_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD (%d) init.\n", r);
+		/*
+		 * At this point rdev->uvd.vcpu_bo is NULL which trickles down
+		 * to early fails uvd_v2_2_resume() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable uvd here.
+		 */
+		rdev->has_uvd = 0;
+		return;
+	}
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096);
+}
+
+static void si_uvd_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_uvd)
+		return;
+
+	r = uvd_v2_2_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed UVD resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
+}
+
+static void si_uvd_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_uvd || !rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
+		return;
+	}
+	r = uvd_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing UVD (%d).\n", r);
+		return;
+	}
+}
+
+static void si_vce_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE (%d) init.\n", r);
+		/*
+		 * At this point rdev->vce.vcpu_bo is NULL which trickles down
+		 * to early fails si_vce_start() and thus nothing happens
+		 * there. So it is pointless to try to go through that code
+		 * hence why we disable vce here.
+		 */
+		rdev->has_vce = 0;
+		return;
+	}
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE1_INDEX], 4096);
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_obj = NULL;
+	r600_ring_init(rdev, &rdev->ring[TN_RING_TYPE_VCE2_INDEX], 4096);
+}
+
+static void si_vce_start(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->has_vce)
+		return;
+
+	r = radeon_vce_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = vce_v1_0_resume(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed VCE resume (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE1_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 fences (%d).\n", r);
+		goto error;
+	}
+	r = radeon_fence_driver_start_ring(rdev, TN_RING_TYPE_VCE2_INDEX);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE2 fences (%d).\n", r);
+		goto error;
+	}
+	return;
+
+error:
+	rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
+	rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
+}
+
+static void si_vce_resume(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	if (!rdev->has_vce || !rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size)
+		return;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, VCE_CMD_NO_OP);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	r = radeon_ring_init(rdev, ring, ring->ring_size, 0, VCE_CMD_NO_OP);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE1 ring (%d).\n", r);
+		return;
+	}
+	r = vce_v1_0_init(rdev);
+	if (r) {
+		dev_err(rdev->dev, "failed initializing VCE (%d).\n", r);
+		return;
+	}
+}
+
 static int si_startup(struct radeon_device *rdev)
 {
 	struct radeon_ring *ring;
@@ -6945,33 +6685,8 @@ static int si_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	if (rdev->has_uvd) {
-		r = uvd_v2_2_resume(rdev);
-		if (!r) {
-			r = radeon_fence_driver_start_ring(rdev,
-							   R600_RING_TYPE_UVD_INDEX);
-			if (r)
-				dev_err(rdev->dev, "UVD fences init error (%d).\n", r);
-		}
-		if (r)
-			rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
-	}
-
-	r = radeon_vce_resume(rdev);
-	if (!r) {
-		r = vce_v1_0_resume(rdev);
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE1_INDEX);
-		if (!r)
-			r = radeon_fence_driver_start_ring(rdev,
-							   TN_RING_TYPE_VCE2_INDEX);
-	}
-	if (r) {
-		dev_err(rdev->dev, "VCE init error (%d).\n", r);
-		rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
-		rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
-	}
+	si_uvd_start(rdev);
+	si_vce_start(rdev);
 
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
@@ -7029,34 +6744,8 @@ static int si_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	if (rdev->has_uvd) {
-		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-		if (ring->ring_size) {
-			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-					     RADEON_CP_PACKET2);
-			if (!r)
-				r = uvd_v1_0_init(rdev);
-			if (r)
-				DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
-		}
-	}
-
-	r = -ENOENT;
-
-	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-	if (ring->ring_size)
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     VCE_CMD_NO_OP);
-
-	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-	if (ring->ring_size)
-		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     VCE_CMD_NO_OP);
-
-	if (!r)
-		r = vce_v1_0_init(rdev);
-	else if (r != -ENOENT)
-		DRM_ERROR("radeon: failed initializing VCE (%d).\n", r);
+	si_uvd_resume(rdev);
+	si_vce_resume(rdev);
 
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
@@ -7116,8 +6805,9 @@ int si_suspend(struct radeon_device *rdev)
 	if (rdev->has_uvd) {
 		uvd_v1_0_fini(rdev);
 		radeon_uvd_suspend(rdev);
-		radeon_vce_suspend(rdev);
 	}
+	if (rdev->has_vce)
+		radeon_vce_suspend(rdev);
 	si_fini_pg(rdev);
 	si_fini_cg(rdev);
 	si_irq_suspend(rdev);
@@ -7215,25 +6905,8 @@ int si_init(struct radeon_device *rdev)
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 64 * 1024);
 
-	if (rdev->has_uvd) {
-		r = radeon_uvd_init(rdev);
-		if (!r) {
-			ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
-			ring->ring_obj = NULL;
-			r600_ring_init(rdev, ring, 4096);
-		}
-	}
-
-	r = radeon_vce_init(rdev);
-	if (!r) {
-		ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-
-		ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
-		ring->ring_obj = NULL;
-		r600_ring_init(rdev, ring, 4096);
-	}
+	si_uvd_init(rdev);
+	si_vce_init(rdev);
 
