lkml.org 
[lkml]   [2019]   [Dec]   [20]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH AUTOSEL 4.19 04/34] drm/amdgpu: add cache flush workaround to gfx8 emit_fence
Date
From: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>

[ Upstream commit bf26da927a1cd57c9deb2db29ae8cf276ba8b17b ]

The same workaround is used for gfx7.
Both PAL and Mesa use it for gfx8 too, so port this commit to
gfx_v8_0_ring_emit_fence_gfx.

Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 22 +++++++++++++++++++---
1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 5a9534a82d409..e1cb7fa89e4d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -6405,7 +6405,23 @@ static void gfx_v8_0_ring_emit_fence_gfx(struct amdgpu_ring *ring, u64 addr,
bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
bool int_sel = flags & AMDGPU_FENCE_FLAG_INT;

- /* EVENT_WRITE_EOP - flush caches, send int */
+ /* Workaround for cache flush problems. First send a dummy EOP
+ * event down the pipe with seq one below.
+ */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
+ amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
+ EOP_TC_ACTION_EN |
+ EOP_TC_WB_ACTION_EN |
+ EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) |
+ EVENT_INDEX(5)));
+ amdgpu_ring_write(ring, addr & 0xfffffffc);
+ amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xffff) |
+ DATA_SEL(1) | INT_SEL(0));
+ amdgpu_ring_write(ring, lower_32_bits(seq - 1));
+ amdgpu_ring_write(ring, upper_32_bits(seq - 1));
+
+ /* Then send the real EOP event down the pipe:
+ * EVENT_WRITE_EOP - flush caches, send int */
amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
EOP_TC_ACTION_EN |
@@ -7154,7 +7170,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = {
5 + /* COND_EXEC */
7 + /* PIPELINE_SYNC */
VI_FLUSH_GPU_TLB_NUM_WREG * 5 + 9 + /* VM_FLUSH */
- 8 + /* FENCE for VM_FLUSH */
+ 12 + /* FENCE for VM_FLUSH */
20 + /* GDS switch */
4 + /* double SWITCH_BUFFER,
the first COND_EXEC jump to the place just
@@ -7166,7 +7182,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = {
31 + /* DE_META */
3 + /* CNTX_CTRL */
5 + /* HDP_INVL */
- 8 + 8 + /* FENCE x2 */
+ 12 + 12 + /* FENCE x2 */
2, /* SWITCH_BUFFER */
.emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_gfx */
.emit_ib = gfx_v8_0_ring_emit_ib_gfx,
--
2.20.1
\
 
 \ /
  Last update: 2019-12-20 15:38    [W:0.097 / U:7.036 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site