From: Tao Zhou <tao.zhou1@amd.com>
Date: Wed, 27 Aug 2025 07:48:06 +0000 (+0800)
Subject: drm/amdgpu: get RAS bad page address from MCA address
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=e84835940e60a7d5263767ee92acc08f9877cb26;p=thirdparty%2Fkernel%2Flinux.git

drm/amdgpu: get RAS bad page address from MCA address

Instead of from physical address.

v2: add comment to make the code more readable

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
---

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 36a5393d6b74e..9e2e098af86c2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -3014,8 +3014,13 @@ 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,
@@ -3162,7 +3167,11 @@ 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 670c0dedf4e92..ec248ca6ef930 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -1022,9 +1022,9 @@ 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;
 
-		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);
 	}