[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PULL 3/5] nbd: fix uninitialized variable warning

From: Eric Blake
Subject: Re: [PULL 3/5] nbd: fix uninitialized variable warning
Date: Thu, 6 Feb 2020 10:55:06 -0600
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.4.1

On 1/8/20 10:02 AM, Laurent Vivier wrote:
From: Pan Nengyuan <address@hidden>

/mnt/sdb/qemu/nbd/server.c: In function 'nbd_handle_request':
/mnt/sdb/qemu/nbd/server.c:2313:9: error: 'ret' may be used uninitialized in 
this function [-Werror=maybe-uninitialized]
      int ret;

Reported-by: Euler Robot <address@hidden>
Signed-off-by: Pan Nengyuan <address@hidden>
Reviewed-by: Richard Henderson <address@hidden>
Reviewed-by: Laurent Vivier <address@hidden>
Message-Id: <address@hidden>
Signed-off-by: Laurent Vivier <address@hidden>
  nbd/server.c | 10 +---------
  1 file changed, 1 insertion(+), 9 deletions(-)

This patch is broken; I first noticed it when testing libnbd, where the symptoms are a hung interop/dirty-bitmap.sh test.

diff --git a/nbd/server.c b/nbd/server.c
index 24ebc1a80571..87fcd2e7bfac 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -2384,20 +2384,12 @@ static coroutine_fn int nbd_handle_request(NBDClient 
-                if (ret < 0) {
-                    return ret;
-                }
-            }
-            if (client->export_meta.bitmap) {
+            } else {              /* client->export_meta.bitmap */
                  ret = nbd_co_send_bitmap(client, request->handle,
                                           request->from, request->len,
                                           true, NBD_META_ID_DIRTY_BITMAP, 
-                if (ret < 0) {
-                    return ret;
-                }

The NBD spec says that clients can request more than one meta context at a time. Qemu does not (and hence we didn't notice the breakage until now), but libnbd does, and this change prevents qemu from obeying the spec (if the client requested both contexts, then we must return both, and not short-circuit after the first).

I'll post a patch to fix this unintended regression soon.

Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org

reply via email to

[Prev in Thread] Current Thread [Next in Thread]