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
@@ -7286,8 +6959,9 @@ void si_fini(struct radeon_device *rdev)
 	if (rdev->has_uvd) {
 		uvd_v1_0_fini(rdev);
 		radeon_uvd_fini(rdev);
-		radeon_vce_fini(rdev);
 	}
+	if (rdev->has_vce)
+		radeon_vce_fini(rdev);
 	si_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
 	radeon_gem_fini(rdev);
@@ -7313,7 +6987,7 @@ uint64_t si_get_gpu_clock_counter(struct radeon_device *rdev)
 	mutex_lock(&rdev->gpu_clock_mutex);
 	WREG32(RLC_CAPTURE_GPU_CLOCK_COUNT, 1);
 	clock = (uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_LSB) |
-	        ((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+		((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
 	mutex_unlock(&rdev->gpu_clock_mutex);
 	return clock;
 }
@@ -7409,17 +7083,13 @@ int si_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 
 static void si_pcie_gen3_enable(struct radeon_device *rdev)
 {
-	struct pci_dev _root;
-	struct pci_dev *root;
+	struct pci_dev *root = rdev->pdev->bus->self;
+	enum pci_bus_speed speed_cap;
 	int bridge_pos, gpu_pos;
-	u32 speed_cntl, mask, current_data_rate;
-	int ret, i;
+	u32 speed_cntl, current_data_rate;
+	int i;
 	u16 tmp16;
 
-	root = &_root;
-	root->pc = rdev->pdev->pc;
-	root->tag = *rdev->ddev->bridgetag;
-
 	if (pci_is_root_bus(rdev->pdev->bus))
 		return;
 
@@ -7432,23 +7102,24 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret != 0)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN)
 		return;
 
-	if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+	if ((speed_cap != PCIE_SPEED_8_0GT) &&
+	    (speed_cap != PCIE_SPEED_5_0GT))
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
 		LC_CURRENT_DATA_RATE_SHIFT;
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		if (current_data_rate == 2) {
 			DRM_INFO("PCIE gen 3 link speeds already enabled\n");
 			return;
 		}
 		DRM_INFO("enabling PCIE gen 3 link speeds, disable with radeon.pcie_gen2=0\n");
-	} else if (mask & DRM_PCIE_SPEED_50) {
+	} else if (speed_cap == PCIE_SPEED_5_0GT) {
 		if (current_data_rate == 1) {
 			DRM_INFO("PCIE gen 2 link speeds already enabled\n");
 			return;
@@ -7464,7 +7135,7 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!gpu_pos)
 		return;
 
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		/* re-try equalization if gen3 is not already enabled */
 		if (current_data_rate != 2) {
 			u16 bridge_cfg, gpu_cfg;
@@ -7552,9 +7223,9 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 
 	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
 	tmp16 &= ~0xf;
-	if (mask & DRM_PCIE_SPEED_80)
+	if (speed_cap == PCIE_SPEED_8_0GT)
 		tmp16 |= 3; /* gen3 */
-	else if (mask & DRM_PCIE_SPEED_50)
+	else if (speed_cap == PCIE_SPEED_5_0GT)
 		tmp16 |= 2; /* gen2 */
 	else
 		tmp16 |= 1; /* gen1 */
@@ -7702,13 +7373,8 @@ static void si_program_aspm(struct radeon_device *rdev)
 
 			if (!disable_clkreq &&
 			    !pci_is_root_bus(rdev->pdev->bus)) {
+				struct pci_dev *root = rdev->pdev->bus->self;
 				u32 lnkcap;
-				struct pci_dev _root;
-				struct pci_dev *root;
-				
-				root = &_root;
-				root->pc = rdev->pdev->pc;
-				root->tag = *rdev->ddev->bridgetag;
 
 				clk_req_support = false;
 				pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap);
@@ -7782,35 +7448,35 @@ static void si_program_aspm(struct radeon_device *rdev)
 	}
 }
 
-int si_vce_send_vcepll_ctlreq(struct radeon_device *rdev)
+static int si_vce_send_vcepll_ctlreq(struct radeon_device *rdev)
 {
-        unsigned i;
+	unsigned i;
 
-        /* make sure VCEPLL_CTLREQ is deasserted */
-        WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, 0, ~UPLL_CTLREQ_MASK);
+	/* make sure VCEPLL_CTLREQ is deasserted */
+	WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, 0, ~UPLL_CTLREQ_MASK);
 
-        mdelay(10);
+	mdelay(10);
 
-        /* assert UPLL_CTLREQ */
-        WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, UPLL_CTLREQ_MASK, ~UPLL_CTLREQ_MASK);
+	/* assert UPLL_CTLREQ */
+	WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, UPLL_CTLREQ_MASK, ~UPLL_CTLREQ_MASK);
 
-        /* wait for CTLACK and CTLACK2 to get asserted */
-        for (i = 0; i < 100; ++i) {
-                uint32_t mask = UPLL_CTLACK_MASK | UPLL_CTLACK2_MASK;
-                if ((RREG32_SMC(CG_VCEPLL_FUNC_CNTL) & mask) == mask)
-                        break;
-                mdelay(10);
-        }
+	/* wait for CTLACK and CTLACK2 to get asserted */
+	for (i = 0; i < 100; ++i) {
+		uint32_t mask = UPLL_CTLACK_MASK | UPLL_CTLACK2_MASK;
+		if ((RREG32_SMC(CG_VCEPLL_FUNC_CNTL) & mask) == mask)
+			break;
+		mdelay(10);
+	}
 
-        /* deassert UPLL_CTLREQ */
-        WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, 0, ~UPLL_CTLREQ_MASK);
+	/* deassert UPLL_CTLREQ */
+	WREG32_SMC_P(CG_VCEPLL_FUNC_CNTL, 0, ~UPLL_CTLREQ_MASK);
 
