Re: [BUG][PPC64] BUG in 2.6.26.5-rt9 causing Hang

[Gregory Haskins]

Hi Chirag

Chirag Jog wrote:
> Hi Gregory,
> We see the following BUG followed by a hang on the latest kernel 2.6.26.5-rt9 on a Power6 blade (PPC64)
> It is easily recreated by running the async_handler or sbrk_mutex (realtime tests from ltp) tests.
>   
FYI I am looking at this now.  I suspect a dequeue_pushable_task()
probably found its way inside a conditional for GROUP_SCHED and
inadventently gets compiled away if you disable the feature. 
Investigating now..

-Greg

> login: cpu 0x2: Vector: 700 (Program Check) at [c0000000e8e875d0]
>     pc: c00000000005110c: .pick_next_pushable_task+0x54/0x9c
>     lr: c000000000059f50: .push_rt_task+0x44/0x2b4
>     sp: c0000000e8e87850
>    msr: 8000000000021032
>   current = 0xc0000000ea5bb2e0
>   paca    = 0xc000000000608700
>     pid   = 2811, comm = async_handler
> kernel BUG at kernel/sched_rt.c:1041! <---------------------
> enter ? for help
> [link register   ] c000000000059f50 .push_rt_task+0x44/0x2b4
> [c0000000e8e87850] c0000000e8e878f0 (unreliable)
> [c0000000e8e87900] c00000000005a1dc .push_rt_tasks+0x1c/0x38
> [c0000000e8e87980] c00000000005a21c .post_schedule_rt+0x24/0x44
> [c0000000e8e87a10] c000000000057cbc .finish_task_switch+0xd0/0x180
> [c0000000e8e87ab0] c0000000003b6e88 .__schedule+0x6e0/0x798
> [c0000000e8e87b90] c0000000003b7148 .schedule+0xec/0x11c
> [c0000000e8e87c10] c0000000003b7a40 .do_nanosleep+0x6c/0xcc
> [c0000000e8e87c90] c000000000080738 .hrtimer_nanosleep+0x7c/0x100
> [c0000000e8e87d90] c000000000080830 .sys_nanosleep+0x74/0x94
> [c0000000e8e87e30] c0000000000086ac syscall_exit+0x0/0x40
> --- Exception: c00 (System Call) at 0000008026449844
> SP (400014185f0) is in userspace
>
>
> This is generated by the BUG_ON lines in the pick_next_pushable function
> introduced by the sched-only-push-once-per-queue.patch .
>
> The -rt kernel prior to this patch didnot give such BUGes.
>
> All this was tried with
> CONFIG_GROUP_SCHED=N
> CONFIG_RT_GROUP_SCHED=N
>
>
> Setting the options
> CONFIG_GROUP_SCHED=y
> CONFIG_RT_GROUP_SCHED=Y,
> seems to solve the problem.
>
>
>
>

Attachment: signature.asc
Description: OpenPGP digital signature