RE: [BUG] run rcutorture BUG: unable to handle page fault for address: ffffffff84d05000

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> Test based on the following branches:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git  branch dev, arch X86_64 
> 
> 
> runqemu kvm slirp nographic qemuparams="-m 1024 -smp 4" bootparams="nokaslr console=ttyS0 rcutorture.torture_type=rcu rcutorture.shuffle_interval=0 isolcpus=0,1 rcu_nocbs=0,1 nohz_full=0,1 rcutree.dump_tree=1 rcutorture.onoff_holdoff=30 rcutorture.onoff_interval=10" -d
> 
> [   39.392925] BUG: unable to handle page fault for address: ffffffff84d05000
> [   39.393244] #PF: supervisor read access in kernel mode
> [   39.393480] #PF: error_code(0x0000) - not-present page
> [   39.393715] PGD 3e19067 P4D 3e19067 PUD 3e1a063 PMD 800ffffffb3ff062
> [   39.394014] Oops: 0000 [#1] PREEMPT SMP KASAN PTI
> [   39.394231] CPU: 0 PID: 16 Comm: rcu_preempt Not tainted 6.2.0-rc1-yocto-standard+ #635
> [   39.394590] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.o4
> [   39.395085] RIP: 0010:do_raw_spin_trylock+0x70/0x120
> [   39.395320] Code: 81 c7 00 f1 f1 f1 f1 c7 40 04 04 f3 f3 f3 65 48 8b 04 25 28 00 00 00 48 89 45 e0 31 c0 e8 c8 0
> [   39.396143] RSP: 0018:ffff8880072d7b30 EFLAGS: 00010046
> [   39.396381] RAX: 0000000000000000 RBX: ffffffff84d05000 RCX: dffffc0000000000
> [   39.396703] RDX: 0000000000000003 RSI: 0000000000000004 RDI: ffffffff84d05000
> [   39.397027] RBP: ffff8880072d7ba8 R08: ffffffff811d74a0 R09: fffffbfff09a0a01
> [   39.397346] R10: ffffffff84d05003 R11: fffffbfff09a0a00 R12: 1ffff11000e5af66
> [   39.397669] R13: ffffffff84d05018 R14: ffffffff84d05000 R15: ffff8880072d7cd8
> [   39.397990] FS:  0000000000000000(0000) GS:ffff888035400000(0000) knlGS:0000000000000000
> [   39.398353] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [   39.398615] CR2: ffffffff84d05000 CR3: 000000000a22c000 CR4: 00000000001506f0
> [   39.398936] Call Trace:
> [   39.399053]  <TASK>
> [   39.399156]  ? __pfx_do_raw_spin_trylock+0x10/0x10
> [   39.399379]  ? trace_preempt_off+0x2a/0x110
> [   39.399576]  _raw_spin_lock+0x41/0x80
> [   39.399751]  ? schedule_timeout+0x242/0x580
> [   39.399945]  schedule_timeout+0x242/0x580
> [   39.400133]  ? __pfx_schedule_timeout+0x10/0x10
> [   39.400346]  ? __pfx_do_raw_spin_trylock+0x10/0x10
> [   39.400567]  ? __pfx_process_timeout+0x10/0x10
> [   39.400776]  ? _raw_spin_unlock_irqrestore+0x46/0x80
> [   39.401006]  ? prepare_to_swait_event+0xb8/0x210
> [   39.401221]  rcu_gp_fqs_loop+0x60b/0xd50
> [   39.401405]  ? rcu_gp_init+0x89c/0x1250
> [   39.401587]  ? __pfx_rcu_gp_fqs_loop+0x10/0x10
> [   39.401793]  ? _raw_spin_unlock_irqrestore+0x46/0x80
> [   39.402022]  rcu_gp_kthread+0x2b7/0x620
> [   39.402201]  ? __pfx_do_raw_spin_trylock+0x10/0x10
> [   39.402421]  ? __pfx_rcu_gp_kthread+0x10/0x10
> [   39.402625]  ? __kasan_check_read+0x11/0x20
> [   39.402818]  ? __kthread_parkme+0xe8/0x110
> [   39.403010]  ? __pfx_rcu_gp_kthread+0x10/0x10
> [   39.403213]  kthread+0x192/0x1d0
> [   39.403366]  ? __pfx_kthread+0x10/0x10
> [   39.403541]  ret_from_fork+0x2c/0x50
> [   39.403712]  </TASK>
> [   39.403818] Modules linked in:
> [   39.403972] CR2: ffffffff84d05000
> [   39.404128] ---[ end trace 0000000000000000 ]---
> [   39.404340] RIP: 0010:do_raw_spin_trylock+0x70/0x120
> [   39.404569] Code: 81 c7 00 f1 f1 f1 f1 c7 40 04 04 f3 f3 f3 65 48 8b 04 25 28 00 00 00 48 89 45 e0 31 c0 e8 c8 0
> [   39.405400] RSP: 0018:ffff8880072d7b30 EFLAGS: 00010046
> [   39.405639] RAX: 0000000000000000 RBX: ffffffff84d05000 RCX: dffffc0000000000
> [   39.405959] RDX: 0000000000000003 RSI: 0000000000000004 RDI: ffffffff84d05000
> [   39.406281] RBP: ffff8880072d7ba8 R08: ffffffff811d74a0 R09: fffffbfff09a0a01
> [   39.406602] R10: ffffffff84d05003 R11: fffffbfff09a0a00 R12: 1ffff11000e5af66
> [   39.406922] R13: ffffffff84d05018 R14: ffffffff84d05000 R15: ffff8880072d7cd8
> [   39.407245] FS:  0000000000000000(0000) GS:ffff888035400000(0000) knlGS:0000000000000000
> [   39.407607] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [   39.407871] CR2: ffffffff84d05000 CR3: 000000000a22c000 CR4: 00000000001506f0
> [   39.408195] Kernel panic - not syncing: Fatal exception
> [   39.408450] Kernel Offset: disabled
> [   39.408615] ---[ end Kernel panic - not syncing: Fatal exception ]---
> 
> After remove isolcpus=0,1 and nohz_full=0,1, there is no Oops.
>

Only add nohz_full=0,1 bootparams, the Oops can be reproduced.

runqemu kvm slirp nographic qemuparams="-m 1024 -smp 4" bootparams=
"console=ttyS0 nohz_full=0,1 rcu_nocbs=0,1 sched_verbose" -d

root@qemux86-64:~# echo 0 > /sys/devices/system/cpu/cpu2/online
root@qemux86-64:~# echo 0 > /sys/devices/system/cpu/cpu3/online

schedule_timeout()
->__mod_timer()
 ->get_target_base(base, timer->flags)
   ->get_timer_cpu_base(tflags, get_nohz_timer_target());
     ->cpu = get_nohz_timer_target()
             ->housekeeping_any_cpu(HK_TYPE_TIMER)
                     /*housekeeping.cpumasks[type] is 2-3*/
                     /*cpu_online_mask is 0-1*/
               ->cpu = cpumask_any_and(housekeeping.cpumasks[type],
                        cpu_online_mask);
             /*cpu value is 4*/
     ->new_base = per_cpu_ptr(&timer_bases[BASE_DEF], cpu);
   /*new_base is illegal address*/
 ->if (base != new_base)
   ->raw_spin_lock(&new_base->lock); ==> trigger Oops

The follow change can fix Oops

diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c
index 373d42c707bc..e255eb83f14f 100644
--- a/kernel/sched/isolation.c
+++ b/kernel/sched/isolation.c
@@ -46,7 +46,9 @@ int housekeeping_any_cpu(enum hk_type type)
                        if (cpu < nr_cpu_ids)
                                return cpu;

-                       return cpumask_any_and(housekeeping.cpumasks[type], cpu_online_mask);
+                       cpu = cpumask_any_and(housekeeping.cpumasks[type], cpu_online_mask);
+                       if (cpu >= nr_cpu_ids)
+                               return smp_processor_id();
                }
        }
        return smp_processor_id();

Have send patch.

Thanks
Zqiang

>That certainly isn't what we want the kernel to be doing!  ;-)
>
>Could you please try bisecting?
>
>							Thanx, Paul




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux