| 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
| ||
|
Message-Id: <20131109065207.124452413@linuxfoundation.org> Date: Fri, 8 Nov 2013 22:51:47 -0800 From: Greg Kroah-Hartman <gregkh@...uxfoundation.org> To: linux-kernel@...r.kernel.org Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>, stable@...r.kernel.org, Bian Yu <bianyu@...acom.com>, Jianpeng Ma <majianpeng@...il.com>, NeilBrown <neilb@...e.de> Subject: [PATCH 3.11 41/94] md: avoid deadlock when md_set_badblocks. 3.11-stable review patch. If anyone has any objections, please let me know. ------------------ From: Bian Yu <bianyu@...acom.com> commit 905b0297a9533d7a6ee00a01a990456636877dd6 upstream. When operate harddisk and hit errors, md_set_badblocks is called after scsi_restart_operations which already disabled the irq. but md_set_badblocks will call write_sequnlock_irq and enable irq. so softirq can preempt the current thread and that may cause a deadlock. I think this situation should use write_sequnlock_irqsave/irqrestore instead. I met the situation and the call trace is below: [ 638.919974] BUG: spinlock recursion on CPU#0, scsi_eh_13/1010 [ 638.921923] lock: 0xffff8800d4d51fc8, .magic: dead4ead, .owner: scsi_eh_13/1010, .owner_cpu: 0 [ 638.923890] CPU: 0 PID: 1010 Comm: scsi_eh_13 Not tainted 3.12.0-rc5+ #37 [ 638.925844] Hardware name: To be filled by O.E.M. To be filled by O.E.M./MAHOBAY, BIOS 4.6.5 03/05/2013 [ 638.927816] ffff880037ad4640 ffff880118c03d50 ffffffff8172ff85 0000000000000007 [ 638.929829] ffff8800d4d51fc8 ffff880118c03d70 ffffffff81730030 ffff8800d4d51fc8 [ 638.931848] ffffffff81a72eb0 ffff880118c03d90 ffffffff81730056 ffff8800d4d51fc8 [ 638.933884] Call Trace: [ 638.935867] <IRQ> [<ffffffff8172ff85>] dump_stack+0x55/0x76 [ 638.937878] [<ffffffff81730030>] spin_dump+0x8a/0x8f [ 638.939861] [<ffffffff81730056>] spin_bug+0x21/0x26 [ 638.941836] [<ffffffff81336de4>] do_raw_spin_lock+0xa4/0xc0 [ 638.943801] [<ffffffff8173f036>] _raw_spin_lock+0x66/0x80 [ 638.945747] [<ffffffff814a73ed>] ? scsi_device_unbusy+0x9d/0xd0 [ 638.947672] [<ffffffff8173fb1b>] ? _raw_spin_unlock+0x2b/0x50 [ 638.949595] [<ffffffff814a73ed>] scsi_device_unbusy+0x9d/0xd0 [ 638.951504] [<ffffffff8149ec47>] scsi_finish_command+0x37/0xe0 [ 638.953388] [<ffffffff814a75e8>] scsi_softirq_done+0xa8/0x140 [ 638.955248] [<ffffffff8130e32b>] blk_done_softirq+0x7b/0x90 [ 638.957116] [<ffffffff8104fddd>] __do_softirq+0xfd/0x330 [ 638.958987] [<ffffffff810b964f>] ? __lock_release+0x6f/0x100 [ 638.960861] [<ffffffff8174a5cc>] call_softirq+0x1c/0x30 [ 638.962724] [<ffffffff81004c7d>] do_softirq+0x8d/0xc0 [ 638.964565] [<ffffffff8105024e>] irq_exit+0x10e/0x150 [ 638.966390] [<ffffffff8174ad4a>] smp_apic_timer_interrupt+0x4a/0x60 [ 638.968223] [<ffffffff817499af>] apic_timer_interrupt+0x6f/0x80 [ 