diff options
Diffstat (limited to 'sys/dev/pci/drm/radeon')
| -rw-r--r-- | sys/dev/pci/drm/radeon/evergreen_blit_kms.c | 730 | ||||
| -rw-r--r-- | sys/dev/pci/drm/radeon/r600_audio.c | 255 | ||||
| -rw-r--r-- | sys/dev/pci/drm/radeon/r600_blit_kms.c | 786 | ||||
| -rw-r--r-- | sys/dev/pci/drm/radeon/radeon_blit_common.h | 45 |
4 files changed, 1816 insertions, 0 deletions
diff --git a/sys/dev/pci/drm/radeon/evergreen_blit_kms.c b/sys/dev/pci/drm/radeon/evergreen_blit_kms.c index e69de29bb2d..149bb1de0f6 100644 --- a/sys/dev/pci/drm/radeon/evergreen_blit_kms.c +++ b/sys/dev/pci/drm/radeon/evergreen_blit_kms.c @@ -0,0 +1,730 @@ +/* $OpenBSD: evergreen_blit_kms.c,v 1.5 2018/04/20 21:12:50 naddy Exp $ */ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * + * 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, sublicense, + * 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 NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) 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. + * + * Authors: + * Alex Deucher <alexander.deucher@amd.com> + */ + +#include <dev/pci/drm/drmP.h> +#include <dev/pci/drm/radeon_drm.h> +#include "radeon.h" + +#include "evergreend.h" +#include "evergreen_blit_shaders.h" +#include "cayman_blit_shaders.h" +#include "radeon_blit_common.h" + +/* emits 17 */ +static void +set_render_target(struct radeon_device *rdev, int format, + int w, int h, u64 gpu_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 cb_color_info; + int pitch, slice; + + h = roundup2(h, 8); + if (h < 8) + h = 8; + + cb_color_info = CB_FORMAT(format) | + CB_SOURCE_FORMAT(CB_SF_EXPORT_NORM) | + CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + pitch = (w / 8) - 1; + slice = ((w * h) / 64) - 1; + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 15)); + radeon_ring_write(ring, (CB_COLOR0_BASE - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, pitch); + radeon_ring_write(ring, slice); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, cb_color_info); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, (w - 1) | ((h - 1) << 16)); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); +} + +/* emits 5dw */ +static void +cp_set_surface_sync(struct radeon_device *rdev, + u32 sync_type, u32 size, + u64 mc_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 cp_coher_size; + + if (size == 0xffffffff) + cp_coher_size = 0xffffffff; + else + cp_coher_size = ((size + 255) >> 8); + + if (rdev->family >= CHIP_CAYMAN) { + /* CP_COHER_CNTL2 has to be set manually when submitting a surface_sync + * to the RB directly. For IBs, the CP programs this as part of the + * surface_sync packet. + */ + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(ring, (0x85e8 - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(ring, 0); /* CP_COHER_CNTL2 */ + } + radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3)); + radeon_ring_write(ring, sync_type); + radeon_ring_write(ring, cp_coher_size); + radeon_ring_write(ring, mc_addr >> 8); + radeon_ring_write(ring, 10); /* poll interval */ +} + +/* emits 11dw + 1 surface sync = 16dw */ +static void +set_shaders(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u64 gpu_addr; + + /* VS */ + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.vs_offset; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 3)); + radeon_ring_write(ring, (SQ_PGM_START_VS - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, 2); + radeon_ring_write(ring, 0); + + /* PS */ + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.ps_offset; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 4)); + radeon_ring_write(ring, (SQ_PGM_START_PS - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, 1); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 2); + + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.vs_offset; + cp_set_surface_sync(rdev, PACKET3_SH_ACTION_ENA, 512, gpu_addr); +} + +/* emits 10 + 1 sync (5) = 15 */ +static void +set_vtx_resource(struct radeon_device *rdev, u64 gpu_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 sq_vtx_constant_word2, sq_vtx_constant_word3; + + /* high addr, stride */ + sq_vtx_constant_word2 = SQ_VTXC_BASE_ADDR_HI(upper_32_bits(gpu_addr) & 0xff) | + SQ_VTXC_STRIDE(16); +#ifdef __BIG_ENDIAN + sq_vtx_constant_word2 |= SQ_VTXC_ENDIAN_SWAP(SQ_ENDIAN_8IN32); +#endif + /* xyzw swizzles */ + sq_vtx_constant_word3 = SQ_VTCX_SEL_X(SQ_SEL_X) | + SQ_VTCX_SEL_Y(SQ_SEL_Y) | + SQ_VTCX_SEL_Z(SQ_SEL_Z) | + SQ_VTCX_SEL_W(SQ_SEL_W); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_RESOURCE, 8)); + radeon_ring_write(ring, 0x580); + radeon_ring_write(ring, gpu_addr & 0xffffffff); + radeon_ring_write(ring, 48 - 1); /* size */ + radeon_ring_write(ring, sq_vtx_constant_word2); + radeon_ring_write(ring, sq_vtx_constant_word3); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, S__SQ_CONSTANT_TYPE(SQ_TEX_VTX_VALID_BUFFER)); + + if ((rdev->family == CHIP_CEDAR) || + (rdev->family == CHIP_PALM) || + (rdev->family == CHIP_SUMO) || + (rdev->family == CHIP_SUMO2) || + (rdev->family == CHIP_CAICOS)) + cp_set_surface_sync(rdev, + PACKET3_TC_ACTION_ENA, 48, gpu_addr); + else + cp_set_surface_sync(rdev, + PACKET3_VC_ACTION_ENA, 48, gpu_addr); + +} + +/* emits 10 */ +static void +set_tex_resource(struct radeon_device *rdev, + int format, int w, int h, int pitch, + u64 gpu_addr, u32 size) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 sq_tex_resource_word0, sq_tex_resource_word1; + u32 sq_tex_resource_word4, sq_tex_resource_word7; + + if (h < 1) + h = 1; + + sq_tex_resource_word0 = TEX_DIM(SQ_TEX_DIM_2D); + sq_tex_resource_word0 |= ((((pitch >> 3) - 1) << 6) | + ((w - 1) << 18)); + sq_tex_resource_word1 = ((h - 1) << 0) | + TEX_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + /* xyzw swizzles */ + sq_tex_resource_word4 = TEX_DST_SEL_X(SQ_SEL_X) | + TEX_DST_SEL_Y(SQ_SEL_Y) | + TEX_DST_SEL_Z(SQ_SEL_Z) | + TEX_DST_SEL_W(SQ_SEL_W); + + sq_tex_resource_word7 = format | + S__SQ_CONSTANT_TYPE(SQ_TEX_VTX_VALID_TEXTURE); + + cp_set_surface_sync(rdev, + PACKET3_TC_ACTION_ENA, size, gpu_addr); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_RESOURCE, 8)); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, sq_tex_resource_word0); + radeon_ring_write(ring, sq_tex_resource_word1); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, sq_tex_resource_word4); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, sq_tex_resource_word7); +} + +/* emits 12 */ +static void +set_scissors(struct radeon_device *rdev, int x1, int y1, + int x2, int y2) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + /* workaround some hw bugs */ + if (x2 == 0) + x1 = 1; + if (y2 == 0) + y1 = 1; + if (rdev->family >= CHIP_CAYMAN) { + if ((x2 == 1) && (y2 == 1)) + x2 = 2; + } + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_SCREEN_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_GENERIC_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16) | (1 << 31)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_WINDOW_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_START) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16) | (1 << 31)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); +} + +/* emits 10 */ +static void +draw_auto(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(ring, (VGT_PRIMITIVE_TYPE - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(ring, DI_PT_RECTLIST); + + radeon_ring_write(ring, PACKET3(PACKET3_INDEX_TYPE, 0)); + radeon_ring_write(ring, +#ifdef __BIG_ENDIAN + (2 << 2) | +#endif + DI_INDEX_SIZE_16_BIT); + + radeon_ring_write(ring, PACKET3(PACKET3_NUM_INSTANCES, 0)); + radeon_ring_write(ring, 1); + + radeon_ring_write(ring, PACKET3(PACKET3_DRAW_INDEX_AUTO, 1)); + radeon_ring_write(ring, 3); + radeon_ring_write(ring, DI_SRC_SEL_AUTO_INDEX); + +} + +/* emits 39 */ +static void +set_default_state(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 sq_config, sq_gpr_resource_mgmt_1, sq_gpr_resource_mgmt_2, sq_gpr_resource_mgmt_3; + u32 sq_thread_resource_mgmt, sq_thread_resource_mgmt_2; + u32 sq_stack_resource_mgmt_1, sq_stack_resource_mgmt_2, sq_stack_resource_mgmt_3; + int num_ps_gprs, num_vs_gprs, num_temp_gprs; + int num_gs_gprs, num_es_gprs, num_hs_gprs, num_ls_gprs; + int num_ps_threads, num_vs_threads, num_gs_threads, num_es_threads; + int num_hs_threads, num_ls_threads; + int num_ps_stack_entries, num_vs_stack_entries, num_gs_stack_entries, num_es_stack_entries; + int num_hs_stack_entries, num_ls_stack_entries; + u64 gpu_addr; + int dwords; + + /* set clear context state */ + radeon_ring_write(ring, PACKET3(PACKET3_CLEAR_STATE, 0)); + radeon_ring_write(ring, 0); + + if (rdev->family < CHIP_CAYMAN) { + switch (rdev->family) { + case CHIP_CEDAR: + default: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 16; + num_gs_threads = 16; + num_es_threads = 16; + num_hs_threads = 16; + num_ls_threads = 16; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_REDWOOD: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_JUNIPER: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_PALM: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 16; + num_gs_threads = 16; + num_es_threads = 16; + num_hs_threads = 16; + num_ls_threads = 16; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_SUMO: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 25; + num_gs_threads = 25; + num_es_threads = 25; + num_hs_threads = 25; + num_ls_threads = 25; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_SUMO2: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 96; + num_vs_threads = 25; + num_gs_threads = 25; + num_es_threads = 25; + num_hs_threads = 25; + num_ls_threads = 25; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_BARTS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 85; + num_vs_stack_entries = 85; + num_gs_stack_entries = 85; + num_es_stack_entries = 85; + num_hs_stack_entries = 85; + num_ls_stack_entries = 85; + break; + case CHIP_TURKS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 20; + num_gs_threads = 20; + num_es_threads = 20; + num_hs_threads = 20; + num_ls_threads = 20; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + case CHIP_CAICOS: + num_ps_gprs = 93; + num_vs_gprs = 46; + num_temp_gprs = 4; + num_gs_gprs = 31; + num_es_gprs = 31; + num_hs_gprs = 23; + num_ls_gprs = 23; + num_ps_threads = 128; + num_vs_threads = 10; + num_gs_threads = 10; + num_es_threads = 10; + num_hs_threads = 10; + num_ls_threads = 10; + num_ps_stack_entries = 42; + num_vs_stack_entries = 42; + num_gs_stack_entries = 42; + num_es_stack_entries = 42; + num_hs_stack_entries = 42; + num_ls_stack_entries = 42; + break; + } + + if ((rdev->family == CHIP_CEDAR) || + (rdev->family == CHIP_PALM) || + (rdev->family == CHIP_SUMO) || + (rdev->family == CHIP_SUMO2) || + (rdev->family == CHIP_CAICOS)) + sq_config = 0; + else + sq_config = VC_ENABLE; + + sq_config |= (EXPORT_SRC_C | + CS_PRIO(0) | + LS_PRIO(0) | + HS_PRIO(0) | + PS_PRIO(0) | + VS_PRIO(1) | + GS_PRIO(2) | + ES_PRIO(3)); + + sq_gpr_resource_mgmt_1 = (NUM_PS_GPRS(num_ps_gprs) | + NUM_VS_GPRS(num_vs_gprs) | + NUM_CLAUSE_TEMP_GPRS(num_temp_gprs)); + sq_gpr_resource_mgmt_2 = (NUM_GS_GPRS(num_gs_gprs) | + NUM_ES_GPRS(num_es_gprs)); + sq_gpr_resource_mgmt_3 = (NUM_HS_GPRS(num_hs_gprs) | + NUM_LS_GPRS(num_ls_gprs)); + sq_thread_resource_mgmt = (NUM_PS_THREADS(num_ps_threads) | + NUM_VS_THREADS(num_vs_threads) | + NUM_GS_THREADS(num_gs_threads) | + NUM_ES_THREADS(num_es_threads)); + sq_thread_resource_mgmt_2 = (NUM_HS_THREADS(num_hs_threads) | + NUM_LS_THREADS(num_ls_threads)); + sq_stack_resource_mgmt_1 = (NUM_PS_STACK_ENTRIES(num_ps_stack_entries) | + NUM_VS_STACK_ENTRIES(num_vs_stack_entries)); + sq_stack_resource_mgmt_2 = (NUM_GS_STACK_ENTRIES(num_gs_stack_entries) | + NUM_ES_STACK_ENTRIES(num_es_stack_entries)); + sq_stack_resource_mgmt_3 = (NUM_HS_STACK_ENTRIES(num_hs_stack_entries) | + NUM_LS_STACK_ENTRIES(num_ls_stack_entries)); + + /* disable dyn gprs */ + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(ring, (SQ_DYN_GPR_CNTL_PS_FLUSH_REQ - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(ring, 0); + + /* setup LDS */ + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(ring, (SQ_LDS_RESOURCE_MGMT - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(ring, 0x10001000); + + /* SQ config */ + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 11)); + radeon_ring_write(ring, (SQ_CONFIG - PACKET3_SET_CONFIG_REG_START) >> 2); + radeon_ring_write(ring, sq_config); + radeon_ring_write(ring, sq_gpr_resource_mgmt_1); + radeon_ring_write(ring, sq_gpr_resource_mgmt_2); + radeon_ring_write(ring, sq_gpr_resource_mgmt_3); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, sq_thread_resource_mgmt); + radeon_ring_write(ring, sq_thread_resource_mgmt_2); + radeon_ring_write(ring, sq_stack_resource_mgmt_1); + radeon_ring_write(ring, sq_stack_resource_mgmt_2); + radeon_ring_write(ring, sq_stack_resource_mgmt_3); + } + + /* CONTEXT_CONTROL */ + radeon_ring_write(ring, 0xc0012800); + radeon_ring_write(ring, 0x80000000); + radeon_ring_write(ring, 0x80000000); + + /* SQ_VTX_BASE_VTX_LOC */ + radeon_ring_write(ring, 0xc0026f00); + radeon_ring_write(ring, 0x00000000); + radeon_ring_write(ring, 0x00000000); + radeon_ring_write(ring, 0x00000000); + + /* SET_SAMPLER */ + radeon_ring_write(ring, 0xc0036e00); + radeon_ring_write(ring, 0x00000000); + radeon_ring_write(ring, 0x00000012); + radeon_ring_write(ring, 0x00000000); + radeon_ring_write(ring, 0x00000000); + + /* set to DX10/11 mode */ + radeon_ring_write(ring, PACKET3(PACKET3_MODE_CONTROL, 0)); + radeon_ring_write(ring, 1); + + /* emit an IB pointing at default state */ + dwords = roundup2(rdev->r600_blit.state_len, 0x10); + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.state_offset; + radeon_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); + radeon_ring_write(ring, gpu_addr & 0xFFFFFFFC); + radeon_ring_write(ring, upper_32_bits(gpu_addr) & 0xFF); + radeon_ring_write(ring, dwords); + +} + +int evergreen_blit_init(struct radeon_device *rdev) +{ + u32 obj_size; + int i, r, dwords; + void *ptr; + u32 packet2s[16]; + int num_packet2s = 0; + + rdev->r600_blit.primitives.set_render_target = set_render_target; + rdev->r600_blit.primitives.cp_set_surface_sync = cp_set_surface_sync; + rdev->r600_blit.primitives.set_shaders = set_shaders; + rdev->r600_blit.primitives.set_vtx_resource = set_vtx_resource; + rdev->r600_blit.primitives.set_tex_resource = set_tex_resource; + rdev->r600_blit.primitives.set_scissors = set_scissors; + rdev->r600_blit.primitives.draw_auto = draw_auto; + rdev->r600_blit.primitives.set_default_state = set_default_state; + + rdev->r600_blit.ring_size_common = 8; /* sync semaphore */ + rdev->r600_blit.ring_size_common += 55; /* shaders + def state */ + rdev->r600_blit.ring_size_common += 16; /* fence emit for VB IB */ + rdev->r600_blit.ring_size_common += 5; /* done copy */ + rdev->r600_blit.ring_size_common += 16; /* fence emit for done copy */ + + rdev->r600_blit.ring_size_per_loop = 74; + if (rdev->family >= CHIP_CAYMAN) + rdev->r600_blit.ring_size_per_loop += 9; /* additional DWs for surface sync */ + + rdev->r600_blit.max_dim = 16384; + + rdev->r600_blit.state_offset = 0; + + if (rdev->family < CHIP_CAYMAN) + rdev->r600_blit.state_len = evergreen_default_size; + else + rdev->r600_blit.state_len = cayman_default_size; + + dwords = rdev->r600_blit.state_len; + while (dwords & 0xf) { + packet2s[num_packet2s++] = cpu_to_le32(PACKET2(0)); + dwords++; + } + + obj_size = dwords * 4; + obj_size = roundup2(obj_size, 256); + + rdev->r600_blit.vs_offset = obj_size; + if (rdev->family < CHIP_CAYMAN) + obj_size += evergreen_vs_size * 4; + else + obj_size += cayman_vs_size * 4; + obj_size = roundup2(obj_size, 256); + + rdev->r600_blit.ps_offset = obj_size; + if (rdev->family < CHIP_CAYMAN) + obj_size += evergreen_ps_size * 4; + else + obj_size += cayman_ps_size * 4; + obj_size = roundup2(obj_size, 256); + + /* pin copy shader into vram if not already initialized */ + if (!rdev->r600_blit.shader_obj) { + r = radeon_bo_create(rdev, obj_size, PAGE_SIZE, true, + RADEON_GEM_DOMAIN_VRAM, + NULL, &rdev->r600_blit.shader_obj); + if (r) { + DRM_ERROR("evergreen failed to allocate shader\n"); + return r; + } + + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (unlikely(r != 0)) + return r; + r = radeon_bo_pin(rdev->r600_blit.shader_obj, RADEON_GEM_DOMAIN_VRAM, + &rdev->r600_blit.shader_gpu_addr); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + if (r) { + dev_err(rdev->dev, "(%d) pin blit object failed\n", r); + return r; + } + } + + DRM_DEBUG("evergreen blit allocated bo %08x vs %08x ps %08x\n", + obj_size, + rdev->r600_blit.vs_offset, rdev->r600_blit.ps_offset); + + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (unlikely(r != 0)) + return r; + r = radeon_bo_kmap(rdev->r600_blit.shader_obj, &ptr); + if (r) { + DRM_ERROR("failed to map blit object %d\n", r); + return r; + } + + if (rdev->family < CHIP_CAYMAN) { + memcpy_toio(ptr + rdev->r600_blit.state_offset, + evergreen_default_state, rdev->r600_blit.state_len * 4); + + if (num_packet2s) + memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), + packet2s, num_packet2s * 4); + for (i = 0; i < evergreen_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(evergreen_vs[i]); + for (i = 0; i < evergreen_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(evergreen_ps[i]); + } else { + memcpy_toio(ptr + rdev->r600_blit.state_offset, + cayman_default_state, rdev->r600_blit.state_len * 4); + + if (num_packet2s) + memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), + packet2s, num_packet2s * 4); + for (i = 0; i < cayman_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(cayman_vs[i]); + for (i = 0; i < cayman_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(cayman_ps[i]); + } + radeon_bo_kunmap(rdev->r600_blit.shader_obj); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + + radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size); + return 0; +} diff --git a/sys/dev/pci/drm/radeon/r600_audio.c b/sys/dev/pci/drm/radeon/r600_audio.c index e69de29bb2d..b7aecc05ea3 100644 --- a/sys/dev/pci/drm/radeon/r600_audio.c +++ b/sys/dev/pci/drm/radeon/r600_audio.c @@ -0,0 +1,255 @@ +/* $OpenBSD: r600_audio.c,v 1.5 2018/04/20 21:12:50 naddy Exp $ */ +/* + * Copyright 2008 Advanced Micro Devices, Inc. + * Copyright 2008 Red Hat Inc. + * Copyright 2009 Christian König. + * + * 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, sublicense, + * 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 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 NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) 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. + * + * Authors: Christian König + */ +#include <dev/pci/drm/drmP.h> +#include "radeon.h" +#include "radeon_reg.h" +#include "radeon_asic.h" +#include "atom.h" + +/* + * check if enc_priv stores radeon_encoder_atom_dig + */ +static bool radeon_dig_encoder(struct drm_encoder *encoder) +{ + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + switch (radeon_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_LVDS: + case ENCODER_OBJECT_ID_INTERNAL_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_LVTM1: + case ENCODER_OBJECT_ID_INTERNAL_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_DDI: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + return true; + } + return false; +} + +/* + * check if the chipset is supported + */ +static int r600_audio_chipset_supported(struct radeon_device *rdev) +{ + return (rdev->family >= CHIP_R600 && !ASIC_IS_DCE6(rdev)) + || rdev->family == CHIP_RS600 + || rdev->family == CHIP_RS690 + || rdev->family == CHIP_RS740; +} + +struct r600_audio r600_audio_status(struct radeon_device *rdev) +{ + struct r600_audio status; + uint32_t value; + + value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL); + + /* number of channels */ + status.channels = (value & 0x7) + 1; + + /* bits per sample */ + switch ((value & 0xF0) >> 4) { + case 0x0: + status.bits_per_sample = 8; + break; + case 0x1: + status.bits_per_sample = 16; + break; + case 0x2: + status.bits_per_sample = 20; + break; + case 0x3: + status.bits_per_sample = 24; + break; + case 0x4: + status.bits_per_sample = 32; + break; + default: + dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n", + (int)value); + status.bits_per_sample = 16; + } + + /* current sampling rate in HZ */ + if (value & 0x4000) + status.rate = 44100; + else + status.rate = 48000; + status.rate *= ((value >> 11) & 0x7) + 1; + status.rate /= ((value >> 8) & 0x7) + 1; + + value = RREG32(R600_AUDIO_STATUS_BITS); + + /* iec 60958 status bits */ + status.status_bits = value & 0xff; + + /* iec 60958 category code */ + status.category_code = (value >> 8) & 0xff; + + return status; +} + +/* + * update all hdmi interfaces with current audio parameters + */ +void r600_audio_update_hdmi(void *arg1) +{ + struct radeon_device *rdev = arg1; + struct drm_device *dev = rdev->ddev; + struct r600_audio audio_status = r600_audio_status(rdev); + struct drm_encoder *encoder; + bool changed = false; + + if (rdev->audio_status.channels != audio_status.channels || + rdev->audio_status.rate != audio_status.rate || + rdev->audio_status.bits_per_sample != audio_status.bits_per_sample || + rdev->audio_status.status_bits != audio_status.status_bits || + rdev->audio_status.category_code != audio_status.category_code) { + rdev->audio_status = audio_status; + changed = true; + } + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (!radeon_dig_encoder(encoder)) + continue; + if (changed || r600_hdmi_buffer_status_changed(encoder)) + r600_hdmi_update_audio_settings(encoder); + } +} + +/* + * turn on/off audio engine + */ +static void r600_audio_engine_enable(struct radeon_device *rdev, bool enable) +{ + u32 value = 0; + DRM_INFO("%s audio support\n", enable ? "Enabling" : "Disabling"); + if (ASIC_IS_DCE4(rdev)) { + if (enable) { + value |= 0x81000000; /* Required to enable audio */ + value |= 0x0e1000f0; /* fglrx sets that too */ + } + WREG32(EVERGREEN_AUDIO_ENABLE, value); + } else { + WREG32_P(R600_AUDIO_ENABLE, + enable ? 0x81000000 : 0x0, ~0x81000000); + } + rdev->audio_enabled = enable; +} + +/* + * initialize the audio vars + */ +int r600_audio_init(struct radeon_device *rdev) +{ + if (!radeon_audio || !r600_audio_chipset_supported(rdev)) + return 0; + + r600_audio_engine_enable(rdev, true); + + rdev->audio_status.channels = -1; + rdev->audio_status.rate = -1; + rdev->audio_status.bits_per_sample = -1; + rdev->audio_status.status_bits = 0; + rdev->audio_status.category_code = 0; + + return 0; +} + +/* + * atach the audio codec to the clock source of the encoder + */ +void r600_audio_set_clock(struct drm_encoder *encoder, int clock) +{ + struct drm_device *dev = encoder->dev; + struct radeon_device *rdev = dev->dev_private; + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; + struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc); + int base_rate = 48000; + + switch (radeon_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_LVTM1: + WREG32_P(R600_AUDIO_TIMING, 0, ~0x301); + break; + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: + WREG32_P(R600_AUDIO_TIMING, 0x100, ~0x301); + break; + default: + dev_err(rdev->dev, "Unsupported encoder type 0x%02X\n", + radeon_encoder->encoder_id); + return; + } + + if (ASIC_IS_DCE4(rdev)) { + /* TODO: other PLLs? */ + WREG32(EVERGREEN_AUDIO_PLL1_MUL, base_rate * 10); + WREG32(EVERGREEN_AUDIO_PLL1_DIV, clock * 10); + WREG32(EVERGREEN_AUDIO_PLL1_UNK, 0x00000071); + + /* Select DTO source */ + WREG32(0x5ac, radeon_crtc->crtc_id); + } else { + switch (dig->dig_encoder) { + case 