From: Alexander Ivanov <alexander.ivanov@virtuozzo.com>
To: Eric Blake <eblake@redhat.com>
Cc: qemu-devel@nongnu.org, qemu-block@nongnu.org, den@virtuozzo.com,
andrey.drobyshev@virtuozzo.com, vsementsov@yandex-team.ru,
kwolf@redhat.com, hreitz@redhat.com
Subject: Re: [PATCH] nbd: Prevent NULL pointer dereference in nbd_blockdev_client_closed()
Date: Tue, 18 Jun 2024 10:44:52 +0200 [thread overview]
Message-ID: <6c103dec-cce7-4ff0-b337-21c130c495fe@virtuozzo.com> (raw)
In-Reply-To: <ooczxb5vojkt6blp2hhyfqcvltgzemtmwmigvgeg4utngwrcpo@2rolx5gvqzlr>
Hello Eric,
Do you have any ideas about the bug?
Thank you.
On 6/10/24 14:33, Eric Blake wrote:
> On Sat, Jun 08, 2024 at 11:36:59AM GMT, Alexander Ivanov wrote:
>> There is a bug reproducer in the attachment.
> Summarizing the reproducer, you are repeatedly calling QMP
> nbd-server-start/nbd-server-stop on qemu as NBD server in one thread,
> and repeatedly calling 'qemu-nbd -L' in another, to try and provoke
> the race where the server is stopped while qemu-nbd -L is still trying
> to grab information.
>
>>
>> On 6/7/24 17:00, Alexander Ivanov wrote:
>>> In some cases, the NBD server can be stopped before
>>> nbd_blockdev_client_closed() is called, causing the nbd_server variable
>>> to be nullified. This leads to a NULL pointer dereference when accessing
>>> nbd_server.
> Am I correct that the NULL pointer deref was in qemu-nbd in your
> reproducer, and not in qemu-kvm?
>
>>> Add a NULL check for nbd_server to the nbd_blockdev_client_closed()
>>> function to prevent NULL pointer dereference.
>>>
>>> Signed-off-by: Alexander Ivanov <alexander.ivanov@virtuozzo.com>
>>> ---
>>> blockdev-nbd.c | 3 +++
>>> 1 file changed, 3 insertions(+)
>>>
>>> diff --git a/blockdev-nbd.c b/blockdev-nbd.c
>>> index 213012435f..fb1f30ae0d 100644
>>> --- a/blockdev-nbd.c
>>> +++ b/blockdev-nbd.c
>>> @@ -52,6 +52,9 @@ int nbd_server_max_connections(void)
>>> static void nbd_blockdev_client_closed(NBDClient *client, bool ignored)
>>> {
>>> nbd_client_put(client);
>>> + if (nbd_server == NULL) {
>>> + return;
>>> + }
>>> assert(nbd_server->connections > 0);
> While this does indeed avoid the NULL dereference right here, I still
> want to understand why nbd_server is getting set to NULL while there
> is still an active client, and make sure we don't have any other NULL
> derefs. I'll respond again once I've studied the code a bit more.
>
>>> nbd_server->connections--;
>>> nbd_update_server_watch(nbd_server);
>> --
>> Best regards,
>> Alexander Ivanov
--
Best regards,
Alexander Ivanov
next prev parent reply other threads:[~2024-06-18 8:45 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-07 15:00 [PATCH] nbd: Prevent NULL pointer dereference in nbd_blockdev_client_closed() Alexander Ivanov
2024-06-08 9:36 ` Alexander Ivanov
2024-06-10 12:33 ` Eric Blake
2024-06-10 12:52 ` Alexander Ivanov
2024-06-18 8:44 ` Alexander Ivanov [this message]
2024-06-28 9:58 ` Alexander Ivanov
2024-07-30 2:35 ` Eric Blake
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=6c103dec-cce7-4ff0-b337-21c130c495fe@virtuozzo.com \
--to=alexander.ivanov@virtuozzo.com \
--cc=andrey.drobyshev@virtuozzo.com \
--cc=den@virtuozzo.com \
--cc=eblake@redhat.com \
--cc=hreitz@redhat.com \
--cc=kwolf@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=vsementsov@yandex-team.ru \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).