ci: tolerate stale NFS file handles on rm -rf build

Suppress errors from stale NFS file handles when wiping the build
directory before a fresh build. Stale-handle files are inaccessible
and cannot cause SIGILL, so ignoring rm failures is safe.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: sbryngelson

.github/scripts/prebuild-case-optimization.sh

@@ -22,7 +22,7 @@ case "$cluster" in
     *) echo "ERROR: Unknown cluster '$cluster'"; exit 1 ;;
 esac
 
-rm -rf build
+rm -rf build 2>/dev/null || true
 
 . ./mfc.sh load -c "$flag" -m g
 

.github/workflows/common/bench.sh

@@ -25,7 +25,7 @@ fi
 # Phoenix builds inside SLURM; Frontier pre-builds via build.sh on the login node.
 # Phoenix: always nuke stale builds (heterogeneous compute nodes → ISA mismatch risk).
 if [ "$job_cluster" = "phoenix" ]; then
-    rm -rf build
+    rm -rf build 2>/dev/null || true
 fi
 
 if [ ! -d "build" ]; then

.github/workflows/common/test.sh

@@ -13,7 +13,8 @@ build_opts="$gpu_opts"
 # Phoenix builds inside SLURM on heterogeneous compute nodes — always start fresh
 # to avoid SIGILL from stale binaries compiled on a different microarchitecture.
 if [ "$job_cluster" = "phoenix" ]; then
-    rm -rf build
+    # Suppress stale NFS file handle errors — those files are inaccessible anyway.
+    rm -rf build 2>/dev/null || true
 fi
 
 if [ ! -d "build" ]; then

.github/workflows/frontier/build.sh

@@ -20,7 +20,7 @@ build_opts="$gpu_opts"
 
 . ./mfc.sh load -c $compiler_flag -m $([ "$job_device" = "gpu" ] && echo "g" || echo "c")
 
-rm -rf build
+rm -rf build 2>/dev/null || true
 
 source .github/scripts/retry-build.sh
 if [ "$run_bench" == "bench" ]; then