0: + WREG32(R600_AUDIO_PLL1_MUL, base_rate * 50); + WREG32(R600_AUDIO_PLL1_DIV, clock * 100); + WREG32(R600_AUDIO_CLK_SRCSEL, 0); + break; + + case 1: + WREG32(R600_AUDIO_PLL2_MUL, base_rate * 50); + WREG32(R600_AUDIO_PLL2_DIV, clock * 100); + WREG32(R600_AUDIO_CLK_SRCSEL, 1); + break; + default: + dev_err(rdev->dev, + "Unsupported DIG on encoder 0x%02X\n", + radeon_encoder->encoder_id); + return; + } + } +} + +/* + * release the audio timer + * TODO: How to do this correctly on SMP systems? + */ +void r600_audio_fini(struct radeon_device *rdev) +{ + if (!rdev->audio_enabled) + return; + + r600_audio_engine_enable(rdev, false); +} diff --git a/sys/dev/pci/drm/radeon/r600_blit_kms.c b/sys/dev/pci/drm/radeon/r600_blit_kms.c index e69de29bb2d..47157d09a4d 100644 --- a/sys/dev/pci/drm/radeon/r600_blit_kms.c +++ b/sys/dev/pci/drm/radeon/r600_blit_kms.c @@ -0,0 +1,786 @@ +/* $OpenBSD: r600_blit_kms.c,v 1.5 2018/04/20 21:12:50 naddy Exp $ */ +/* + * Copyright 2009 Advanced Micro Devices, Inc. + * Copyright 2009 Red Hat Inc. + * + * 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, sublicense, + * 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 NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) 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. + * + */ + +#include <dev/pci/drm/drmP.h> +#include <dev/pci/drm/radeon_drm.h> +#include "radeon.h" + +#include "r600d.h" +#include "r600_blit_shaders.h" +#include "radeon_blit_common.h" + +/* emits 21 on rv770+, 23 on r600 */ +static void +set_render_target(struct radeon_device *rdev, int format, + int w, int h, u64 gpu_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 cb_color_info; + int pitch, slice; + + h = roundup2(h, 8); + if (h < 8) + h = 8; + + cb_color_info = CB_FORMAT(format) | + CB_SOURCE_FORMAT(CB_SF_EXPORT_NORM) | + CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + pitch = (w / 8) - 1; + slice = ((w * h) / 64) - 1; + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_BASE - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + + if (rdev->family > CHIP_R600 && rdev->family < CHIP_RV770) { + radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_BASE_UPDATE, 0)); + radeon_ring_write(ring, 2 << 0); + } + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_SIZE - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, (pitch << 0) | (slice << 10)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_VIEW - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_INFO - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, cb_color_info); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_TILE - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_FRAG - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (CB_COLOR0_MASK - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); +} + +/* emits 5dw */ +static void +cp_set_surface_sync(struct radeon_device *rdev, + u32 sync_type, u32 size, + u64 mc_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 cp_coher_size; + + if (size == 0xffffffff) + cp_coher_size = 0xffffffff; + else + cp_coher_size = ((size + 255) >> 8); + + radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3)); + radeon_ring_write(ring, sync_type); + radeon_ring_write(ring, cp_coher_size); + radeon_ring_write(ring, mc_addr >> 8); + radeon_ring_write(ring, 10); /* poll interval */ +} + +/* emits 21dw + 1 surface sync = 26dw */ +static void +set_shaders(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u64 gpu_addr; + u32 sq_pgm_resources; + + /* setup shader regs */ + sq_pgm_resources = (1 << 0); + + /* VS */ + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.vs_offset; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_START_VS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_RESOURCES_VS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, sq_pgm_resources); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_CF_OFFSET_VS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); + + /* PS */ + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.ps_offset; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_START_PS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, gpu_addr >> 8); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_RESOURCES_PS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, sq_pgm_resources | (1 << 28)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_EXPORTS_PS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 2); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + radeon_ring_write(ring, (SQ_PGM_CF_OFFSET_PS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, 0); + + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.vs_offset; + cp_set_surface_sync(rdev, PACKET3_SH_ACTION_ENA, 512, gpu_addr); +} + +/* emits 9 + 1 sync (5) = 14*/ +static void +set_vtx_resource(struct radeon_device *rdev, u64 gpu_addr) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 sq_vtx_constant_word2; + + sq_vtx_constant_word2 = SQ_VTXC_BASE_ADDR_HI(upper_32_bits(gpu_addr) & 0xff) | + SQ_VTXC_STRIDE(16); +#ifdef __BIG_ENDIAN + sq_vtx_constant_word2 |= SQ_VTXC_ENDIAN_SWAP(SQ_ENDIAN_8IN32); +#endif + + radeon_ring_write(ring, PACKET3(PACKET3_SET_RESOURCE, 7)); + radeon_ring_write(ring, 0x460); + radeon_ring_write(ring, gpu_addr & 0xffffffff); + radeon_ring_write(ring, 48 - 1); + radeon_ring_write(ring, sq_vtx_constant_word2); + radeon_ring_write(ring, 1 << 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, SQ_TEX_VTX_VALID_BUFFER << 30); + + if ((rdev->family == CHIP_RV610) || + (rdev->family == CHIP_RV620) || + (rdev->family == CHIP_RS780) || + (rdev->family == CHIP_RS880) || + (rdev->family == CHIP_RV710)) + cp_set_surface_sync(rdev, + PACKET3_TC_ACTION_ENA, 48, gpu_addr); + else + cp_set_surface_sync(rdev, + PACKET3_VC_ACTION_ENA, 48, gpu_addr); +} + +/* emits 9 */ +static void +set_tex_resource(struct radeon_device *rdev, + int format, int w, int h, int pitch, + u64 