linux-kernel - Re: [PATCH] net: guard against coalescing packets from buggy network drivers

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: <1398440595.29914.77.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Fri, 25 Apr 2014 08:43:15 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Svenning Sørensen <sss@...omea.com>
Cc:	"David S. Miller" <davem@...emloft.net>,
	linux-kernel@...r.kernel.org, netdev@...r.kernel.org
Subject: Re: [PATCH] net: guard against coalescing packets from buggy
 network drivers

On Fri, 2014-04-25 at 17:02 +0200, Svenning Sørensen wrote:
> This patch adds a guard against coalescing packets received by buggy network
> drivers that don't initialize skb->tail properly.
> 
> Coalescing such packets will corrupt data in a way that makes it hard to
> track down the real cause of the problem. Observed symptoms are a flood of
> WARNs by tcp_recvmsg at random times, leaving no clear indication as to why
> the packet buffers were corrupted. See for example this thread:
> https://groups.google.com/forum/#!topic/linux.kernel/A8ZI9aXooSU
> 
> Obviously the correct thing to do is fixing the drivers in question (eg. by
> using skb_put() instead of just setting skb->len) - this patch will 
> hopefully
> help to track them down with less effort than I had to put into it :)
> 
> NB: in my case it was an out-of-tree Bluegiga WiFi driver; a quick look 
> indicates
> that there are in-tree drivers (for hardware that I don't own) with the 
> same bug.
> 
> 
> Signed-off-by: Svenning Soerensen <sss@...omea.com>

If you want a debugging patch, this should happen way before TCP stack
or even IP stack.

Say you want to use this buggy driver on a router, this code path will
never bit hit.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/