[SWDEV-230863] Improve the functionality of RdcSmiHealth module.

Memory check:get the threshold of retired page number
EEPROM check:read and verify the checksum
Power/Thermal check: power/thermal throttle status counter

Signed-off-by: Meng Li <li.meng@amd.com>
Change-Id: Id2c751416eb5bf007e6e1da8dc05966a6ba1324e
This commit is contained in:
limeng12
2025-01-09 13:52:56 +08:00
committed by Meng, Li (Jassmine)
parent 83f36f1673
commit 016a1d9d39
11 changed files with 307 additions and 72 deletions
+5 -4
View File
@@ -169,8 +169,9 @@ FLD_DESC_ENT(RDC_EVNT_NOTIF_RING_HANG, "GPU ring hang just occured",
// RDC health related fields
FLD_DESC_ENT(RDC_HEALTH_XGMI_ERROR, "XGMI one or more errors detected", "XGMI_ERROR", true)
FLD_DESC_ENT(RDC_HEALTH_PCIE_REPLAY_COUNT, "Total PCIE replay count", "PCIE_REPLAY_COUNT", true)
FLD_DESC_ENT(RDC_HEALTH_RETIRED_PAGE_NUM, "Retired page number", "RETIRED_PAGE_NUM", true)
FLD_DESC_ENT(RDC_HEALTH_PENDING_PAGE_NUM, "Pending page number", "PENDING_PAGE_NUM", true)
FLD_DESC_ENT(RDC_HEALTH_RETIRED_PAGE_LIMIT, "Retired page limit", "RETIRED_PAGE_LIMIT", false)
FLD_DESC_ENT(RDC_HEALTH_UNCORRECTABLE_PAGE_LIMIT, "Uncorrectable page limit", "UNCORRECTABLE_PAGE_LIMIT", false)
FLD_DESC_ENT(RDC_HEALTH_POWER_THROTTLE_TIME, "Power throttle status counter", "POWER_THROTTLE_TIME", false)
FLD_DESC_ENT(RDC_HEALTH_THERMAL_THROTTLE_TIME, "Total time(ms) in thermal throttle status", "THERMAL_THROTTLE_TIME", false)
FLD_DESC_ENT(RDC_HEALTH_RETIRED_PAGE_LIMIT, "Retired page limit", "RETIRED_PAGE_LIMIT", true)
FLD_DESC_ENT(RDC_HEALTH_EEPROM_CONFIG_VALID, "Verify checksum of EEPROM", "EEPROM_CONFIG_VALID", true)
FLD_DESC_ENT(RDC_HEALTH_POWER_THROTTLE_TIME, "Power throttle status counter", "POWER_THROTTLE_TIME", true)
FLD_DESC_ENT(RDC_HEALTH_THERMAL_THROTTLE_TIME, "Total time(ms) in thermal throttle status", "THERMAL_THROTTLE_TIME", true)