Instead of from physical address. v2: add comment to make the code more readable
Signed-off-by: Tao Zhou <[email protected]> --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 15 ++++++++++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c | 4 ++-- 2 files changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 81135cd45d0e..ac00bc93f34c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -3010,8 +3010,13 @@ static int amdgpu_ras_mca2pa_by_idx(struct amdgpu_device *adev, addr_in.ma.err_addr = bps->address; addr_in.ma.socket_id = socket; addr_in.ma.ch_inst = bps->mem_channel; - /* tell RAS TA the node instance is not used */ - addr_in.ma.node_inst = TA_RAS_INV_NODE; + if (!amdgpu_ras_smu_eeprom_supported(adev)) { + /* tell RAS TA the node instance is not used */ + addr_in.ma.node_inst = TA_RAS_INV_NODE; + } else { + addr_in.ma.umc_inst = bps->mcumc_id; + addr_in.ma.node_inst = bps->cu; + } if (adev->umc.ras && adev->umc.ras->convert_ras_err_addr) ret = adev->umc.ras->convert_ras_err_addr(adev, err_data, @@ -3158,7 +3163,11 @@ static int __amdgpu_ras_convert_rec_from_rom(struct amdgpu_device *adev, save_nps = (bps->retired_page >> UMC_NPS_SHIFT) & UMC_NPS_MASK; bps->retired_page &= ~(UMC_NPS_MASK << UMC_NPS_SHIFT); } else { - save_nps = nps; + /* if pmfw manages eeprom, save_nps is not stored on eeprom, + * we should always convert mca address into physical address, + * make save_nps different from nps + */ + save_nps = nps + 1; } if (save_nps == nps) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c index 64c004a2b767..425aa44446cc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c @@ -1019,10 +1019,10 @@ int amdgpu_ras_eeprom_read_idx(struct amdgpu_ras_eeprom_control *control, record[i - rec_idx].retired_page = 0x1ULL; record[i - rec_idx].ts = ts; record[i - rec_idx].err_type = AMDGPU_RAS_EEPROM_ERR_NON_RECOVERABLE; - record[i - rec_idx].cu = 0; if (adev->umc.ras->mca_ipid_parse) - adev->umc.ras->mca_ipid_parse(adev, ipid, NULL, + adev->umc.ras->mca_ipid_parse(adev, ipid, + (uint32_t *)&(record[i - rec_idx].cu), (uint32_t *)&(record[i - rec_idx].mem_channel), (uint32_t *)&(record[i - rec_idx].mcumc_id), NULL); else -- 2.34.1
