From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from out1-smtp.messagingengine.com ([66.111.4.25]:37437 "EHLO out1-smtp.messagingengine.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726546AbeLROkn (ORCPT ); Tue, 18 Dec 2018 09:40:43 -0500 Message-ID: <71040d877679690375cea73177c5c095ab4f9b00.camel@themaw.net> Subject: Re: kernel BUG at fs/inode.c:LINE! From: Ian Kent To: Dmitry Vyukov Cc: Al Viro , syzbot , Andrew Morton , linux-fsdevel , LKML , syzkaller-bugs Date: Tue, 18 Dec 2018 22:40:36 +0800 In-Reply-To: References: <00000000000051e9c2057d31a563@google.com> <20181217072144.GQ2217@ZenIV.linux.org.uk> <95ae4c9893c89189d4309fe673ade6f389280101.camel@themaw.net> <66d497c00cffb3e4109ca0d5287c8277954d7132.camel@themaw.net> <8917bbe175172b4fd9c9f92bf2a52443eb2b1827.camel@themaw.net> Content-Type: text/plain; charset="UTF-8" Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Tue, 2018-12-18 at 14:19 +0100, Dmitry Vyukov wrote: > On Tue, Dec 18, 2018 at 1:42 PM Ian Kent wrote: > > > > On Tue, 2018-12-18 at 13:27 +0100, Dmitry Vyukov wrote: > > > On Tue, Dec 18, 2018 at 12:35 PM Ian Kent wrote: > > > > > > > > On Tue, 2018-12-18 at 18:42 +0800, Ian Kent wrote: > > > > > On Mon, 2018-12-17 at 07:21 +0000, Al Viro wrote: > > > > > > On Sun, Dec 16, 2018 at 10:11:04PM -0800, syzbot wrote: > > > > > > > Hello, > > > > > > > > > > > > > > syzbot found the following crash on: > > > > > > > > > > > > > > HEAD commit: d14b746c6c1c Add linux-next specific files for > > > > > > > 20181214 > > > > > > > git tree: linux-next > > > > > > > console output: > > > > > > > https://syzkaller.appspot.com/x/log.txt?x=13706347400000 > > > > > > > kernel config: > > > > > > > https://syzkaller.appspot.com/x/.config?x=1da6d2d18f803140 > > > > > > > dashboard link: > > > > > > > https://syzkaller.appspot.com/bug?extid=5399ed0832693e29f392 > > > > > > > compiler: gcc (GCC) 8.0.1 20180413 (experimental) > > > > > > > syz repro: > > > > > > > https://syzkaller.appspot.com/x/repro.syz?x=101032b3400000 > > > > > > > C reproducer: > > > > > > > https://syzkaller.appspot.com/x/repro.c?x=16534063400000 > > > > > > > > > > > > > > IMPORTANT: if you fix the bug, please add the following tag to the > > > > > > > commit: > > > > > > > Reported-by: syzbot+5399ed0832693e29f392@syzkaller.appspotmail.com > > > > > > > > > > > > > > slab_pre_alloc_hook mm/slab.h:423 [inline] > > > > > > > slab_alloc mm/slab.c:3365 [inline] > > > > > > > kmem_cache_alloc+0x2c4/0x730 mm/slab.c:3539 > > > > > > > __d_alloc+0xc8/0xb90 fs/dcache.c:1599 > > > > > > > ------------[ cut here ]------------ > > > > > > > kernel BUG at fs/inode.c:1566! > > > > > > > d_alloc_anon fs/dcache.c:1698 [inline] > > > > > > > d_make_root+0x43/0xc0 fs/dcache.c:1885 > > > > > > > autofs_fill_super+0x6f1/0x1c30 fs/autofs/inode.c:273 > > > > > > > > > > > > Huh? BUG is in iput(), AFAICS, so the stack trace is rather > > > > > > misreported. > > > > > > iput() can be called by d_make_root(), provided that dentry > > > > > > allocation > > > > > > fails. So the most straightforward interpretation would be that we > > > > > > had an allocation failure (injected?), followed by iput() of the > > > > > > inode > > > > > > passed to d_make_root(). Which happened to find I_CLEAR in > > > > > > ->i_state > > > > > > of that inode somehow, which should be impossible short of seriously > > > > > > buggered inode refcounting somewhere - the inode has just been > > > > > > returned > > > > > > by new_inode(), which clears i_state, and it would have to have > > > > > > passed > > > > > > clear_inode() (i.e. has been through inode eviction) since then... > > > > > > > > > > Sorry Al, that's my bad. > > > > > > > > > > See > > > > > > > > > https://www.ozlabs.org/~akpm/mmotm/broken-out/autofs-fix-possible-inode-leak-in-autofs_fill_super.patch > > > > > > > > > > I think this will fix it, I'll forward it to Andrew if you agree: > > > > > > > > Actually, looking at it again the above patch is plain not needed, > > > > dropping it and updating the patch which follows it in the series > > > > is what needs to be done. > > > > > > > > Andrew, what should I do to make this easiest for you to handle, > > > > a respost with v2 in the subject of the patch affected by dropping > > > > the above patch? > > > > > > > > Or I could repost the series with above patch dropped and the affected > > > > patch corrected? > > > > > > Hi Ian, > > > > > > If you going to amend any commits, please add: > > > Tested-by: syzbot+5399ed0832693e29f392@syzkaller.appspotmail.com > > > otherwise: > > > Reported-by: syzbot+5399ed0832693e29f392@syzkaller.appspotmail.com > > > > I was thinking about how to handle loosing that information. > > > > I don't think this amounts amending commits since Andrews mmotm is > > based on patch series so dropping a patch and updating patches before > > being merged won't capture this. > > > > Adding the "Tested-by" attribution to the updated patch prior to syzbot > > actually performing the test might be ok since it will get tested along > > the way. Although the problem patch itself won't exist any more so ... ? > > I am a bit lost. > There should be some patch that will fix this (new or amended). Either > that patch needs to include Reporter/Tested-by tag, or we will need to > tell syzbot about the fix manually with the "#syz fix:" command. One of the patches that I forwarded to Andrew was plain wrong, it isn't actually needed so it needs to be dropped altogether (this one caused the breakage). Doing so then requires a minor (unrelated) change to the next of my patches in the series I sent. Andrew forwards certain patches in his patch queue to to the linux-next tree. I'm not sure if that is done via a cumulative pull request or via actual patches. Whether dropping a patch from Andrews mmotm tree when forwarding his patch queue to linux-next requires a revert patch rather than just omitting the dropped patch (which would carry the annotation) or not I don't know, we'll have to wait and see what Andrew would like me to do. > > > OTOH, if I repost the series I think I can send them to syzbot for testing > > before forwarding to Andrew (I've done something like that before but can't > > remember how now) and add the attribution to the series. > > > > But this all depends on what is best for Andrew and what Al would like to > > see done.