tap: With pasta, don't reset on tap errors, handle write failures

Since commit 0515adceaa ("passt, pasta: Namespace-based sandboxing,
defer seccomp policy application"), it makes no sense to close and
reopen the tap device on error: we don't have access to /dev/net/tun
after the initial setup phase.

If we hit ENOBUFS while writing (as reported: in one case because
the kernel actually ran out of memory, with another case under
investigation), or ENOSPC, we're supposed to drop whatever data we
were trying to send: there's no room for it.

Handle EINTR just like we handled EAGAIN/EWOULDBLOCK: there's no
particular reason why sending the same data should fail again.

Anything else I can think of would be an unrecoverable error: exit
with failure then.

While at it, drop a useless cast on the write() call: it takes a
const void * anyway.

Reported-by: Gianluca Stivan <me@yawnt.com>
Reported-by: Chris Kuhn <kuhnchris@kuhnchris.eu>
Fixes: 0515adceaa ("passt, pasta: Namespace-based sandboxing, defer seccomp policy application")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

tap.c

@@ -320,12 +320,23 @@ static size_t tap_send_frames_pasta(struct ctx *c,
 	size_t i;
 
 	for (i = 0; i < n; i++) {
-		if (write(c->fd_tap, (char *)iov[i].iov_base,
+		if (write(c->fd_tap, iov[i].iov_base, iov[i].iov_len) < 0) {
-			  iov[i].iov_len) < 0) {
 			debug("tap write: %s", strerror(errno));
-			if (errno != EAGAIN && errno != EWOULDBLOCK)
+
-				tap_handler(c, c->fd_tap, EPOLLERR, NULL);
+			switch (errno) {
+			case EAGAIN:
+#if EAGAIN != EWOULDBLOCK
+			case EWOULDBLOCK:
+#endif
+			case EINTR:
 				i--;
+				break;
+			case ENOBUFS:
+			case ENOSPC:
+				break;
+			default:
+				die("Write error on tap device, exiting");
+			}
 		}
 	}
 
@@ -1237,6 +1248,9 @@ void tap_handler(struct ctx *c, int fd, uint32_t events,
 			exit(EXIT_SUCCESS);
 		}
 
+		if (c->mode == MODE_PASTA)
+			die("Error on tap device, exiting");
+
 		tap_sock_init(c);
 	}
 }