linux-kernel - [3.0-rc1] mce: unsafe smp_processor

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>] [day] [month] [year] [list]

Message-ID: <BANLkTi=_37v5C3-yK8Nn2Kbo3gC0mUyY=g@mail.gmail.com>
Date:	Wed, 1 Jun 2011 00:36:39 +0800
From:	Daniel J Blueman <daniel.blueman@...il.com>
To:	Huang Ying <ying.huang@...el.com>
Cc:	"H. Peter Anvin" <hpa@...or.com>,
	Linux Kernel <linux-kernel@...r.kernel.org>
Subject: [3.0-rc1] mce: unsafe smp_processor_id usage

In 3.0-rc1 with preemption enabled, we hit a warning in
mce_raise_notify [1]; the value of 'cpu' in mce_raise_notify isn't
guaranteed correct if preemption occurs. It looks like we need to
disable preemption along part of the path, or to use
raw_smp_processor_id, since the changing the cpumask is safe.

If raw_smp_processor_id, I can cook and test a patch pretty quick.

Thanks,
  Daniel

--- [1]

BUG: using smp_processor_id() in preemptible [00000000] code: Xorg/990
caller is mce_raise_notify+0x2e/0x120
Pid: 990, comm: Xorg Tainted: G        WC  3.0.0-rc1-100debug+ #1
Call Trace:
 [<ffffffff81343b64>] debug_smp_processor_id+0xe4/0x100
 [<ffffffff8101a5ce>] mce_raise_notify+0x2e/0x120
 [<ffffffff81030344>] ? __change_page_attr+0x164/0x240
 [<ffffffff8108421b>] notifier_call_chain+0x8b/0x100
 [<ffffffff8108483c>] __atomic_notifier_call_chain+0x6c/0xb0
 [<ffffffff810847d0>] ? register_reboot_notifier+0x20/0x20
 [<ffffffff81084891>] atomic_notifier_call_chain+0x11/0x20
 [<ffffffff810849fe>] notify_die+0x2e/0x30
 [<ffffffff81003695>] do_general_protection+0x145/0x180
 [<ffffffff8171aff4>] ? restore_args+0x30/0x30
 [<ffffffff8171b1af>] general_protection+0x1f/0x30
 [<ffffffff81509e49>] ? evdev_poll+0x29/0x70
 [<ffffffff81153c92>] do_select+0x312/0x610
 [<ffffffff81153980>] ? poll_select_set_timeout+0x90/0x90
 [<ffffffff81153400>] ? poll_freewait+0xb0/0xb0
 [<ffffffff811534f0>] ? __pollwait+0xf0/0xf0
 [<ffffffff811534f0>] ? __pollwait+0xf0/0xf0
 [<ffffffff811534f0>] ? __pollwait+0xf0/0xf0
 [<ffffffff811534f0>] ? __pollwait+0xf0/0xf0
 [<ffffffff8112b72f>] ? check_object+0xaf/0x270
 [<ffffffff8112bd96>] ? init_object+0x46/0x80
 [<ffffffff8112d17a>] ? free_debug_processing+0x17a/0x290
 [<ffffffff810933dd>] ? trace_hardirqs_off+0xd/0x10
 [<ffffffff811541af>] core_sys_select+0x21f/0x430
 [<ffffffff81153fd7>] ? core_sys_select+0x47/0x430
 [<ffffffff810704ae>] ? set_current_blocked+0x4e/0x60
 [<ffffffff81002612>] ? handle_signal+0x102/0x220
 [<ffffffff8100279e>] ? do_signal+0x6e/0x180
 [<ffffffff8115459a>] sys_select+0x3a/0x100
 [<ffffffff8171b73b>] system_call_fastpath+0x16/0x1b
-- 
Daniel J Blueman
--
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/