-        if (i == 100) {
-                DRM_ERROR("Timeout setting UVD clocks!\n");
-                return -ETIMEDOUT;
-        }
+	if (i == 100) {
+		DRM_ERROR("Timeout setting UVD clocks!\n");
+		return -ETIMEDOUT;
+	}
 
-        return 0;
+	return 0;
 }
 
 int si_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk)
diff --git a/sys/dev/pci/drm/radeon/si_blit_shaders.c b/sys/dev/pci/drm/radeon/si_blit_shaders.c
index 7e1ff00997d..ec415e7dfa4 100644
--- a/sys/dev/pci/drm/radeon/si_blit_shaders.c
+++ b/sys/dev/pci/drm/radeon/si_blit_shaders.c
@@ -24,7 +24,9 @@
  *     Alex Deucher <alexander.deucher@amd.com>
  */
 
-#include <dev/pci/drm/drm_linux.h>
+#include <linux/types.h>
+#include <linux/bug.h>
+#include <linux/kernel.h>
 
 const u32 si_default_state[] =
 {
diff --git a/sys/dev/pci/drm/radeon/si_dma.c b/sys/dev/pci/drm/radeon/si_dma.c
index 9b2ab3044fa..83207929fc6 100644
--- a/sys/dev/pci/drm/radeon/si_dma.c
+++ b/sys/dev/pci/drm/radeon/si_dma.c
@@ -21,7 +21,7 @@
  *
  * Authors: Alex Deucher
  */
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_trace.h"
diff --git a/sys/dev/pci/drm/radeon/si_dpm.c b/sys/dev/pci/drm/radeon/si_dpm.c
index 688f90bde23..0a785ef0ab6 100644
--- a/sys/dev/pci/drm/radeon/si_dpm.c
+++ b/sys/dev/pci/drm/radeon/si_dpm.c
@@ -21,13 +21,15 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "sid.h"
 #include "r600_dpm.h"
 #include "si_dpm.h"
 #include "atom.h"
+#include <linux/math64.h>
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -497,7 +499,7 @@ static const struct si_cac_config_reg lcac_pitcairn[] =
 
 static const struct si_cac_config_reg cac_override_pitcairn[] =
 {
-    { 0xFFFFFFFF }
+	{ 0xFFFFFFFF }
 };
 
 static const struct si_powertune_data powertune_data_pitcairn =
@@ -989,7 +991,7 @@ static const struct si_cac_config_reg lcac_cape_verde[] =
 
 static const struct si_cac_config_reg cac_override_cape_verde[] =
 {
-    { 0xFFFFFFFF }
+	{ 0xFFFFFFFF }
 };
 
 static const struct si_powertune_data powertune_data_cape_verde =
@@ -1760,9 +1762,9 @@ static void si_fan_ctrl_set_default_mode(struct radeon_device *rdev);
 
 static struct si_power_info *si_get_pi(struct radeon_device *rdev)
 {
-        struct si_power_info *pi = rdev->pm.dpm.priv;
+	struct si_power_info *pi = rdev->pm.dpm.priv;
 
-        return pi;
+	return pi;
 }
 
 static void si_calculate_leakage_for_v_and_t_formula(const struct ni_leakage_coeffients *coeff,
@@ -2910,29 +2912,6 @@ static int si_init_smc_spll_table(struct radeon_device *rdev)
 	return ret;
 }
 
-struct si_dpm_quirk {
-	u32 chip_vendor;
-	u32 chip_device;
-	u32 subsys_vendor;
-	u32 subsys_device;
-	u32 max_sclk;
-	u32 max_mclk;
-};
-
-/* cards with dpm stability problems */
-static struct si_dpm_quirk si_dpm_quirk_list[] = {
-	/* PITCAIRN - https://bugs.freedesktop.org/show_bug.cgi?id=76490 */
-	{ PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6811, 0x174b, 0xe271, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6811, 0x174b, 0x2015, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6810, 0x174b, 0xe271, 85000, 90000 },
-	{ PCI_VENDOR_ID_ATI, 0x6811, 0x1462, 0x2015, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6811, 0x1043, 0x2015, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6811, 0x148c, 0x2015, 0, 120000 },
-	{ PCI_VENDOR_ID_ATI, 0x6810, 0x1682, 0x9275, 0, 120000 },
-	{ 0, 0, 0, 0 },
-};
-
 static u16 si_get_lower_of_leakage_and_vce_voltage(struct radeon_device *rdev,
 						   u16 vce_voltage)
 {
@@ -2995,29 +2974,8 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	u32 max_sclk = 0, max_mclk = 0;
 	int i;
-	struct si_dpm_quirk *p = si_dpm_quirk_list;
 
-	/* limit all SI kickers */
-	if (rdev->family == CHIP_PITCAIRN) {
-		if ((rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->device == 0x6810) ||
-		    (rdev->pdev->device == 0x6811) ||
-		    (rdev->pdev->device == 0x6816) ||
-		    (rdev->pdev->device == 0x6817) ||
-		    (rdev->pdev->device == 0x6806))
-			max_mclk = 120000;
-	} else if (rdev->family == CHIP_OLAND) {
-		if ((rdev->pdev->revision == 0xC7) ||
-		    (rdev->pdev->revision == 0x80) ||
-		    (rdev->pdev->revision == 0x81) ||
-		    (rdev->pdev->revision == 0x83) ||
-		    (rdev->pdev->revision == 0x87) ||
-		    (rdev->pdev->device == 0x6604) ||
-		    (rdev->pdev->device == 0x6605)) {
-			max_sclk = 75000;
-			max_mclk = 80000;
-		}
-	} else if (rdev->family == CHIP_HAINAN) {
+	if (rdev->family == CHIP_HAINAN) {
 		if ((rdev->pdev->revision == 0x81) ||
 		    (rdev->pdev->revision == 0x83) ||
 		    (rdev->pdev->revision == 0xC3) ||
@@ -3025,6 +2983,10 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		    (rdev->pdev->device == 0x6665) ||
 		    (rdev->pdev->device == 0x6667)) {
 			max_sclk = 75000;
+		}
+		if ((rdev->pdev->revision == 0xC3) ||
+		    (rdev->pdev->device == 0x6665)) {
+			max_sclk = 60000;
 			max_mclk = 80000;
 		}
 	} else if (rdev->family == CHIP_OLAND) {
@@ -3038,18 +3000,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 			max_sclk = 75000;
 		}
 	}
-	/* Apply dpm quirks */
-	while (p && p->chip_device != 0) {
-		if (rdev->pdev->vendor == p->chip_vendor &&
-		    rdev->pdev->device == p->chip_device &&
-		    rdev->pdev->subsystem_vendor == p->subsys_vendor &&
-		    rdev->pdev->subsystem_device == p->subsys_device) {
-			max_sclk = p->max_sclk;
-			max_mclk = p->max_mclk;
-			break;
-		}
-		++p;
-	}
 
 	if (rps->vce_active) {
 		rps->evclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].evclk;
@@ -3192,9 +3142,9 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-        for (i = 0; i < ps->performance_level_count; i++)
-                btc_adjust_clock_combinations(rdev, max_limits,
-                                              &ps->performance_levels[i]);
+	for (i = 0; i < ps->performance_level_count; i++)
+		btc_adjust_clock_combinations(rdev, max_limits,
+					      &ps->performance_levels[i]);
 
 	for (i = 0; i < ps->performance_level_count; i++) {
 		if (ps->performance_levels[i].vddc < min_vce_voltage)
@@ -3333,7 +3283,7 @@ static void si_set_dpm_event_sources(struct radeon_device *rdev, u32 sources)
 	case 0:
 	default:
 		want_thermal_protection = false;
-                break;
+		break;
 	case (1 << RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL):
 		want_thermal_protection = true;
 		dpm_event_src = RADEON_DPM_EVENT_SRC_DIGITAL;
@@ -3535,7 +3485,7 @@ static int si_process_firmware_header(struct radeon_device *rdev)
 	if (ret)
 		return ret;
 
-        si_pi->state_table_start = tmp;
+	si_pi->state_table_start = tmp;
 
 	ret = si_read_smc_sram_dword(rdev,
 				     SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
@@ -3694,7 +3644,7 @@ static void si_program_response_times(struct radeon_device *rdev)
 	si_write_smc_soft_register(rdev, SI_SMC_SOFT_REGISTER_mvdd_chg_time, 1);
 
 	voltage_response_time = (u32)rdev->pm.dpm.voltage_response_time;
-        backbias_response_time = (u32)rdev->pm.dpm.backbias_response_time;
+	backbias_response_time = (u32)rdev->pm.dpm.backbias_response_time;
 
 	if (voltage_response_time == 0)
 		voltage_response_time = 1000;
@@ -3802,7 +3752,7 @@ static void si_setup_bsp(struct radeon_device *rdev)
 			       &pi->pbsu);
 
 
-        pi->dsp = BSP(pi->bsp) | BSU(pi->bsu);
+	pi->dsp = BSP(pi->bsp) | BSU(pi->bsu);
 	pi->psp = BSP(pi->pbsp) | BSU(pi->pbsu);
 
 	WREG32(CG_BSP, pi->dsp);
@@ -4350,7 +4300,7 @@ static int si_populate_memory_timing_parameters(struct radeon_device *rdev,
 
 	radeon_atom_set_engine_dram_timings(rdev,
 					    pl->sclk,
-                                            pl->mclk);
+					    pl->mclk);
 
 	dram_timing  = RREG32(MC_ARB_DRAM_TIMING);
 	dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
@@ -4385,7 +4335,7 @@ static int si_do_program_memory_timing_parameters(struct radeon_device *rdev,
 					   si_pi->sram_end);
 		if (ret)
 			break;
-        }
+	}
 
 	return ret;
 }
@@ -4863,9 +4813,9 @@ static int si_calculate_sclk_params(struct radeon_device *rdev,
 	spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK;
 	spll_func_cntl_2 |= SCLK_MUX_SEL(2);
 
-        spll_func_cntl_3 &= ~SPLL_FB_DIV_MASK;
-        spll_func_cntl_3 |= SPLL_FB_DIV(fbdiv);
-        spll_func_cntl_3 |= SPLL_DITHEN;
+	spll_func_cntl_3 &= ~SPLL_FB_DIV_MASK;
+	spll_func_cntl_3 |= SPLL_FB_DIV(fbdiv);
+	spll_func_cntl_3 |= SPLL_DITHEN;
 
 	if (pi->sclk_ss) {
 		struct radeon_atom_ss ss;
@@ -4972,15 +4922,15 @@ static int si_populate_mclk_value(struct radeon_device *rdev,
 		tmp = freq_nom / reference_clock;
 		tmp = tmp * tmp;
 		if (radeon_atombios_get_asic_ss_info(rdev, &ss,
-                                                     ASIC_INTERNAL_MEMORY_SS, freq_nom)) {
+						     ASIC_INTERNAL_MEMORY_SS, freq_nom)) {
 			u32 clks = reference_clock * 5 / ss.rate;
 			u32 clkv = (u32)((((131 * ss.percentage * ss.rate) / 100) * tmp) / freq_nom);
 
