| 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
| ||
|
Date: Mon, 20 Jul 2020 15:10:11 -0700
From: "Paul E. McKenney" <paulmck@...nel.org>
To: Qian Cai <cai@....pw>
Cc: Al Viro <viro@...IV.linux.org.uk>,
LKML <linux-kernel@...r.kernel.org>,
linux-fsdevel@...r.kernel.org
Subject: Re: splice() rcu_sched self-detected stall on CPU
On Mon, Jul 20, 2020 at 04:12:04PM -0400, Qian Cai wrote:
> On Tue, May 05, 2020 at 03:00:34PM -0400, Qian Cai wrote:
> >
> >
> > > On May 5, 2020, at 2:57 PM, Paul E. McKenney <paulmck@...nel.org> wrote:
> > >
> > > On Mon, May 04, 2020 at 03:11:09PM -0400, Qian Cai wrote:
> > >> Running a syscall fuzzer inside a container on linux-next floods
> > >> systems with soft lockups. It looks like stuck in this line at
> > >> iov_iter_copy_from_user_atomic(), Thoughts?
> > >>
> > >> iterate_all_kinds(i, bytes, v,
> > >> copyin((p += v.iov_len) - v.iov_len, v.iov_base, v.iov_len),
> > >> memcpy_from_page((p += v.bv_len) - v.bv_len, v.bv_page,
> > >> v.bv_offset, v.bv_len),
> > >> memcpy((p += v.iov_len) - v.iov_len, v.iov_base, v.iov_len)
> > >> )
> > >
> > > If the size being copied is large enough, something like this might happen.
> > >
> > > Is this a CONFIG_PREEMPT=n kernel? And is the size passed in to
> >
> > Yes, CONFIG_PREEMPT=n.
> >
> > > iov_iter_copy_from_user_atomic() quite large, given that this is generated
> > > by a fuzzer? If so, one thing to try is to add cond_resched() in the
> > > iterate_bvec(), iterate_kvec(), and iterate_iovec() macros.
> >
> > I’ll try that.
>
> Finally, had a chance to try it, but in
> iov_iter_copy_from_user_atomic(), it calls kmap_atomic() first which
> calls preempt_disable(). Thus, in_atomic() == 1. Later, those
> cond_resched() would trigger the "sleeping function called from invalid
> context" warnings.
OK, no cond_resched() in those macros, then! :-(
This situation sounds bad for response time as well as from a soft-lockup
viewpoint. So I will ask the stupid question... Why does this all need
preemption disabled?
I suppose that one other alternative is to limit the size.
Other thoughts?
Thanx, Paul
> Just to refresh the memory, it is quite easy to still reproduce this
> today.
>
> # git clone https://gitlab.com/cailca/linux-mm
> # cd linux-mm; make
> # ./random -x 0-100 -f
>
> [10605.410827][ C34] watchdog: BUG: soft lockup - CPU#34 stuck for 23s! [trinity-c47:134866]
> [10605.419192][ C34] Modules linked in: vfio_pci vfio_virqfd vfio_iommu_type1 vfio loop processor ip_tables x_tables sd_mod ahci libahci mlx5_core firmware_class libata dm_mirror dm_region_hash dm_log dm_mod efivarfs
> [10605.438424][ C34] irq event stamp: 439794
> [10605.442615][ C34] hardirqs last enabled at (439793): [<ffffa000101d38c4>] el1_irq+0xc4/0x140
> [10605.451313][ C34] hardirqs last disabled at (439794): [<ffffa000101d387c>] el1_irq+0x7c/0x140
> [10605.460013][ C34] softirqs last enabled at (439790): [<ffffa000101d1b50>] efi_header_end+0xb50/0x14d4
> [10605.469494][ C34] softirqs last disabled at (439783): [<ffffa000102bd898>] irq_exit+0x440/0x510
> [10605.478368][ C34] CPU: 34 PID: 134866 Comm: trinity-c47 Not tainted 5.8.0-rc5-next-20200717+ #2
> [10605.487236][ C34] Hardware name: HPE Apollo 70 /C01_APACHE_MB , BIOS L50_5.13_1.11 06/18/2019
> [10605.497669][ C34] pstate: 40400009 (nZcv daif +PAN -UAO BTYPE=--)
> [10605.503938][ C34] pc : iov_iter_copy_from_user_atomic+0x640/0xa68
> [10605.510208][ C34] lr : generic_perform_write+0x24c/0x370
> [10605.515691][ C34] sp : ffff008921b57580
> [10605.519697][ C34] x29: ffff008921b57580 x28: 0000000000000000
> [10605.525712][ C34] x27: dfffa00000000000 x26: ffff00895cbe6500
> [10605.531726][ C34] x25: 0000000000000000 x24: 0000000000000003
> [10605.537740][ C34] x23: ffff008921b57b90 x22: 0000000000000000
> [10605.543754][ C34] x21: 0000000000000000 x20: 0000000000000004
> [10605.549769][ C34] x19: ffff000c5e817d39 x18: 0000000000000000
> [10605.555782][ C34] x17: 0000000000000000 x16: 0000000000000000
> [10605.561796][ C34] x15: 0000000000000000 x14: 1fffe0112436ae4a
> [10605.567809][ C34] x13: ffff8012adcac617 x12: 1fffe012adcac616
> [10605.573823][ C34] x11: 1fffe012adcac616 x10: ffff8012adcac616
> [10605.579838][ C34] x9 : dfffa00000000000 x8 : ffff00956e5630b7
