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

Add support for deferred RAS errors in API

Browse Source 
The API will support the deferred errors

Change-Id: I221a146f09fefde1fc31e5f746d0870e07c93561
This commit is contained in:
Bill(Shuzhou) Liu
2024-03-04 09:27:28 -06:00
committed by Maisam Arif
parent c8c03dfab0
commit c489cb8f3f
4 changed files with 25 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rocm_smi/src/rocm_smi.cc
+14
View File
@@ -766,6 +766,20 @@ rsmi_dev_ecc_count_get(uint32_t dv_ind, rsmi_gpu_block_t block,
assert(junk == "ce:");
fs2 >> ec->correctable_err;
ec->deferred_err = 0;
if (val_vec.size() > 2) {
std::istringstream fs3(val_vec[2]);
fs3 >> junk;
if (junk == "de:") {
fs3 >> ec->deferred_err;
} else {
ss << __PRETTY_FUNCTION__
<< "Trying to get the de count, but got " << junk
<< " ignore the defer count";
LOG_ERROR(ss);
}
}
ss << __PRETTY_FUNCTION__ << " | ======= end ======="
<< ", reporting " << amd::smi::getRSMIStatusString(ret);;
LOG_TRACE(ss);