-                        mpll_ss1 &= ~CLKV_MASK;
-                        mpll_ss1 |= CLKV(clkv);
+			mpll_ss1 &= ~CLKV_MASK;
+			mpll_ss1 |= CLKV(clkv);
 
-                        mpll_ss2 &= ~CLKS_MASK;
-                        mpll_ss2 |= CLKS(clks);
+			mpll_ss2 &= ~CLKS_MASK;
+			mpll_ss2 |= CLKS(clks);
 		}
 	}
 
@@ -5307,7 +5257,7 @@ static int si_convert_power_state_to_smc(struct radeon_device *rdev,
 		ni_pi->enable_power_containment = false;
 
 	ret = si_populate_sq_ramping_values(rdev, radeon_state, smc_state);
-        if (ret)
+	if (ret)
 		ni_pi->enable_sq_ramping = false;
 
 	return si_populate_smc_t(rdev, radeon_state, smc_state);
@@ -5478,46 +5428,46 @@ static bool si_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg)
 	case  MC_SEQ_RAS_TIMING >> 2:
 		*out_reg = MC_SEQ_RAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_CAS_TIMING >> 2:
+	case MC_SEQ_CAS_TIMING >> 2:
 		*out_reg = MC_SEQ_CAS_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING >> 2:
+	case MC_SEQ_MISC_TIMING >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING_LP >> 2;
 		break;
-        case MC_SEQ_MISC_TIMING2 >> 2:
+	case MC_SEQ_MISC_TIMING2 >> 2:
 		*out_reg = MC_SEQ_MISC_TIMING2_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D0 >> 2:
+	case MC_SEQ_RD_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_RD_CTL_D1 >> 2:
+	case MC_SEQ_RD_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_RD_CTL_D1_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D0 >> 2:
+	case MC_SEQ_WR_CTL_D0 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D0_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_D1 >> 2:
+	case MC_SEQ_WR_CTL_D1 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_D1_LP >> 2;
 		break;
-        case MC_PMG_CMD_EMRS >> 2:
+	case MC_PMG_CMD_EMRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_EMRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS >> 2:
+	case MC_PMG_CMD_MRS >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS1 >> 2:
+	case MC_PMG_CMD_MRS1 >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS1_LP >> 2;
 		break;
-        case MC_SEQ_PMG_TIMING >> 2:
+	case MC_SEQ_PMG_TIMING >> 2:
 		*out_reg = MC_SEQ_PMG_TIMING_LP >> 2;
 		break;
-        case MC_PMG_CMD_MRS2 >> 2:
+	case MC_PMG_CMD_MRS2 >> 2:
 		*out_reg = MC_SEQ_PMG_CMD_MRS2_LP >> 2;
 		break;
-        case MC_SEQ_WR_CTL_2 >> 2:
+	case MC_SEQ_WR_CTL_2 >> 2:
 		*out_reg = MC_SEQ_WR_CTL_2_LP >> 2;
 		break;
-        default:
+	default:
 		result = false;
 		break;
 	}
@@ -5604,19 +5554,19 @@ static int si_initialize_mc_reg_table(struct radeon_device *rdev)
 	WREG32(MC_SEQ_PMG_CMD_MRS2_LP, RREG32(MC_PMG_CMD_MRS2));
 	WREG32(MC_SEQ_WR_CTL_2_LP, RREG32(MC_SEQ_WR_CTL_2));
 
-        ret = radeon_atom_init_mc_reg_table(rdev, module_index, table);
-        if (ret)
-                goto init_mc_done;
+	ret = radeon_atom_init_mc_reg_table(rdev, module_index, table);
+	if (ret)
+		goto init_mc_done;
 
-        ret = si_copy_vbios_mc_reg_table(table, si_table);
-        if (ret)
-                goto init_mc_done;
+	ret = si_copy_vbios_mc_reg_table(table, si_table);
+	if (ret)
+		goto init_mc_done;
 
 	si_set_s0_mc_reg_index(si_table);
 
 	ret = si_set_mc_special_registers(rdev, si_table);
-        if (ret)
-                goto init_mc_done;
+	if (ret)
+		goto init_mc_done;
 
 	si_set_valid_flag(si_table);
 
@@ -5757,10 +5707,10 @@ static int si_upload_mc_reg_table(struct radeon_device *rdev,
 
 static void si_enable_voltage_control(struct radeon_device *rdev, bool enable)
 {
-        if (enable)
-                WREG32_P(GENERAL_PWRMGT, VOLT_PWRMGT_EN, ~VOLT_PWRMGT_EN);
-        else
-                WREG32_P(GENERAL_PWRMGT, 0, ~VOLT_PWRMGT_EN);
+	if (enable)
+		WREG32_P(GENERAL_PWRMGT, VOLT_PWRMGT_EN, ~VOLT_PWRMGT_EN);
+	else
+		WREG32_P(GENERAL_PWRMGT, 0, ~VOLT_PWRMGT_EN);
 }
 
