netdev - linux-next - crashes in rhashtable

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-next>] [day] [month] [year] [list]

Date:	Thu, 29 Jan 2015 09:57:42 -0500
From:	Valdis Kletnieks <Valdis.Kletnieks@...edu>
To:	Thomas Graf <tgraf@...g.ch>
Cc:	linux-kernel@...r.kernel.org, netdev@...r.kernel.org
Subject: linux-next - crashes in rhashtable_shrink

For the past few kernels, I've seen occasional crashes in rhashtable_shrink.
Looks like the last thing to touch that code was the patch series:

Subject	[PATCH 0/9 net-next v2] rhashtable: Per bucket locks & deferred table resizing
Date	Fri, 2 Jan 2015 23:00:13 +0100

Most common appears to be during boot, while ip6tables-restore is
running (though I admit I am not sure if ip6tables-restore was actually
running, or if it ended up hung because somebody else had died while holding
a lock.). But I have also seen it after 10-12 hours of uptime

[   53.762550] BUG: unable to handle kernel NULL pointer dereference at           (null)
[   53.763607] IP: [<ffffffffab2a2898>] bucket_tail+0x57/0xaf

[   53.771622] Workqueue: events rht_deferred_worker
[   53.772715] task: ffff880224d20890 ti: ffff880224d24000 task.ti: ffff880224d24000

[   53.793546] Call Trace:
[   53.795374]  [<ffffffffab2a38fd>] rhashtable_shrink+0x13e/0x1a0
[   53.797197]  [<ffffffffab2a3a02>] rht_deferred_worker+0xa3/0xb4
[   53.798857]  [<ffffffffab058ef6>] process_one_work+0x2d8/0x54d
[   53.800070]  [<ffffffffab0594d8>] worker_thread+0x343/0x46b
[   53.801290]  [<ffffffffab059195>] ? process_scheduled_works+0x2a/0x2a
[   53.802513]  [<ffffffffab05e19d>] kthread+0xea/0xf2
[   53.803733]  [<ffffffffab6c50ba>] ? _raw_spin_unlock_irq+0x29/0x74
[   53.804957]  [<ffffffffab05e0b3>] ? __kthread_parkme+0xa8/0xa8
[   53.806176]  [<ffffffffab6c5a6c>] ret_from_fork+0x7c/0xb0
[   53.807400]  [<ffffffffab05e0b3>] ? __kthread_parkme+0xa8/0xa8
[   53.808609] Code: 4c 89 e7 e8 a1 ff ff ff 85 c0 75 1f 48 c7 c2 a9 89 b0 ab be 76 00 00 00 48 c7 c7 0d 61 b2 ab c6 05 a9 3d a3 00 01 e8 62 d6 dd ff <f6> 03 01 75 48 e8 d6 d7 de ff 85 c0 74 37 80 3d 90 3d a3 00 00
[   53.810092] RIP  [<ffffffffab2a2898>] bucket_tail+0x57/0xaf
[   53.811414]  RSP <ffff880224d27cc8>
[   53.812721] CR2: 0000000000000000
[   53.814033] ---[ end trace 82b77fc1a90284ff ]---

Unfortunately, I can't easily bisect to verify or narrow down further, as I'm only
seeing it 2-3 times a week....


Content of type "application/pgp-signature" skipped