drm/amdkfd: Use same SQ prefetch setting as amdgpu

0 causes instruction fetch stall at cache line boundary under some
conditions on Navi10. A non-zero prefetch is the preferred default
in any case.

Fixes soft hang in Luxmark.

Signed-off-by: Jay Cornwall <jay.cornwall@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
This commit is contained in:
Jay Cornwall 2020-10-17 08:38:43 -05:00 committed by Alex Deucher
parent a6c42e8431
commit d56b1980d7
1 changed files with 3 additions and 2 deletions

View File

@ -58,8 +58,9 @@ static int update_qpd_v10(struct device_queue_manager *dqm,
/* check if sh_mem_config register already configured */
if (qpd->sh_mem_config == 0) {
qpd->sh_mem_config =
SH_MEM_ALIGNMENT_MODE_UNALIGNED <<
SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT;
(SH_MEM_ALIGNMENT_MODE_UNALIGNED <<
SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) |
(3 << SH_MEM_CONFIG__INITIAL_INST_PREFETCH__SHIFT);
#if 0
/* TODO:
* This shouldn't be an issue with Navi10. Verify.