 static enum radeon_pcie_gen si_get_maximum_link_speed(struct radeon_device *rdev,
@@ -6862,7 +6812,7 @@ static int si_parse_power_table(struct radeon_device *rdev)
 	struct _NonClockInfoArray *non_clock_info_array;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 	struct ni_ps *ps;
@@ -6882,8 +6832,9 @@ static int si_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -6948,8 +6899,9 @@ int si_dpm_init(struct radeon_device *rdev)
 	struct ni_power_info *ni_pi;
 	struct si_power_info *si_pi;
 	struct atom_clock_dividers dividers;
+	enum pci_bus_speed speed_cap = PCI_SPEED_UNKNOWN;
+	struct pci_dev *root = rdev->pdev->bus->self;
 	int ret;
-	u32 mask;
 
 	si_pi = kzalloc(sizeof(struct si_power_info), GFP_KERNEL);
 	if (si_pi == NULL)
@@ -6959,11 +6911,21 @@ int si_dpm_init(struct radeon_device *rdev)
 	eg_pi = &ni_pi->eg;
 	pi = &eg_pi->rv7xx;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret)
+	if (!pci_is_root_bus(rdev->pdev->bus))
+		speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN) {
 		si_pi->sys_pcie_mask = 0;
-	else
-		si_pi->sys_pcie_mask = mask;
+	} else {
+		if (speed_cap == PCIE_SPEED_8_0GT)
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50 |
+				RADEON_PCIE_SPEED_80;
+		else if (speed_cap == PCIE_SPEED_5_0GT)
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50;
+		else
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25;
+	}
 	si_pi->force_pcie_gen = RADEON_PCIE_GEN_INVALID;
 	si_pi->boot_pcie_gen = si_get_current_pcie_speed(rdev);
 
@@ -6991,7 +6953,9 @@ int si_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/sys/dev/pci/drm/radeon/si_smc.c b/sys/dev/pci/drm/radeon/si_smc.c
index 1e76cfb7aee..51155abda8d 100644
--- a/sys/dev/pci/drm/radeon/si_smc.c
+++ b/sys/dev/pci/drm/radeon/si_smc.c
@@ -22,7 +22,8 @@
  * Authors: Alex Deucher
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "sid.h"
 #include "ppsmc.h"
diff --git a/sys/dev/pci/drm/radeon/sid.h b/sys/dev/pci/drm/radeon/sid.h
index d1a7b58dd29..65a911ddd50 100644
--- a/sys/dev/pci/drm/radeon/sid.h
+++ b/sys/dev/pci/drm/radeon/sid.h
@@ -1145,6 +1145,7 @@
 #define	SPI_LB_CU_MASK					0x9354
 
 #define	TA_CNTL_AUX					0x9508
+#define	TA_CS_BC_BASE_ADDR				0x950C
 
 #define CC_RB_BACKEND_DISABLE				0x98F4
 #define		BACKEND_DISABLE(x)     			((x) << 16)
@@ -1559,6 +1560,7 @@
 #define UVD_UDEC_ADDR_CONFIG				0xEF4C
 #define UVD_UDEC_DB_ADDR_CONFIG				0xEF50
 #define UVD_UDEC_DBW_ADDR_CONFIG			0xEF54
+#define UVD_NO_OP					0xEFFC
 #define UVD_RBC_RB_RPTR					0xF690
 #define UVD_RBC_RB_WPTR					0xF694
 #define UVD_STATUS					0xf6bc
diff --git a/sys/dev/pci/drm/radeon/sumo_dpm.c b/sys/dev/pci/drm/radeon/sumo_dpm.c
index a2afe0963b0..1e4975f3374 100644
--- a/sys/dev/pci/drm/radeon/sumo_dpm.c
+++ b/sys/dev/pci/drm/radeon/sumo_dpm.c
@@ -21,13 +21,14 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "sumod.h"
 #include "r600_dpm.h"
 #include "cypress_dpm.h"
 #include "sumo_dpm.h"
+#include <linux/seq_file.h>
 
 #define SUMO_MAX_DEEPSLEEP_DIVIDER_ID 5
 #define SUMO_MINIMUM_ENGINE_CLOCK 800
@@ -786,8 +787,8 @@ static void sumo_program_acpi_power_level(struct radeon_device *rdev)
 	struct atom_clock_dividers dividers;
 	int ret;
 
-        ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
-                                             pi->acpi_pl.sclk,
+	ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+					     pi->acpi_pl.sclk,
 					     false, &dividers);
 	if (ret)
 		return;
@@ -1461,7 +1462,7 @@ static int sumo_parse_power_table(struct radeon_device *rdev)
 	struct _NonClockInfoArray *non_clock_info_array;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 	struct sumo_ps *ps;
@@ -1481,8 +1482,9 @@ static int sumo_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
diff --git a/sys/dev/pci/drm/radeon/sumo_smc.c b/sys/dev/pci/drm/radeon/sumo_smc.c
index dcf9447e786..cc051be4236 100644
--- a/sys/dev/pci/drm/radeon/sumo_smc.c
+++ b/sys/dev/pci/drm/radeon/sumo_smc.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "sumod.h"
 #include "sumo_dpm.h"
diff --git a/sys/dev/pci/drm/radeon/trinity_dpm.c b/sys/dev/pci/drm/radeon/trinity_dpm.c
index 93b869e7eb7..5d317f763ee 100644
--- a/sys/dev/pci/drm/radeon/trinity_dpm.c
+++ b/sys/dev/pci/drm/radeon/trinity_dpm.c
@@ -21,12 +21,13 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "trinityd.h"
 #include "r600_dpm.h"
 #include "trinity_dpm.h"
+#include <linux/seq_file.h>
 
