Uploaded sanitized BC250/ROCm Repository.

2026-08-20 00:45:43 +02:00
parent 7d2184f1e8
commit d7d22e93b3
678 changed files with 65963 additions and 1 deletions
@@ -0,0 +1,21 @@
+#!/bin/bash
+# GPU health check after hung process
+echo "=== GPU processes ==="
+lsof /dev/kfd 2>/dev/null | head -10
+lsof /dev/dri/renderD128 2>/dev/null | head -10
+
+echo ""
+echo "=== dmesg GPU errors (last 50 lines) ==="
+dmesg | grep -iE "amdgpu|gfx|gpu|drm|kfd|reset|fault|error|hang" | tail -50
+
+echo ""
+echo "=== rocm-smi ==="
+rocm-smi 2>&1 | head -20
+
+echo ""
+echo "=== Any python processes? ==="
+pgrep -la python | head -10
+
+echo ""
+echo "=== System uptime ==="
+uptime