netdev - Re: [PATCH] conn_client: Add check for rxpc channel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <1159570.1590937744@warthog.procyon.org.uk>
Date:   Sun, 31 May 2020 16:09:04 +0100
From:   David Howells <dhowells@...hat.com>
To:     gaurav singh <gaurav1086@...il.com>
Cc:     dhowells@...hat.com, davem@...emloft.net, kuba@...nel.org,
        linux-afs@...ts.infradead.org, netdev@...r.kernel.org,
        trivial@...nel.org
Subject: Re: [PATCH] conn_client: Add check for rxpc channel

Hi,

Note that "conn_client:" isn't a suitable tag for the subject.  You should use
"rxrpc:" instead.

How did you find this by the way?  You shouldn't get a NULL pointer there.
Either the call is waiting for a channel to be assigned (in which case the
condition on:

	if (!list_empty(&call->chan_wait_link)) {

will be true) or it should have been assigned a channel, in which case chan
will not be NULL.

Note that the function takes the lock under which this is managed
(conn->channel_lock) across this, so it shouldn't change.

Even __rxrpc_disconnect_call(), which is called to implicitly close out a call
that gets superseded on its channel, doesn't stop
rxrpc_disconnect_client_call() from finding the channel.

David