gpu_addr, u32 size) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + uint32_t sq_tex_resource_word0, sq_tex_resource_word1, sq_tex_resource_word4; + + if (h < 1) + h = 1; + + sq_tex_resource_word0 = S_038000_DIM(V_038000_SQ_TEX_DIM_2D) | + S_038000_TILE_MODE(V_038000_ARRAY_1D_TILED_THIN1); + sq_tex_resource_word0 |= S_038000_PITCH((pitch >> 3) - 1) | + S_038000_TEX_WIDTH(w - 1); + + sq_tex_resource_word1 = S_038004_DATA_FORMAT(format); + sq_tex_resource_word1 |= S_038004_TEX_HEIGHT(h - 1); + + sq_tex_resource_word4 = S_038010_REQUEST_SIZE(1) | + S_038010_DST_SEL_X(SQ_SEL_X) | + S_038010_DST_SEL_Y(SQ_SEL_Y) | + S_038010_DST_SEL_Z(SQ_SEL_Z) | + S_038010_DST_SEL_W(SQ_SEL_W); + + cp_set_surface_sync(rdev, + PACKET3_TC_ACTION_ENA, size, gpu_addr); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_RESOURCE, 7)); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, sq_tex_resource_word0); + radeon_ring_write(ring, sq_tex_resource_word1); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, gpu_addr >> 8); + radeon_ring_write(ring, sq_tex_resource_word4); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, SQ_TEX_VTX_VALID_TEXTURE << 30); +} + +/* emits 12 */ +static void +set_scissors(struct radeon_device *rdev, int x1, int y1, + int x2, int y2) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_SCREEN_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_GENERIC_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16) | (1 << 31)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); + + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + radeon_ring_write(ring, (PA_SC_WINDOW_SCISSOR_TL - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2); + radeon_ring_write(ring, (x1 << 0) | (y1 << 16) | (1 << 31)); + radeon_ring_write(ring, (x2 << 0) | (y2 << 16)); +} + +/* emits 10 */ +static void +draw_auto(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(ring, (VGT_PRIMITIVE_TYPE - PACKET3_SET_CONFIG_REG_OFFSET) >> 2); + radeon_ring_write(ring, DI_PT_RECTLIST); + + radeon_ring_write(ring, PACKET3(PACKET3_INDEX_TYPE, 0)); + radeon_ring_write(ring, +#ifdef __BIG_ENDIAN + (2 << 2) | +#endif + DI_INDEX_SIZE_16_BIT); + + radeon_ring_write(ring, PACKET3(PACKET3_NUM_INSTANCES, 0)); + radeon_ring_write(ring, 1); + + radeon_ring_write(ring, PACKET3(PACKET3_DRAW_INDEX_AUTO, 1)); + radeon_ring_write(ring, 3); + radeon_ring_write(ring, DI_SRC_SEL_AUTO_INDEX); + +} + +/* emits 14 */ +static void +set_default_state(struct radeon_device *rdev) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + u32 sq_config, sq_gpr_resource_mgmt_1, sq_gpr_resource_mgmt_2; + u32 sq_thread_resource_mgmt, sq_stack_resource_mgmt_1, sq_stack_resource_mgmt_2; + int num_ps_gprs, num_vs_gprs, num_temp_gprs, num_gs_gprs, num_es_gprs; + int num_ps_threads, num_vs_threads, num_gs_threads, num_es_threads; + int num_ps_stack_entries, num_vs_stack_entries, num_gs_stack_entries, num_es_stack_entries; + u64 gpu_addr; + int dwords; + + switch (rdev->family) { + case CHIP_R600: + num_ps_gprs = 192; + num_vs_gprs = 56; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 136; + num_vs_threads = 48; + num_gs_threads = 4; + num_es_threads = 4; + num_ps_stack_entries = 128; + num_vs_stack_entries = 128; + num_gs_stack_entries = 0; + num_es_stack_entries = 0; + break; + case CHIP_RV630: + case CHIP_RV635: + num_ps_gprs = 84; + num_vs_gprs = 36; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 144; + num_vs_threads = 40; + num_gs_threads = 4; + num_es_threads = 4; + num_ps_stack_entries = 40; + num_vs_stack_entries = 40; + num_gs_stack_entries = 32; + num_es_stack_entries = 16; + break; + case CHIP_RV610: + case CHIP_RV620: + case CHIP_RS780: + case CHIP_RS880: + default: + num_ps_gprs = 84; + num_vs_gprs = 36; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 136; + num_vs_threads = 48; + num_gs_threads = 4; + num_es_threads = 4; + num_ps_stack_entries = 40; + num_vs_stack_entries = 40; + num_gs_stack_entries = 32; + num_es_stack_entries = 16; + break; + case CHIP_RV670: + num_ps_gprs = 144; + num_vs_gprs = 40; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 136; + num_vs_threads = 48; + num_gs_threads = 4; + num_es_threads = 4; + num_ps_stack_entries = 40; + num_vs_stack_entries = 40; + num_gs_stack_entries = 32; + num_es_stack_entries = 16; + break; + case CHIP_RV770: + num_ps_gprs = 192; + num_vs_gprs = 56; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 188; + num_vs_threads = 60; + num_gs_threads = 0; + num_es_threads = 0; + num_ps_stack_entries = 256; + num_vs_stack_entries = 256; + num_gs_stack_entries = 0; + num_es_stack_entries = 0; + break; + case CHIP_RV730: + case CHIP_RV740: + num_ps_gprs = 84; + num_vs_gprs = 36; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 188; + num_vs_threads = 60; + num_gs_threads = 0; + num_es_threads = 0; + num_ps_stack_entries = 128; + num_vs_stack_entries = 128; + num_gs_stack_entries = 0; + num_es_stack_entries = 0; + break; + case CHIP_RV710: + num_ps_gprs = 192; + num_vs_gprs = 56; + num_temp_gprs = 4; + num_gs_gprs = 0; + num_es_gprs = 0; + num_ps_threads = 144; + num_vs_threads = 48; + num_gs_threads = 0; + num_es_threads = 0; + num_ps_stack_entries = 128; + num_vs_stack_entries = 128; + num_gs_stack_entries = 0; + num_es_stack_entries = 0; + break; + } + + if ((rdev->family == CHIP_RV610) || + (rdev->family == CHIP_RV620) || + (rdev->family == CHIP_RS780) || + (rdev->family == CHIP_RS880) || + (rdev->family == CHIP_RV710)) + sq_config = 0; + else + sq_config = VC_ENABLE; + + sq_config |= (DX9_CONSTS | + ALU_INST_PREFER_VECTOR | + PS_PRIO(0) | + VS_PRIO(1) | + GS_PRIO(2) | + ES_PRIO(3)); + + sq_gpr_resource_mgmt_1 = (NUM_PS_GPRS(num_ps_gprs) | + NUM_VS_GPRS(num_vs_gprs) | + NUM_CLAUSE_TEMP_GPRS(num_temp_gprs)); + sq_gpr_resource_mgmt_2 = (NUM_GS_GPRS(num_gs_gprs) | + NUM_ES_GPRS(num_es_gprs)); + sq_thread_resource_mgmt = (NUM_PS_THREADS(num_ps_threads) | + NUM_VS_THREADS(num_vs_threads) | + NUM_GS_THREADS(num_gs_threads) | + NUM_ES_THREADS(num_es_threads)); + sq_stack_resource_mgmt_1 = (NUM_PS_STACK_ENTRIES(num_ps_stack_entries) | + NUM_VS_STACK_ENTRIES(num_vs_stack_entries)); + sq_stack_resource_mgmt_2 = (NUM_GS_STACK_ENTRIES(num_gs_stack_entries) | + NUM_ES_STACK_ENTRIES(num_es_stack_entries)); + + /* emit an IB pointing at default state */ + dwords = roundup2(rdev->r600_blit.state_len, 0x10); + gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.state_offset; + radeon_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); + radeon_ring_write(ring, +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (gpu_addr & 0xFFFFFFFC)); + radeon_ring_write(ring, upper_32_bits(gpu_addr) & 0xFF); + radeon_ring_write(ring, dwords); + + /* SQ config */ + radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 6)); + radeon_ring_write(ring, (SQ_CONFIG - PACKET3_SET_CONFIG_REG_OFFSET) >> 2); + radeon_ring_write(ring, sq_config); + radeon_ring_write(ring, sq_gpr_resource_mgmt_1); + radeon_ring_write(ring, sq_gpr_resource_mgmt_2); + radeon_ring_write(ring, sq_thread_resource_mgmt); + radeon_ring_write(ring, sq_stack_resource_mgmt_1); + radeon_ring_write(ring, sq_stack_resource_mgmt_2); +} + +int r600_blit_init(struct radeon_device *rdev) +{ + u32 obj_size; + int i, r, dwords; + void *ptr; + u32 packet2s[16]; + int num_packet2s = 0; + + rdev->r600_blit.primitives.set_render_target = set_render_target; + rdev->r600_blit.primitives.cp_set_surface_sync = cp_set_surface_sync; + rdev->r600_blit.primitives.set_shaders = set_shaders; + rdev->r600_blit.primitives.set_vtx_resource = set_vtx_resource; + rdev->r600_blit.primitives.set_tex_resource = set_tex_resource; + rdev->r600_blit.primitives.set_scissors = set_scissors; + rdev->r600_blit.primitives.draw_auto = draw_auto; + rdev->r600_blit.primitives.set_default_state = set_default_state; + + rdev->r600_blit.ring_size_common = 8; /* sync semaphore */ + rdev->r600_blit.ring_size_common += 40; /* shaders + def state */ + rdev->r600_blit.ring_size_common += 5; /* done copy */ + rdev->r600_blit.ring_size_common += 16; /* fence emit for done copy */ + + rdev->r600_blit.ring_size_per_loop = 76; + /* set_render_target emits 2 extra dwords on rv6xx */ + if (rdev->family > CHIP_R600 && rdev->family < CHIP_RV770) + rdev->r600_blit.ring_size_per_loop += 2; + + rdev->r600_blit.max_dim = 8192; + + rdev->r600_blit.state_offset = 0; + + if (rdev->family >= CHIP_RV770) + rdev->r600_blit.state_len = r7xx_default_size; + else + rdev->r600_blit.state_len = r6xx_default_size; + + dwords = rdev->r600_blit.state_len; + while (dwords & 0xf) { + packet2s[num_packet2s++] = cpu_to_le32(PACKET2(0)); + dwords++; + } + + obj_size = dwords * 4; + obj_size = roundup2(obj_size, 256); + + rdev->r600_blit.vs_offset = obj_size; + obj_size += r6xx_vs_size * 4; + obj_size = roundup2(obj_size, 256); + + rdev->r600_blit.ps_offset = obj_size; + obj_size += r6xx_ps_size * 4; + obj_size = roundup2(obj_size, 256); + + /* pin copy shader into vram if not already initialized */ + if (rdev->r600_blit.shader_obj == NULL) { + r = radeon_bo_create(rdev, obj_size, PAGE_SIZE, true, + RADEON_GEM_DOMAIN_VRAM, + NULL, &rdev->r600_blit.shader_obj); + if (r) { + DRM_ERROR("r600 failed to allocate shader\n"); + return r; + } + + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (unlikely(r != 0)) + return r; + r = radeon_bo_pin(rdev->r600_blit.shader_obj, RADEON_GEM_DOMAIN_VRAM, + &rdev->r600_blit.shader_gpu_addr); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + if (r) { + dev_err(rdev->dev, "(%d) pin blit object failed\n", r); + return r; + } + } + + DRM_DEBUG("r6xx blit allocated bo %08x vs %08x ps %08x\n", + obj_size, + rdev->r600_blit.vs_offset, rdev->r600_blit.ps_offset); + + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (unlikely(r != 0)) + return r; + r = radeon_bo_kmap(rdev->r600_blit.shader_obj, &ptr); + if (r) { + DRM_ERROR("failed to map blit object %d\n", r); + return r; + } + if (rdev->family >= CHIP_RV770) + memcpy_toio(ptr + rdev->r600_blit.state_offset, + r7xx_default_state, rdev->r600_blit.state_len * 4); + else + memcpy_toio(ptr + rdev->r600_blit.state_offset, + r6xx_default_state, rdev->r600_blit.state_len * 4); + if (num_packet2s) + memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), + packet2s, num_packet2s * 4); + for (i = 0; i < r6xx_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(r6xx_vs[i]); + for (i = 0; i < r6xx_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(r6xx_ps[i]); + radeon_bo_kunmap(rdev->r600_blit.shader_obj); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + + radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size); + return 0; +} + +void r600_blit_fini(struct radeon_device *rdev) +{ + int r; + + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); + if (rdev->r600_blit.shader_obj == NULL) + return; + /* If we can't reserve the bo, unref should be enough to destroy + * it when it becomes idle. + */ + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (!r) { + radeon_bo_unpin(rdev->r600_blit.shader_obj); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + } + radeon_bo_unref(&rdev->r600_blit.shader_obj); +} + +static unsigned r600_blit_create_rect(unsigned num_gpu_pages, + int *width, int *height, int max_dim) +{ + unsigned max_pages; + unsigned pages = num_gpu_pages; + int w, h; + + if (num_gpu_pages == 0) { + /* not supposed to be called with no pages, but just in case */ + h = 0; + w = 0; + pages = 0; + WARN_ON(1); + } else { + int rect_order = 2; + h = RECT_UNIT_H; + while (num_gpu_pages / rect_order) { + h *= 2; + rect_order *= 4; + if (h >= max_dim) { + h = max_dim; + break; + } + } + max_pages = (max_dim * h) / (RECT_UNIT_W * RECT_UNIT_H); + if (pages > max_pages) + pages = max_pages; + w = (pages * RECT_UNIT_W * RECT_UNIT_H) / h; + w = (w / RECT_UNIT_W) * RECT_UNIT_W; + pages = (w * h) / (RECT_UNIT_W * RECT_UNIT_H); + BUG_ON(pages == 0); + } + + + DRM_DEBUG("blit_rectangle: h=%d, w=%d, pages=%d\n", h, w, pages); + + /* return width and height only of the caller wants it */ + if (height) + *height = h; + if (width) + *width = w; + + return pages; +} + + +int r600_blit_prepare_copy(struct radeon_device *rdev, unsigned num_gpu_pages, + struct radeon_fence **fence, struct radeon_sa_bo **vb, + struct radeon_semaphore **sem) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + int r; + int ring_size; + int num_loops = 0; + int dwords_per_loop = rdev->r600_blit.ring_size_per_loop; + + /* num loops */ + while (num_gpu_pages) { + num_gpu_pages -= + r600_blit_create_rect(num_gpu_pages, NULL, NULL, + rdev->r600_blit.max_dim); + num_loops++; + } + + /* 48 bytes for vertex per loop */ + r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, vb, + (num_loops*48)+256, 256, true); + if (r) { + return r; + } + + r = radeon_semaphore_create(rdev, sem); + if (r) { + radeon_sa_bo_free(rdev, vb, NULL); + return r; + } + + /* calculate number of loops correctly */ + ring_size = num_loops * dwords_per_loop; + ring_size += rdev->r600_blit.ring_size_common; + r = radeon_ring_lock(rdev, ring, ring_size); + if (r) { + radeon_sa_bo_free(rdev, vb, NULL); + radeon_semaphore_free(rdev, sem, NULL); + return r; + } + + if (radeon_fence_need_sync(*fence, RADEON_RING_TYPE_GFX_INDEX)) { + radeon_semaphore_sync_rings(rdev, *sem, (*fence)->ring, + RADEON_RING_TYPE_GFX_INDEX); + radeon_fence_note_sync(*fence, RADEON_RING_TYPE_GFX_INDEX); + } else { + radeon_semaphore_free(rdev, sem, NULL); + } + + rdev->r600_blit.primitives.set_default_state(rdev); + rdev->r600_blit.primitives.set_shaders(rdev); + return 0; +} + +void r600_blit_done_copy(struct radeon_device *rdev, struct radeon_fence **fence, + struct radeon_sa_bo *vb, struct radeon_semaphore *sem) +{ + struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + int r; + + r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX); + if (r) { + radeon_ring_unlock_undo(rdev, ring); + return; + } + + radeon_ring_unlock_commit(rdev, ring); + radeon_sa_bo_free(rdev, &vb, *fence); + radeon_semaphore_free(rdev, &sem, *fence); +} + +void r600_kms_blit_copy(struct radeon_device *rdev, + u64 src_gpu_addr, u64 dst_gpu_addr, + unsigned num_gpu_pages, + struct radeon_sa_bo *vb) +{ + u64 vb_gpu_addr; + u32 *vb_cpu_addr; + + DRM_DEBUG("emitting copy %16llx %16llx %d\n", + src_gpu_addr, dst_gpu_addr, num_gpu_pages); + vb_cpu_addr = (u32 *)radeon_sa_bo_cpu_addr(vb); + vb_gpu_addr = radeon_sa_bo_gpu_addr(vb); + + while (num_gpu_pages) { + int w, h; + unsigned size_in_bytes; + unsigned pages_per_loop = + r600_blit_create_rect(num_gpu_pages, &w, &h, + rdev->r600_blit.max_dim); + + size_in_bytes = pages_per_loop * RADEON_GPU_PAGE_SIZE; + DRM_DEBUG("rectangle w=%d h=%d\n", w, h); + + vb_cpu_addr[0] = 0; + vb_cpu_addr[1] = 0; + vb_cpu_addr[2] = 0; + vb_cpu_addr[3] = 0; + + vb_cpu_addr[4] = 0; + vb_cpu_addr[5] = int2float(h); + vb_cpu_addr[6] = 0; + vb_cpu_addr[7] = int2float(h); + + vb_cpu_addr[8] = int2float(w); + vb_cpu_addr[9] = int2float(h); + vb_cpu_addr[10] = int2float(w); + vb_cpu_addr[11] = int2float(h); + + rdev->r600_blit.primitives.set_tex_resource(rdev, FMT_8_8_8_8, + w, h, w, src_gpu_addr, size_in_bytes); + rdev->r600_blit.primitives.set_render_target(rdev, COLOR_8_8_8_8, + w, h, dst_gpu_addr); + rdev->r600_blit.primitives.set_scissors(rdev, 0, 0, w, h); + rdev->r600_blit.primitives.set_vtx_resource(rdev, vb_gpu_addr); + rdev->r600_blit.primitives.draw_auto(rdev); + rdev->r600_blit.primitives.cp_set_surface_sync(rdev, + PACKET3_CB_ACTION_ENA | PACKET3_CB0_DEST_BASE_ENA, + size_in_bytes, dst_gpu_addr); + + vb_cpu_addr += 12; + vb_gpu_addr += 4*12; + src_gpu_addr += size_in_bytes; + dst_gpu_addr += size_in_bytes; + num_gpu_pages -= pages_per_loop; + } +} + +/* 23 bits of float fractional data */ +#define I2F_FRAC_BITS 23 +#define I2F_MASK ((1 << I2F_FRAC_BITS) - 1) + +/* + * Converts unsigned integer into 32-bit IEEE floating point representation. + * Will be exact from 0 to 2^24. Above that, we round towards zero + * as the fractional bits will not fit in a float. (It would be better to + * round towards even as the fpu does, but that is slower.) + */ +__pure uint32_t int2float(uint32_t x) +{ + uint32_t msb, exponent, fraction; + + /* Zero is special */ + if (!x) return 0; + + /* Get location of the most significant bit */ + msb = fls(x); + + /* + * Use a rotate instead of a shift because that works both leftwards + * and rightwards due to the mod(32) behaviour. This means we don't + * need to check to see if we are above 2^24 or not. + */ + fraction = ror32(x, (msb - I2F_FRAC_BITS) & 0x1f) & I2F_MASK; + exponent = (127 + msb) << I2F_FRAC_BITS; + + return fraction + exponent; +} diff --git a/sys/dev/pci/drm/radeon/radeon_blit_common.h b/sys/dev/pci/drm/radeon/radeon_blit_common.h index e69de29bb2d..3a18d4b22ee 100644 --- a/sys/dev/pci/drm/radeon/radeon_blit_common.h +++ b/sys/dev/pci/drm/radeon/radeon_blit_common.h @@ -0,0 +1,45 @@ +/* $OpenBSD: radeon_blit_common.h,v 1.3 2018/04/20 21:12:50 naddy Exp $ */ +/* + * Copyright 2009 Advanced Micro Devices, Inc. + * Copyright 2009 Red Hat Inc. + * Copyright 2012 Alcatel-Lucent, Inc. + * + * 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, sublicense, + * 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 NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) 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. + * + */ + +#ifndef __RADEON_BLIT_COMMON_H__ + +#define DI_PT_RECTLIST 0x11 +#define DI_INDEX_SIZE_16_BIT 0x0 +#define DI_SRC_SEL_AUTO_INDEX 0x2 + +#define FMT_8 0x1 +#define FMT_5_6_5 0x8 +#define FMT_8_8_8_8 0x1a +#define COLOR_8 0x1 +#define COLOR_5_6_5 0x8 +#define COLOR_8_8_8_8 0x1a + +#define RECT_UNIT_H 32 +#define RECT_UNIT_W (RADEON_GPU_PAGE_SIZE / 4 / RECT_UNIT_H) + +#define __RADEON_BLIT_COMMON_H__ +#endif |