> [10605.585852][ C34] x7 : ffff80112436af75 x6 : ffff8012adcac617
> [10605.591865][ C34] x5 : 0000000000000000 x4 : 0000000000000004
> [10605.597886][ C34] x3 : ffff00895cbe6500 x2 : 1fffe0112b97cca1
> [10605.603926][ C34] x1 : 0000000000000007 x0 : ffff00895cbe650c
> [10605.609941][ C34] Call trace:
> [10605.613083][ C34] iov_iter_copy_from_user_atomic+0x640/0xa68
> [10605.619003][ C34] generic_perform_write+0x24c/0x370
> [10605.624140][ C34] __generic_file_write_iter+0x280/0x428
> [10605.629624][ C34] generic_file_write_iter+0x2a4/0x8e0
> [10605.634936][ C34] do_iter_readv_writev+0x348/0x590
> [10605.639986][ C34] do_iter_write+0x124/0x468
> [10605.644428][ C34] vfs_iter_write+0x54/0x98
> [10605.648786][ C34] iter_file_splice_write+0x4a0/0x860
> [10605.654009][ C34] do_splice_from+0x60/0x120
> [10605.658452][ C34] do_splice+0x7ac/0x1310
> [10605.662633][ C34] __arm64_sys_splice+0x118/0x1e0
> [10605.667510][ C34] do_el0_svc+0x124/0x228
> [10605.671693][ C34] el0_sync_handler+0x260/0x410
> [10605.676395][ C34] el0_sync+0x140/0x180
> [10648.198030][ C34] rcu: INFO: rcu_sched self-detected stall on CPU
> [10648.204756][ C34] rcu: 34-....: (6472 ticks this GP) idle=666/1/0x4000000000000002 softirq=36485/36485 fqs=1901
> [10648.216008][ C34] (t=6501 jiffies g=587637 q=293554)
> [10648.221240][ C34] Task dump for CPU 34:
> [10648.225248][ C34] trinity-c47 R running task 25856 134866 106496 0x00000007
> [10648.233097][ C34] Call trace:
> [10648.236245][ C34] dump_backtrace+0x0/0x398
> [10648.240603][ C34] show_stack+0x14/0x20
> [10648.244615][ C34] sched_show_task+0x498/0x608
> [10648.249232][ C34] dump_cpu_task+0xcc/0xdc
> [10648.253503][ C34] rcu_dump_cpu_stacks+0x248/0x294
> [10648.258467][ C34] rcu_sched_clock_irq+0x12f4/0x1998
> [10648.263606][ C34] update_process_times+0x2c/0x90
> [10648.268484][ C34] tick_sched_handle+0x68/0x120
> [10648.273187][ C34] tick_sched_timer+0x60/0x128
> [10648.277802][ C34] __hrtimer_run_queues+0x7b4/0x1388
> [10648.282940][ C34] hrtimer_interrupt+0x284/0x640
> [10648.287732][ C34] arch_timer_handler_phys+0x48/0x68
> [10648.292871][ C34] handle_percpu_devid_irq+0x248/0xde0
> [10648.298194][ C34] generic_handle_irq+0x74/0xa0
> [10648.302907][ C34] __handle_domain_irq+0xa0/0x160
> [10648.307784][ C34] gic_handle_irq+0xc0/0x1a0
> [10648.312227][ C34] el1_irq+0xbc/0x140
> [10648.316064][ C34] iov_iter_copy_from_user_atomic+0x668/0xa68
> [10648.321984][ C34] generic_perform_write+0x24c/0x370
> [10648.327121][ C34] __generic_file_write_iter+0x280/0x428
> [10648.332605][ C34] generic_file_write_iter+0x2a4/0x8e0
> [10648.337916][ C34] do_iter_readv_writev+0x348/0x590
> [10648.342966][ C34] do_iter_write+0x124/0x468
> [10648.347407][ C34] vfs_iter_write+0x54/0x98
> [10648.351764][ C34] iter_file_splice_write+0x4a0/0x860
> [10648.356987][ C34] do_splice_from+0x60/0x120
> [10648.361430][ C34] do_splice+0x7ac/0x1310
> [10648.365611][ C34] __arm64_sys_splice+0x118/0x1e0
> [10648.370489][ C34] do_el0_svc+0x124/0x228
> [10648.374672][ C34] el0_sync_handler+0x260/0x410
> [10648.379374][ C34] el0_sync+0x140/0x180
>
> [10718.696867][ T1302] INFO: task trinity-c44:135125 blocked for more than 123 seconds.
> [10718.704644][ T1302] Tainted: G L 5.8.0-rc5-next-20200717+ #2
> [10718.712163][ T1302] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [10718.720728][ T1302] trinity-c44 D25760 135125 1 0x00000001
> [10718.727040][ T1302] Call trace:
> [10718.730222][ T1302] __switch_to+0x1a0/0x258
> [10718.734496][ T1302] __schedule+0x770/0x18d0
> [10718.738804][ T1302] schedule+0x1e4/0x3e8
> [10718.742824][ T1302] rwsem_down_write_slowpath+0x80c/0xe28
> [10718.748359][ T1302] down_write+0x208/0x230
> [10718.752547][ T1302] generic_file_write_iter+0xd0/0x8e0
> [10718.757774][ T1302] do_iter_readv_writev+0x348/0x590
> [10718.762864][ T1302] do_iter_write+0x124/0x468
> [10718.767311][ T1302] vfs_writev+0x15c/0x250
> [10718.771543][ T1302] do_pwritev+0x128/0x160
> [10718.775732][ T1302] __arm64_sys_pwritev+0x88/0xc8
> [10718.780561][ T1302] do_el0_svc+0x124/0x228
> [10718.784749][ T1302] el0_sync_handler+0x260/0x410
> [10718.789491][ T1302] el0_sync+0x140/0x180
Powered by blists - more mailing lists