search

ROCm / ROC-smi

ROC System Management Interface
https://github.com/RadeonOpenCompute/ROC-smi/blob/master/README.md
179 stars 55 forks source link

rsmi_pci_dev_bandwidth_get returns RSMI_STATUS_NOT_YET_IMPLEMENTED #99

Open gcongiu opened 2 years ago

gcongiu commented 2 years ago

Since the latest updates of rocm toolkit I get a RSMI_STATUS_NOT_YET_IMPLEMENTED error when calling rsmi_pci_dev_bandwidth_get. The error appears with all versions of rocm which might indicate the problem is with the latest driver version.

gcongiu commented 2 years ago

OLCF responded to this with the following:

Our engineering team has indicated that the "RSMI_STATUS_NOT_YET_IMPLEMENTED" error message is expected output since the system is a Bard Peak system that doesn't have PCIe (uses xGMI) connection.

The error is misleading. If xGMI overrides PCIe connection for data transfer then the error should say PCIe disabled (or an equivalent error).