netdev - Re: [PATCH v2 net 1/1] tipc: fix kernel warning when sending SYN message

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

Message-ID: <20230210191606.29b8db03@kernel.org>
Date:   Fri, 10 Feb 2023 19:16:06 -0800
From:   Jakub Kicinski <kuba@...nel.org>
To:     Tung Quang Nguyen <tung.q.nguyen@...tech.com.au>
Cc:     Paolo Abeni <pabeni@...hat.com>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "davem@...emloft.net" <davem@...emloft.net>,
        "edumazet@...gle.com" <edumazet@...gle.com>,
        "jmaloy@...hat.com" <jmaloy@...hat.com>,
        "ying.xue@...driver.com" <ying.xue@...driver.com>,
        "viro@...iv.linux.org.uk" <viro@...iv.linux.org.uk>,
        "syzbot+d43608d061e8847ec9f3@...kaller.appspotmail.com" 
        <syzbot+d43608d061e8847ec9f3@...kaller.appspotmail.com>
Subject: Re: [PATCH v2 net 1/1] tipc: fix kernel warning when sending SYN
 message

On Thu, 9 Feb 2023 11:10:16 +0000 Tung Quang Nguyen wrote:
> >>  	msg_set_size(mhdr, msz);
> >>
> >> +	if (!dsz)
> >> +		iov_iter_init(&m->msg_iter, ITER_SOURCE, NULL, 0, 0);  
> >
> >It looks like the root cause of the problem is that not all (indirect)
> >callers of tipc_msg_build() properly initialize the iter.
> >
> >tipc_connect() is one of such caller, but AFAICS even tipc_accept() can
> >reach tipc_msg_build() without proper iter initialization - via
> >__tipc_sendstream -> __tipc_sendmsg.
> >
> >I think it's better if you address the issue in relevant callers,
> >avoiding unneeded and confusing code in tipc_msg_build().  
>
> I am fully aware of callers (without initializing iovec) of this
> function. My intention was to make as less change as possible.

General kernel guidance is to fix things "right" (i.e. so that the fix
doesn't have to be refactored later). 

> Do you think using  iov_iter_kvec() instead in the callers make sense
> if I go for what you suggested ?

I think so. These are the potential culprits?

$ git grep 'struct msghdr [^*]*;' -- net/tipc/
net/tipc/socket.c:      struct msghdr m = {NULL,};
net/tipc/socket.c:      struct msghdr m = {NULL,};
net/tipc/topsrv.c:      struct msghdr msg;
net/tipc/topsrv.c:      struct msghdr msg = {};