mount: surface ENOSPC instead of endless waiting when the cluster is full (#10341)

When every volume is full, a chunk upload fails only after the assign
retry budget, the failed chunk's data is dropped, and the mount kept
accepting writes anyway. cp would crawl for hours pushing the rest of
the file through a pipeline that could not persist it, and only close()
reported an error - a generic EIO.

Poison the file handle on the first failed chunk upload so subsequent
writes fail immediately, and map "no writable volumes" / "no free
volumes" upload errors to ENOSPC so the writing process aborts with
"No space left on device". Also guard lastErr with a mutex: it was
written concurrently by uploader goroutines, and keep the first error
so later failures do not mask the root cause.

weed/mount/error_classifier.go

@@ -48,6 +48,28 @@ func grpcErrorToFuseStatus(err error) fuse.Status {
 	return fuse.EIO
 }
 
+// isStorageFullError reports whether a chunk upload failed because the cluster
+// has no writable volume left (disks full and volume growth failed). The
+// master's message reaches the mount as plain text inside
+// AssignVolumeResponse.Error, so match on text rather than a status code.
+func isStorageFullError(err error) bool {
+	if err == nil {
+		return false
+	}
+	errStr := strings.ToLower(err.Error())
+	return strings.Contains(errStr, "no writable volumes") ||
+		strings.Contains(errStr, "no free volumes")
+}
+
+// writeErrorToFuseStatus maps a dirty-page write/flush failure to the status
+// returned to the kernel: ENOSPC when the cluster is out of space, EIO otherwise.
+func writeErrorToFuseStatus(err error) fuse.Status {
+	if isStorageFullError(err) {
+		return fuse.Status(syscall.ENOSPC)
+	}
+	return fuse.EIO
+}
+
 // isRetryableFilerError reports whether a filer RPC error looks transient
 // enough to retry. It takes a conservative whitelist approach: only errors
 // that clearly describe a permanent application-level failure