diff options
| author | Ce Sun <cesun102@amd.com> | 2026-07-24 12:03:30 +0800 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2026-07-28 19:17:32 -0400 |
| commit | 9ba8e75d2ee850294be33afc6aeba6ecdb95a6c6 (patch) | |
| tree | 449289ae4755c6e5e87ac81e0eea53cc599c7e27 | |
| parent | cba4928cdffaa0f9012acff0ec4f896320107077 (diff) | |
| download | linux-next-9ba8e75d2ee850294be33afc6aeba6ecdb95a6c6.tar.gz linux-next-9ba8e75d2ee850294be33afc6aeba6ecdb95a6c6.zip | |
drm/amd/ras: add device lost check to early exit psp cmd wait loop
Add device lost status check in PSP fence wait loop to exit
polling early when GPU device lost
Signed-off-by: Ce Sun <cesun102@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
| -rw-r--r-- | drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c | 3 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/ras/rascore/ras.h | 2 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/ras/rascore/ras_core.c | 14 | ||||
| -rw-r--r-- | drivers/gpu/drm/amd/ras/rascore/ras_psp.c | 3 |
4 files changed, 22 insertions, 0 deletions
diff --git a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c index e4444798bc73..0b36b689fcfc 100644 --- a/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c +++ b/drivers/gpu/drm/amd/ras/ras_mgr/amdgpu_ras_sys.c @@ -171,6 +171,9 @@ static int amdgpu_ras_sys_check_gpu_status(struct ras_core_context *ras_core, if (amdgpu_in_reset(adev) || amdgpu_ras_in_recovery(adev)) gpu_status |= RAS_GPU_STATUS__IN_RESET; + if (amdgpu_device_bus_status_check(adev)) + gpu_status |= RAS_GPU_STATUS__DEVICE_LOST; + if (amdgpu_sriov_vf(adev)) gpu_status |= RAS_GPU_STATUS__IS_VF; diff --git a/drivers/gpu/drm/amd/ras/rascore/ras.h b/drivers/gpu/drm/amd/ras/rascore/ras.h index d0018f87172e..8c3d603ba801 100644 --- a/drivers/gpu/drm/amd/ras/rascore/ras.h +++ b/drivers/gpu/drm/amd/ras/rascore/ras.h @@ -137,6 +137,7 @@ enum ras_gpu_status { RAS_GPU_STATUS__IN_RESET = 0x2, RAS_GPU_STATUS__IS_RMA = 0x4, RAS_GPU_STATUS__IS_VF = 0x8, + RAS_GPU_STATUS__DEVICE_LOST = 0x10, }; enum ras_fw_eeprom_cmd { @@ -374,6 +375,7 @@ int ras_core_query_block_ecc_data(struct ras_core_context *ras_core, bool ras_core_gpu_in_reset(struct ras_core_context *ras_core); bool ras_core_gpu_is_rma(struct ras_core_context *ras_core); bool ras_core_gpu_is_vf(struct ras_core_context *ras_core); +bool ras_core_gpu_device_lost(struct ras_core_context *ras_core); bool ras_core_handle_nbio_irq(struct ras_core_context *ras_core, void *data); int ras_core_handle_fatal_error(struct ras_core_context *ras_core); diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_core.c b/drivers/gpu/drm/amd/ras/rascore/ras_core.c index f8a027b60b38..75e8216fae8a 100644 --- a/drivers/gpu/drm/amd/ras/rascore/ras_core.c +++ b/drivers/gpu/drm/amd/ras/rascore/ras_core.c @@ -147,6 +147,20 @@ bool ras_core_gpu_is_vf(struct ras_core_context *ras_core) return (status & RAS_GPU_STATUS__IS_VF) ? true : false; } +bool ras_core_gpu_device_lost(struct ras_core_context *ras_core) +{ + uint32_t status = 0; + + if (!ras_core) + return false; + + if (ras_core->sys_fn && + ras_core->sys_fn->check_gpu_status) + ras_core->sys_fn->check_gpu_status(ras_core, &status); + + return (status & RAS_GPU_STATUS__DEVICE_LOST) ? true : false; +} + bool ras_core_gpu_is_rma(struct ras_core_context *ras_core) { if (!ras_core) diff --git a/drivers/gpu/drm/amd/ras/rascore/ras_psp.c b/drivers/gpu/drm/amd/ras/rascore/ras_psp.c index 358f602b167d..35264c119bbb 100644 --- a/drivers/gpu/drm/amd/ras/rascore/ras_psp.c +++ b/drivers/gpu/drm/amd/ras/rascore/ras_psp.c @@ -294,6 +294,9 @@ static int send_psp_cmd(struct ras_core_context *ras_core, psp_ctx->in_fence_value) { if (--timeout == 0) break; + + if (ras_core_gpu_device_lost(ras_core)) + break; /* * Shouldn't wait for timeout when err_event_athub occurs, * because gpu reset thread triggered and lock resource should |