AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[SWDEV-474474] - Changed Monitor PCIE_REPLAY count to use gpu_metrics

Browse Source 
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I4351a23e8412875bb4b23b30747ac6d0bf3d3c56
This commit is contained in:
Maisam Arif
2024-07-17 17:10:37 -05:00
parent 0fc779b14f
commit 5b7be3bf99
1 changed files with 11 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
amdsmi_cli/amdsmi_commands.py
+11 -3
View File
@@ -4300,11 +4300,19 @@ class AMDSMICommands():
self.logger.table_header += 'DOUBLE_ECC'.rjust(12) self.logger.table_header += 'DOUBLE_ECC'.rjust(12)
try: try:
pcie_replay = amdsmi_interface.amdsmi_get_gpu_pci_replay_counter(args.gpu) pcie_metric = amdsmi_interface.amdsmi_get_pcie_info(args.gpu)['pcie_metric']
monitor_values['pcie_replay'] = pcie_replay logging.debug("PCIE Metric for %s | %s", gpu_id, pcie_metric)
monitor_values['pcie_replay'] = pcie_metric['pcie_replay_count']
except amdsmi_exception.AmdSmiLibraryException as e: except amdsmi_exception.AmdSmiLibraryException as e:
monitor_values['pcie_replay'] = "N/A" monitor_values['pcie_replay'] = "N/A"
logging.debug("Failed to get pcie replay counter on gpu %s | %s", gpu_id, e.get_error_info()) logging.debug("Failed to get gpu_metrics pcie replay counter on gpu %s | %s", gpu_id, e.get_error_info())
if monitor_values['pcie_replay'] == "N/A":
try:
pcie_replay = amdsmi_interface.amdsmi_get_gpu_pci_replay_counter(args.gpu)
monitor_values['pcie_replay'] = pcie_replay
except amdsmi_exception.AmdSmiLibraryException as e:
logging.debug("Failed to get sysfs pcie replay counter on gpu %s | %s", gpu_id, e.get_error_info())
self.logger.table_header += 'PCIE_REPLAY'.rjust(13) self.logger.table_header += 'PCIE_REPLAY'.rjust(13)
if args.vram_usage: if args.vram_usage: