c8000::kernel hangs with 16G DRAM

c8000::kernel hangs with 16G DRAM

[Simone Mannori]

Dear All,

after upgrading the RAM to 16GB (8 x 2GB modules), the kernel hangs
just after the "nouveau" module loading.
With X1 ATI video card, the "radeon" driver has the very same behaviour.
Booting without ANY video card is OK and the c8000 works perfectly
(this exlude any memory realted issues like memory
errors/defective/unmatched DIMM)

With "nouveau" the serial console dump is:

[  OK  ] Reached target Graphical Interface.
[   73.932000] nuveau 0000:80:00.0: DRM: 0xABCE: Parsing digital
output script table
         Starting Update UTMP about Syste[   74.104000] nouveau
0000:80:00.0: DRM: 0xABCE: Parsing digital output script table
m Runlevel Changes...
[  OK  ] Started Update UTMP about System Runlevel Changes.
[   95.616000] INFO: rcu_sched detected stalls on CPUs/tasks:
[   95.680000]  (detected by 1, t=5272 jiffies, g=2108, c=2107, q=121)
[   95.756000] All QSes seen, last rcu_sched kthread activity 5284
(4294916235-4294910951), jiffies_till_next_fqs=1, root ->qsmask 0x0
[   95.896000] dpkg            R  running task        0   941    938 0x00000014
[   95.984000] Backtrace:
[   96.012000]  [<0000000040215bc0>] show_stack+0x20/0x38
[   96.072000]  [<00000000402756e4>] sched_show_task+0x134/0x1d0
[   96.140000]  [<00000000402adc04>] rcu_check_callbacks+0x9bc/0x9c8
[   96.216000]  [<00000000402b2494>] update_process_times+0x74/0xd8
[   96.288000]  [<0000000040216edc>] timer_interrupt+0x1bc/0x220
[   96.356000]  [<000000004029d714>] handle_irq_event_percpu+0xb4/0x250
[   96.432000]  [<00000000402a3028>] handle_percpu_irq+0xa8/0xe8
[   96.500000]  [<000000004029c99c>] generic_handle_irq+0x4c/0x68
[   96.568000]  [<0000000040205308>] call_on_stack+0x18/0x24
[   96.636000]
[   96.652000] rcu_sched kthread starved for 5508 jiffies! g2108 c2107
f0x2 s3 ->state=0x100
[   96.752000] timer_interrupt(CPU 1): delayed! cycles 4370CCA7 rem
82659  next/now 5AA779EC42/5AA771C5E9

  Simone Mannori - ITALY

Re: c8000::kernel hangs with 16G DRAM

[Simone Mannori]

Please forget the previus post: it was a defective DIMM.
Booting with 12GB is OK and using another 2GB couple has fixed the issue.
I'm sorry for the false alarm.

   Simone Mannori - ITALY

On 14 April 2016 at 10:46, Simone Mannori <simone.mannori@gmail.com> wrote:
> Dear All,
>
> after upgrading the RAM to 16GB (8 x 2GB modules), the kernel hangs
> just after the "nouveau" module loading.
> With X1 ATI video card, the "radeon" driver has the very same behaviour.
> Booting without ANY video card is OK and the c8000 works perfectly
> (this exlude any memory realted issues like memory
> errors/defective/unmatched DIMM)
>
> With "nouveau" the serial console dump is:
>
> [  OK  ] Reached target Graphical Interface.
> [   73.932000] nuveau 0000:80:00.0: DRM: 0xABCE: Parsing digital
> output script table
>          Starting Update UTMP about Syste[   74.104000] nouveau
> 0000:80:00.0: DRM: 0xABCE: Parsing digital output script table
> m Runlevel Changes...
> [  OK  ] Started Update UTMP about System Runlevel Changes.
> [   95.616000] INFO: rcu_sched detected stalls on CPUs/tasks:
> [   95.680000]  (detected by 1, t=5272 jiffies, g=2108, c=2107, q=121)
> [   95.756000] All QSes seen, last rcu_sched kthread activity 5284
> (4294916235-4294910951), jiffies_till_next_fqs=1, root ->qsmask 0x0
> [   95.896000] dpkg            R  running task        0   941    938 0x00000014
> [   95.984000] Backtrace:
> [   96.012000]  [<0000000040215bc0>] show_stack+0x20/0x38
> [   96.072000]  [<00000000402756e4>] sched_show_task+0x134/0x1d0
> [   96.140000]  [<00000000402adc04>] rcu_check_callbacks+0x9bc/0x9c8
> [   96.216000]  [<00000000402b2494>] update_process_times+0x74/0xd8
> [   96.288000]  [<0000000040216edc>] timer_interrupt+0x1bc/0x220
> [   96.356000]  [<000000004029d714>] handle_irq_event_percpu+0xb4/0x250
> [   96.432000]  [<00000000402a3028>] handle_percpu_irq+0xa8/0xe8
> [   96.500000]  [<000000004029c99c>] generic_handle_irq+0x4c/0x68
> [   96.568000]  [<0000000040205308>] call_on_stack+0x18/0x24
> [   96.636000]
> [   96.652000] rcu_sched kthread starved for 5508 jiffies! g2108 c2107
> f0x2 s3 ->state=0x100
> [   96.752000] timer_interrupt(CPU 1): delayed! cycles 4370CCA7 rem
> 82659  next/now 5AA779EC42/5AA771C5E9
>
>   Simone Mannori - ITALY