 #define TRINITY_MAX_DEEPSLEEP_DIVIDER_ID 5
 #define TRINITY_MINIMUM_ENGINE_CLOCK 800
@@ -368,8 +369,8 @@ static void trinity_gfx_powergating_initialize(struct radeon_device *rdev)
 	int ret;
 	u32 hw_rev = (RREG32(HW_REV) & ATI_REV_ID_MASK) >> ATI_REV_ID_SHIFT;
 
-        ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
-                                             25000, false, &dividers);
+	ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+					     25000, false, &dividers);
 	if (ret)
 		return;
 
@@ -586,8 +587,8 @@ static void trinity_set_divider_value(struct radeon_device *rdev,
 	u32 value;
 	u32 ix = index * TRINITY_SIZEOF_DPM_STATE_TABLE;
 
-        ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
-                                             sclk, false, &dividers);
+	ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+					     sclk, false, &dividers);
 	if (ret)
 		return;
 
@@ -596,8 +597,8 @@ static void trinity_set_divider_value(struct radeon_device *rdev,
 	value |= CLK_DIVIDER(dividers.post_div);
 	WREG32_SMC(SMU_SCLK_DPM_STATE_0_CNTL_0 + ix, value);
 
-        ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
-                                             sclk/2, false, &dividers);
+	ret = radeon_atom_get_clock_dividers(rdev, COMPUTE_ENGINE_PLL_PARAM,
+					     sclk/2, false, &dividers);
 	if (ret)
 		return;
 
@@ -1044,14 +1045,14 @@ static int trinity_set_thermal_temperature_range(struct radeon_device *rdev,
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
 
-        if (low_temp < min_temp)
+	if (low_temp < min_temp)
 		low_temp = min_temp;
-        if (high_temp > max_temp)
+	if (high_temp > max_temp)
 		high_temp = max_temp;
-        if (high_temp < low_temp) {
+	if (high_temp < low_temp) {
 		DRM_ERROR("invalid thermal range: %d - %d\n", low_temp, high_temp);
-                return -EINVAL;
-        }
+		return -EINVAL;
+	}
 
 	WREG32_P(CG_THERMAL_INT_CTRL, DIG_THERM_INTH(49 + (high_temp / 1000)), ~DIG_THERM_INTH_MASK);
 	WREG32_P(CG_THERMAL_INT_CTRL, DIG_THERM_INTL(49 + (low_temp / 1000)), ~DIG_THERM_INTL_MASK);
@@ -1736,7 +1737,7 @@ static int trinity_parse_power_table(struct radeon_device *rdev)
 	struct _NonClockInfoArray *non_clock_info_array;
 	union power_info *power_info;
 	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-        u16 data_offset;
+	u16 data_offset;
 	u8 frev, crev;
 	u8 *power_state_offset;
 	struct sumo_ps *ps;
@@ -1756,8 +1757,9 @@ static int trinity_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
diff --git a/sys/dev/pci/drm/radeon/trinity_smc.c b/sys/dev/pci/drm/radeon/trinity_smc.c
index 86013436d48..0310e36e315 100644
--- a/sys/dev/pci/drm/radeon/trinity_smc.c
+++ b/sys/dev/pci/drm/radeon/trinity_smc.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "trinityd.h"
 #include "trinity_dpm.h"
diff --git a/sys/dev/pci/drm/radeon/uvd_v1_0.c b/sys/dev/pci/drm/radeon/uvd_v1_0.c
index 1f22d1ea8b1..0dbeb504a42 100644
--- a/sys/dev/pci/drm/radeon/uvd_v1_0.c
+++ b/sys/dev/pci/drm/radeon/uvd_v1_0.c
@@ -22,7 +22,8 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "r600d.h"
@@ -123,12 +124,13 @@ int uvd_v1_0_resume(struct radeon_device *rdev)
 	WREG32(UVD_VCPU_CACHE_SIZE0, size);
 
 	addr += size;
-	size = RADEON_UVD_STACK_SIZE >> 3;
+	size = RADEON_UVD_HEAP_SIZE >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET1, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE1, size);
 
 	addr += size;
-	size = RADEON_UVD_HEAP_SIZE >> 3;
+	size = (RADEON_UVD_STACK_SIZE +
+	       (RADEON_UVD_SESSION_SIZE * rdev->uvd.max_handles)) >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET2, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE2, size);
 
@@ -521,11 +523,17 @@ int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 		goto error;
 	}
 
-	r = radeon_fence_wait(fence, false);
-	if (r) {
+	r = radeon_fence_wait_timeout(fence, false, usecs_to_jiffies(
+		RADEON_USEC_IB_TEST_TIMEOUT));
+	if (r < 0) {
 		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
 		goto error;
+	} else if (r == 0) {
+		DRM_ERROR("radeon: fence wait timed out.\n");
+		r = -ETIMEDOUT;
+		goto error;
 	}
+	r = 0;
 	DRM_INFO("ib test on ring %d succeeded\n",  ring->idx);
 error:
 	radeon_fence_unref(&fence);
diff --git a/sys/dev/pci/drm/radeon/uvd_v2_2.c b/sys/dev/pci/drm/radeon/uvd_v2_2.c
index 8eb919a87d9..9071e656a56 100644
--- a/sys/dev/pci/drm/radeon/uvd_v2_2.c
+++ b/sys/dev/pci/drm/radeon/uvd_v2_2.c
@@ -22,7 +22,8 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv770d.h"
@@ -115,12 +116,13 @@ int uvd_v2_2_resume(struct radeon_device *rdev)
 	WREG32(UVD_VCPU_CACHE_SIZE0, size);
 
