drm/amdgpu: Added RAS UMC error query support for Arcturus

Updated UMC 6.1 function set to support UMC 6.1.1 and 6.1.2 devices

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: John Clements <john.clements@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
John Clements 2019-12-11 10:18:55 +08:00 committed by Alex Deucher
parent a0250689cb
commit 4cf781c24c
3 changed files with 74 additions and 16 deletions

View File

@ -708,11 +708,18 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev)
adev->umc.funcs = &umc_v6_0_funcs;
break;
case CHIP_VEGA20:
adev->umc.max_ras_err_cnt_per_query = UMC_V6_1_TOTAL_CHANNEL_NUM;
adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM;
adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM;
adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_VG20;
adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0];
adev->umc.funcs = &umc_v6_1_funcs;
break;
case CHIP_ARCTURUS:
adev->umc.max_ras_err_cnt_per_query = UMC_V6_1_TOTAL_CHANNEL_NUM;
adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM;
adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM;
adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET;
adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_ARCT;
adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0];
adev->umc.funcs = &umc_v6_1_funcs;
break;

View File

@ -31,6 +31,14 @@
#define smnMCA_UMC0_MCUMC_ADDRT0 0x50f10
/* UMC 6_1_2 register offsets */
#define mmUMCCH0_0_EccErrCntSel_ARCT 0x0360
#define mmUMCCH0_0_EccErrCntSel_ARCT_BASE_IDX 1
#define mmUMCCH0_0_EccErrCnt_ARCT 0x0361
#define mmUMCCH0_0_EccErrCnt_ARCT_BASE_IDX 1
#define mmMCA_UMC_UMC0_MCUMC_STATUST0_ARCT 0x03c2
#define mmMCA_UMC_UMC0_MCUMC_STATUST0_ARCT_BASE_IDX 1
/*
* (addr / 256) * 8192, the higher 26 bits in ErrorAddr
* is the index of 8KB block
@ -95,12 +103,25 @@ static void umc_v6_1_query_correctable_error_count(struct amdgpu_device *adev,
uint64_t mc_umc_status;
uint32_t mc_umc_status_addr;
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt);
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
if (adev->asic_type == CHIP_ARCTURUS) {
/* UMC 6_1_2 registers */
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel_ARCT);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt_ARCT);
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0_ARCT);
} else {
/* UMC 6_1_1 registers */
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt);
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
}
/* select the lower chip and check the error count */
ecc_err_cnt_sel = RREG32(ecc_err_cnt_sel_addr + umc_reg_offset);
@ -141,8 +162,17 @@ static void umc_v6_1_querry_uncorrectable_error_count(struct amdgpu_device *adev
uint64_t mc_umc_status;
uint32_t mc_umc_status_addr;
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
if (adev->asic_type == CHIP_ARCTURUS) {
/* UMC 6_1_2 registers */
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0_ARCT);
} else {
/* UMC 6_1_1 registers */
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
}
/* check the MCUMC_STATUS */
mc_umc_status = RREG64_UMC(mc_umc_status_addr + umc_reg_offset);
@ -179,8 +209,17 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev,
uint64_t mc_umc_status, err_addr, retired_page;
struct eeprom_table_record *err_rec;
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
if (adev->asic_type == CHIP_ARCTURUS) {
/* UMC 6_1_2 registers */
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0_ARCT);
} else {
/* UMC 6_1_1 registers */
mc_umc_status_addr =
SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0);
}
/* skip error address process if -ENOMEM */
if (!err_data->err_addr) {
@ -241,10 +280,21 @@ static void umc_v6_1_err_cnt_init_per_channel(struct amdgpu_device *adev,
uint32_t ecc_err_cnt_sel, ecc_err_cnt_sel_addr;
uint32_t ecc_err_cnt_addr;
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt);
if (adev->asic_type == CHIP_ARCTURUS) {
/* UMC 6_1_2 registers */
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel_ARCT);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt_ARCT);
} else {
/* UMC 6_1_1 registers */
ecc_err_cnt_sel_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCntSel);
ecc_err_cnt_addr =
SOC15_REG_OFFSET(UMC, 0, mmUMCCH0_0_EccErrCnt);
}
/* select the lower chip and check the error count */
ecc_err_cnt_sel = RREG32(ecc_err_cnt_sel_addr + umc_reg_offset);

View File

@ -35,7 +35,8 @@
/* total channel instances in one umc block */
#define UMC_V6_1_TOTAL_CHANNEL_NUM (UMC_V6_1_CHANNEL_INSTANCE_NUM * UMC_V6_1_UMC_INSTANCE_NUM)
/* UMC regiser per channel offset */
#define UMC_V6_1_PER_CHANNEL_OFFSET 0x800
#define UMC_V6_1_PER_CHANNEL_OFFSET_VG20 0x800
#define UMC_V6_1_PER_CHANNEL_OFFSET_ARCT 0x400
/* EccErrCnt max value */
#define UMC_V6_1_CE_CNT_MAX 0xffff