linux-kernel - Re: INFO: rcu detected stall in sys

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

Message-ID: <96b956f4-62cb-83e6-38c2-ca698a862282@moonlit-rail.com>
Date:   Wed, 4 Mar 2020 03:40:20 -0500
From:   Kris Karas <linux-1993@...nlit-rail.com>
To:     syzbot <syzbot+0c5c2dbf76930df91489@...kaller.appspotmail.com>
Cc:     David Miller <davem@...emloft.net>,
        Herbert Xu <herbert@...dor.apana.org.au>,
        "open list:HARDWARE RANDOM NUMBER GENERATOR CORE" 
        <linux-crypto@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        syzkaller-bugs <syzkaller-bugs@...glegroups.com>,
        Kate Stewart <kstewart@...uxfoundation.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Eric Biggers <ebiggers@...nel.org>, allison@...utok.net
Subject: Re: INFO: rcu detected stall in sys_keyctl

Resending this to all the original CCs per suggestion of Dmitry.
I'm not a member of linux-crypto, no idea if it will bounce; in any 
case, the OOPS I saw does not appear to be crypto related.

Dmitry Vyukov wrote:
> syzbot wrote:
>> Call Trace:
>>   <IRQ>
>>   __dump_stack lib/dump_stack.c:77 [inline]
>>   dump_stack+0x197/0x210 lib/dump_stack.c:118
>>   nmi_cpu_backtrace.cold+0x70/0xb2 lib/nmi_backtrace.c:101
>>   nmi_trigger_cpumask_backtrace+0x23b/0x28b lib/nmi_backtrace.c:62
>>   arch_trigger_cpumask_backtrace+0x14/0x20 
>> arch/x86/kernel/apic/hw_nmi.c:38
>>   trigger_single_cpu_backtrace include/linux/nmi.h:164 [inline]
>>   rcu_dump_cpu_stacks+0x183/0x1cf kernel/rcu/tree_stall.h:254
>>   print_cpu_stall kernel/rcu/tree_stall.h:475 [inline]
>>   check_cpu_stall kernel/rcu/tree_stall.h:549 [inline]
>>   rcu_pending kernel/rcu/tree.c:3030 [inline]
>>   rcu_sched_clock_irq.cold+0x51a/0xc37 kernel/rcu/tree.c:2276
>>   update_process_times+0x2d/0x70 kernel/time/timer.c:1726
>>   tick_sched_handle+0xa2/0x190 kernel/time/tick-sched.c:171
>>   tick_sched_timer+0x53/0x140 kernel/time/tick-sched.c:1314
>>   __run_hrtimer kernel/time/hrtimer.c:1517 [inline]
>>   __hrtimer_run_queues+0x364/0xe40 kernel/time/hrtimer.c:1579
>>   hrtimer_interrupt+0x314/0x770 kernel/time/hrtimer.c:1641
>>   local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1119 [inline]
>>   smp_apic_timer_interrupt+0x160/0x610 arch/x86/kernel/apic/apic.c:1144
>>   apic_timer_interrupt+0xf/0x20 arch/x86/entry/entry_64.S:829
>>   </IRQ>
>>
> +lib/mpi maintainers
>
> I wonder if this can also be triggered by remote actors (tls, wifi, 
> usb, etc).
>

This looks somewhat similar to an OOPS + rcu stall I reported earlier in 
reply to Greg KH's announcement of 5.5.7:

     rcu: INFO: rcu_sched self-detected stall on CPU
     rcu:    14-....: (20999 ticks this GP) 
idle=216/1/0x4000000000000002 softirq=454/454 fqs=5250
             (t=21004 jiffies g=-755 q=1327)
     NMI backtrace for cpu 14
     CPU: 14 PID: 520 Comm: pidof Tainted: G      D           5.5.7 #1
     Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./X470 
Taichi, BIOS P3.50 07/18/2019
     Call Trace:
      <IRQ>
      dump_stack+0x50/0x70
      nmi_cpu_backtrace.cold+0x14/0x53
      ? lapic_can_unplug_cpu.cold+0x44/0x44
      nmi_trigger_cpumask_backtrace+0x7b/0x88
      rcu_dump_cpu_stacks+0x7b/0xa9
      rcu_sched_clock_irq.cold+0x152/0x39b
      update_process_times+0x1f/0x50
      tick_sched_timer+0x40/0x90
      ? tick_sched_do_timer+0x50/0x50
      __hrtimer_run_queues+0xdd/0x180
      hrtimer_interrupt+0x108/0x230
      smp_apic_timer_interrupt+0x53/0xa0
      apic_timer_interrupt+0xf/0x20
      </IRQ>

I don't have a reproducer for it, either.  It showed up in 5.5.7 (but 
might be from earlier as it reproduces so infrequently).

Kris