 	addr += size;
-	size = RADEON_UVD_STACK_SIZE >> 3;
+	size = RADEON_UVD_HEAP_SIZE >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET1, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE1, size);
 
 	addr += size;
-	size = RADEON_UVD_HEAP_SIZE >> 3;
+	size = (RADEON_UVD_STACK_SIZE +
+	       (RADEON_UVD_SESSION_SIZE * rdev->uvd.max_handles)) >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET2, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE2, size);
 
diff --git a/sys/dev/pci/drm/radeon/uvd_v3_1.c b/sys/dev/pci/drm/radeon/uvd_v3_1.c
index 0d55c16e02f..d722db2cf34 100644
--- a/sys/dev/pci/drm/radeon/uvd_v3_1.c
+++ b/sys/dev/pci/drm/radeon/uvd_v3_1.c
@@ -22,7 +22,7 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "nid.h"
diff --git a/sys/dev/pci/drm/radeon/uvd_v4_2.c b/sys/dev/pci/drm/radeon/uvd_v4_2.c
index 53aab066e68..91613b8a9dc 100644
--- a/sys/dev/pci/drm/radeon/uvd_v4_2.c
+++ b/sys/dev/pci/drm/radeon/uvd_v4_2.c
@@ -22,7 +22,8 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "cikd.h"
@@ -40,18 +41,25 @@ int uvd_v4_2_resume(struct radeon_device *rdev)
 	uint32_t size;
 
 	/* programm the VCPU memory controller bits 0-27 */
-	addr = rdev->uvd.gpu_addr >> 3;
+
+	/* skip over the header of the new firmware format */
+	if (rdev->uvd.fw_header_present)
+		addr = (rdev->uvd.gpu_addr + 0x200) >> 3;
+	else
+		addr = rdev->uvd.gpu_addr >> 3;
+
 	size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 4) >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET0, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE0, size);
 
 	addr += size;
-	size = RADEON_UVD_STACK_SIZE >> 3;
+	size = RADEON_UVD_HEAP_SIZE >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET1, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE1, size);
 
 	addr += size;
-	size = RADEON_UVD_HEAP_SIZE >> 3;
+	size = (RADEON_UVD_STACK_SIZE +
+	       (RADEON_UVD_SESSION_SIZE * rdev->uvd.max_handles)) >> 3;
 	WREG32(UVD_VCPU_CACHE_OFFSET2, addr);
 	WREG32(UVD_VCPU_CACHE_SIZE2, size);
 
@@ -63,5 +71,8 @@ int uvd_v4_2_resume(struct radeon_device *rdev)
 	addr = (rdev->uvd.gpu_addr >> 32) & 0xFF;
 	WREG32(UVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31));
 
+	if (rdev->uvd.fw_header_present)
+		WREG32(UVD_GP_SCRATCH4, rdev->uvd.max_handles);
+
 	return 0;
 }
diff --git a/sys/dev/pci/drm/radeon/vce_v1_0.c b/sys/dev/pci/drm/radeon/vce_v1_0.c
index 391555d8f96..f541a4b5ac5 100644
--- a/sys/dev/pci/drm/radeon/vce_v1_0.c
+++ b/sys/dev/pci/drm/radeon/vce_v1_0.c
@@ -25,7 +25,8 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "sid.h"
@@ -195,7 +196,7 @@ int vce_v1_0_load_fw(struct radeon_device *rdev, uint32_t *data)
 	memset(&data[5], 0, 44);
 	memcpy(&data[16], &sign[1], rdev->vce_fw->size - sizeof(*sign));
 
-	data += le32_to_cpu(data[4]) / 4;
+	data += (le32_to_cpu(sign->len) + 64) / 4;
 	data[0] = sign->val[i].sigval[0];
 	data[1] = sign->val[i].sigval[1];
 	data[2] = sign->val[i].sigval[2];
diff --git a/sys/dev/pci/drm/radeon/vce_v2_0.c b/sys/dev/pci/drm/radeon/vce_v2_0.c
index a3c9f1b6549..b0a43b68776 100644
--- a/sys/dev/pci/drm/radeon/vce_v2_0.c
+++ b/sys/dev/pci/drm/radeon/vce_v2_0.c
@@ -25,7 +25,8 @@
  * Authors: Christian König <christian.koenig@amd.com>
  */
 
-#include <dev/pci/drm/drmP.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "cikd.h"
@@ -52,7 +53,7 @@ static void vce_v2_0_set_sw_cg(struct radeon_device *rdev, bool gated)
 		WREG32(VCE_UENC_REG_CLOCK_GATING, tmp);
 
 		WREG32(VCE_CGTT_CLK_OVERRIDE, 0);
-    } else {
+	} else {
 		tmp = RREG32(VCE_CLOCK_GATING_B);
 		tmp |= 0xe7;
 		tmp &= ~0xe70000;
@@ -103,6 +104,10 @@ static void vce_v2_0_disable_cg(struct radeon_device *rdev)
 	WREG32(VCE_CGTT_CLK_OVERRIDE, 7);
 }
 
+/*
+ * Local variable sw_cg is used for debugging purposes, in case we
+ * ran into problems with dynamic clock gating. Don't remove it.
+ */
 void vce_v2_0_enable_mgcg(struct radeon_device *rdev, bool enable)
 {
 	bool sw_cg = false;