[PATCH Review 1/1] drm/amdgpu: Skip ring test during ras in recovery
Stanley.Yang
Stanley.Yang at amd.com
Thu Sep 28 03:42:03 UTC 2023
This is workaround due to ring test failed during ras
do gpu recovery for aqua vanjaram.
Signed-off-by: Stanley.Yang <Stanley.Yang at amd.com>
---
drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
index fbfe0a1c4b19..9fff58d073a7 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
@@ -248,10 +248,16 @@ static int gfx_v9_4_3_ring_test_ring(struct amdgpu_ring *ring)
{
uint32_t scratch_reg0_offset, xcc_offset;
struct amdgpu_device *adev = ring->adev;
+ struct amdgpu_ras *ras;
uint32_t tmp = 0;
unsigned i;
int r;
+ /* This is workaround: ring test failed during ras recovery */
+ ras = amdgpu_ras_get_context(adev);
+ if (ras && atomic_read(&ras->in_recovery))
+ return 0;
+
/* Use register offset which is local to XCC in the packet */
xcc_offset = SOC15_REG_OFFSET(GC, 0, regSCRATCH_REG0);
scratch_reg0_offset = SOC15_REG_OFFSET(GC, GET_INST(GC, ring->xcc_id), regSCRATCH_REG0);
--
2.25.1
More information about the amd-gfx
mailing list