[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20161222012101.GF1555@ZenIV.linux.org.uk>
Date: Thu, 22 Dec 2016 01:21:01 +0000
From: Al Viro <viro@...IV.linux.org.uk>
To: Jon Maloy <jon.maloy@...csson.com>
Cc: davem@...emloft.net, netdev@...r.kernel.org,
parthasarathy.bhuvaragan@...csson.com, ying.xue@...driver.com,
maloy@...jonn.com, tipc-discussion@...ts.sourceforge.net
Subject: Re: [PATCH net 1/1] tipc: revert use of copy_from_iter_full()
On Wed, Dec 21, 2016 at 08:01:37PM -0500, Jon Maloy wrote:
> commit cbbd26b8b1a6 ("[iov_iter] new primitives - copy_from_iter_full()
> and friends") replaced calls to copy_from_iter() in the function
> tipc_msg_build(). This causes a an immediate crash as follows:
Very interesting.
> [ 1209.597076] BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
> [ 1209.607025] IP: copy_from_iter_full+0x43/0x290
> [ 1209.689257] tipc_msg_build+0xe1/0x590 [tipc]
> [ 1209.691479] ? _raw_spin_unlock_bh+0x1e/0x20
> [ 1209.694641] ? tipc_node_find+0x30/0xa0 [tipc]
> [ 1209.696789] __tipc_sendmsg+0x189/0x480 [tipc]
> [ 1209.699017] ? remove_wait_queue+0x4d/0x60
> [ 1209.700354] tipc_connect+0x15f/0x1b0 [tipc]
> [ 1209.701684] SYSC_connect+0xd9/0x110
I don't believe that it's something tipc-specific; could you post an objdump
of copy_from_iter_full() in your kernel? That smells like a bug in there
and it really ought to be fixed...
Powered by blists - more mailing lists