Print out the hipRuntimeVersion message from WARN to always show up (#1911)

amd-jiali · web-flow · commit 5978d2f9ab5c · 2025-10-02T11:32:32.000-05:00
Authored-by: Jiali Li &lt;jialili@amd.com&gt;
diff --git a/src/init.cc b/src/init.cc
@@ -150,7 +150,8 @@ ncclResult_t checkHsaEnvSetting() {
 
   INFO(NCCL_INIT, "Hipruntime version: %d, firmware version: %d", hipRuntimeVersion, firmwareVersion);
   if (!validHsaScratchEnvSetting(hsaScratchEnv, hipRuntimeVersion, firmwareVersion, devProp.gcnArchName)) {
-    WARN("HSA_NO_SCRATCH_RECLAIM=1 must be set to avoid RCCL perf hit, rocm ver:%d", hipRuntimeVersion);
+    // Always print out this warning message
+    printf("Fatal Error: HSA_NO_SCRATCH_RECLAIM=1 must be set to avoid performance degradation with HIP Runtime version:%d, GPU Firmware version:%d\n", hipRuntimeVersion, firmwareVersion);
     return ncclSystemError;
   }
   return ncclSuccess;

Original file line number	Diff line number	Diff line change
`@@ -150,7 +150,8 @@ ncclResult_t checkHsaEnvSetting() {`
`150`	`150`
`151`	`151`	`INFO(NCCL_INIT, "Hipruntime version: %d, firmware version: %d", hipRuntimeVersion, firmwareVersion);`
`152`	`152`	`if (!validHsaScratchEnvSetting(hsaScratchEnv, hipRuntimeVersion, firmwareVersion, devProp.gcnArchName)) {`
`153`		`- WARN("HSA_NO_SCRATCH_RECLAIM=1 must be set to avoid RCCL perf hit, rocm ver:%d", hipRuntimeVersion);`
	`153`	`+ // Always print out this warning message`
	`154`	`+ printf("Fatal Error: HSA_NO_SCRATCH_RECLAIM=1 must be set to avoid performance degradation with HIP Runtime version:%d, GPU Firmware version:%d\n", hipRuntimeVersion, firmwareVersion);`
`154`	`155`	`return ncclSystemError;`
`155`	`156`	`}`
`156`	`157`	`return ncclSuccess;`