638.970079] <EOI> [<ffffffff810b964f>] ? __lock_release+0x6f/0x100 [ 638.971899] [<ffffffff8173fa6a>] ? _raw_spin_unlock_irq+0x3a/0x50 [ 638.973691] [<ffffffff8173fa60>] ? _raw_spin_unlock_irq+0x30/0x50 [ 638.975475] [<ffffffff81562393>] md_set_badblocks+0x1f3/0x4a0 [ 638.977243] [<ffffffff81566e07>] rdev_set_badblocks+0x27/0x80 [ 638.978988] [<ffffffffa00d97bb>] raid5_end_read_request+0x36b/0x4e0 [raid456] [ 638.980723] [<ffffffff811b5a1d>] bio_endio+0x1d/0x40 [ 638.982463] [<ffffffff81304ff3>] req_bio_endio.isra.65+0x83/0xa0 [ 638.984214] [<ffffffff81306b9f>] blk_update_request+0x7f/0x350 [ 638.985967] [<ffffffff81306ea1>] blk_update_bidi_request+0x31/0x90 [ 638.987710] [<ffffffff813085e0>] __blk_end_bidi_request+0x20/0x50 [ 638.989439] [<ffffffff8130862f>] __blk_end_request_all+0x1f/0x30 [ 638.991149] [<ffffffff81308746>] blk_peek_request+0x106/0x250 [ 638.992861] [<ffffffff814a62a9>] ? scsi_kill_request.isra.32+0xe9/0x130 [ 638.994561] [<ffffffff814a633a>] scsi_request_fn+0x4a/0x3d0 [ 638.996251] [<ffffffff813040a7>] __blk_run_queue+0x37/0x50 [ 638.997900] [<ffffffff813045af>] blk_run_queue+0x2f/0x50 [ 638.999553] [<ffffffff814a5750>] scsi_run_queue+0xe0/0x1c0 [ 639.001185] [<ffffffff814a7721>] scsi_run_host_queues+0x21/0x40 [ 639.002798] [<ffffffff814a2e87>] scsi_restart_operations+0x177/0x200 [ 639.004391] [<ffffffff814a4fe9>] scsi_error_handler+0xc9/0xe0 [ 639.005996] [<ffffffff814a4f20>] ? scsi_unjam_host+0xd0/0xd0 [ 639.007600] [<ffffffff81072f6b>] kthread+0xdb/0xe0 [ 639.009205] [<ffffffff81072e90>] ? flush_kthread_worker+0x170/0x170 [ 639.010821] [<ffffffff81748cac>] ret_from_fork+0x7c/0xb0 [ 639.012437] [<ffffffff81072e90>] ? flush_kthread_worker+0x170/0x170 This bug was introduce in commit 2e8ac30312973dd20e68073653 (the first time rdev_set_badblock was call from interrupt context), so this patch is appropriate for 3.5 and subsequent kernels. Signed-off-by: Bian Yu <bianyu@...acom.com> Reviewed-by: Jianpeng Ma <majianpeng@...il.com> Signed-off-by: NeilBrown <neilb@...e.de> Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org> --- drivers/md/md.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -8093,6 +8093,7 @@ static int md_set_badblocks(struct badbl u64 *p; int lo, hi; int rv = 1; + unsigned long flags; if (bb->shift < 0) /* badblocks are disabled */ @@ -8107,7 +8108,7 @@ static int md_set_badblocks(struct badbl sectors = next - s; } - write_seqlock_irq(&bb->lock); + write_seqlock_irqsave(&bb->lock, flags); p = bb->page; lo = 0; @@ -8223,7 +8224,7 @@ static int md_set_badblocks(struct badbl bb->changed = 1; if (!acknowledged) bb->unacked_exist = 1; - write_sequnlock_irq(&bb->lock); + write_sequnlock_irqrestore(&bb->lock, flags); return rv; } -- 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/
Powered by blists - more mailing lists