fix: Ignore failure to cleanup tmpdir (#8102)

It seems this happens when podman is still holding a handle in the dir.
We hope to share the cache and builds soon with
https://dfinity.atlassian.net/browse/NODE-1048, so that this cleanup is
no longer necessary.

Author: Bownairo

toolchains/sysimage/proc_wrapper.sh

@@ -12,5 +12,6 @@ tmpfs_tmpdir=$(mktemp -d --tmpdir "icosbuildXXXX")
 sudo mount -t tmpfs none "${tmpfs_tmpdir}"
 
 tmpdir=$(mktemp -d --tmpdir "icosbuildXXXX")
-trap 'sudo umount "${tmpfs_tmpdir}"; sudo rm -rf "$tmpdir" "${tmpfs_tmpdir}"' INT TERM EXIT
+# NOTE: Ignore failure to cleanup the directory for now. This should not be a problem after NODE-1048.
+trap 'sudo umount "${tmpfs_tmpdir}"; sudo rm -rf "$tmpdir" "${tmpfs_tmpdir}" || true' INT TERM EXIT
 TMPDIR="$tmpdir" TMPFS_TMPDIR="${tmpfs_tmpdir}" "$@"