On 7/16/19 5:07 AM, Marc-André Lureau wrote:
> ../block/nbd.c: In function 'nbd_co_request':
> ../block/nbd.c:745:8: error: 'local_reply.type' may be used uninitialized in
> this function [-Werror=maybe-uninitialized]
> if (chunk->type == NBD_REPLY_TYPE_NONE) {
> ^
> ../block/nbd.c:710:14: note: 'local_reply.type' was declared here
> NBDReply local_reply;
> ^~~~~~~~~~~
> ../block/nbd.c:710:14: error: 'local_reply.flags' may be used uninitialized
> in this function [-Werror=maybe-uninitialized]
> ../block/nbd.c:738:8: error: 'local_reply.<U4be0>.magic' may be used
> uninitialized in this function [-Werror=maybe-uninitialized]
> if (nbd_reply_is_simple(reply) || s->quit) {
> ^
> ../block/nbd.c:710:14: note: 'local_reply.<U4be0>.magic' was declared here
> NBDReply local_reply;
> ^~~~~~~~~~~
> cc1: all warnings being treated as errors
>
> Reported-by: Thomas Huth <th...@redhat.com>
> Signed-off-by: Marc-André Lureau <marcandre.lur...@redhat.com>
> Reviewed-by: Philippe Mathieu-Daudé <phi...@redhat.com>
> ---
> block/nbd.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
Actually, since this patch didn't fix all the cases of use of an uninit
variable, I'm wondering if this one-liner would be a better patch for
the issue (that is, fix it so that nbd_co_receive_one_chunk never leaves
reply uninit, rather than chasing down which callers might need init),
and with a perk of not relying on a gcc extension:
diff --git i/block/nbd.c w/block/nbd.c
index 8d565cc624ec..f751a8e633e5 100644
--- i/block/nbd.c
+++ w/block/nbd.c
@@ -640,6 +640,7 @@ static coroutine_fn int nbd_co_receive_one_chunk(
request_ret, qiov, payload,
errp);
if (ret < 0) {
+ memset(reply, 0, sizeof *reply);
s->quit = true;
} else {
/* For assert at loop start in nbd_connection_entry */
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization: qemu.org | libvirt.org
signature.asc
Description: OpenPGP digital signature
