radeon: Make cache flushing for R300 series less invasive.

Instead of calling the DRM CP idle ioctl, just emit the cache flush commands
into the CP stream.

1 files changed, 3 insertions, 43 deletions

diff --git a/src/radeon_dri.c b/src/radeon_dri.c
index 70ff288b..e2238252 100644
--- a/src/radeon_dri.c
+++ b/src/radeon_dri.c
@@ -357,50 +357,10 @@ static void RADEONEnterServer(ScreenPtr pScreen)
     RADEON_MARK_SYNC(info, pScrn);
 
     pSAREAPriv = DRIGetSAREAPrivate(pScrn->pScreen);
-    if (pSAREAPriv->ctxOwner != DRIGetContext(pScrn->pScreen))
+    if (pSAREAPriv->ctxOwner != DRIGetContext(pScrn->pScreen)) {
 	info->XInited3D = FALSE;
-
-
-    /* TODO: Fix this more elegantly.
-     * Sometimes (especially with multiple DRI clients), this code
-     * runs immediately after a DRI client issues a rendering command.
-     *
-     * The accel code regularly inserts WAIT_UNTIL_IDLE into the
-     * command buffer that is sent with the indirect buffer below.
-     * The accel code fails to set the 3D cache flush registers for
-     * the R300 before sending WAIT_UNTIL_IDLE. Sending a cache flush
-     * on these new registers is not necessary for pure 2D functionality,
-     * but it *is* necessary after 3D operations.
-     * Without the cache flushes before WAIT_UNTIL_IDLE, the R300 locks up.
-     *
-     * The CP_IDLE call into the DRM indirectly flushes all caches and
-     * thus avoids the lockup problem, but the solution is far from ideal.
-     * Better solutions could be:
-     *  - always flush caches when entering the X server
-     *  - track the type of rendering commands somewhere and issue
-     *    cache flushes when they change
-     * However, I don't feel confident enough with the control flow
-     * inside the X server to implement either fix. -- nh
-     */
-    
-    /* On my computer (Radeon Mobility M10)
-       The fix below results in x11perf -shmput500 rate of 245.0/sec
-       which is lower than 264.0/sec I get without it.
-       
-       Doing the same each time before indirect buffer is submitted
-       results in x11perf -shmput500 rate of 225.0/sec.
-       
-       On the other hand, not using CP acceleration at all benchmarks
-       at 144.0/sec.
-      
-       For now let us accept this as a lesser evil, especially as the
-       DRM driver for R300 is still in flux.
-       
-       Once the code is more stable this should probably be moved into DRM driver.
-    */ 
-     
-    if (info->ChipFamily>=CHIP_FAMILY_R300)
-        drmCommandNone(info->drmFD, DRM_RADEON_CP_IDLE);
+	info->needCacheFlush = (info->ChipFamily >= CHIP_FAMILY_R300);
+    }
 
 #ifdef DAMAGE
     if (!info->pDamage && info->allowPageFlip) {