Hi,
Sebastian Andrzej Siewior <bigeasy <at> linutronix.de> writes:
> This might be a .config thing, device driver or user land behaviour.
> Usually ksoftirqd should remain (almost) idle because most of BH work
> remains in task-context and not in ksoftirqd. tasklets for instance
> should run in ksoftirqd. So a function trace or event tracing to figure
> out what is scheduling the softirq might give a pointer.
I have a problem that might be related, or at least I suffer from the same
symptoms:
Trying to run various Debian 4.1 (and 4.0 as well, IIRC) -rt kernels on a
Core i7-720QM Thinkpad, it boots up, I can log in, everything seems normal,
but then after a few minutes, not doing anything special, things get
sluggish and within seconds everything grinds to a complete halt. In the
last screen update I had, there was no double-digit percentage of ksoftirqd
though.
Below is what I get in the logs. Note that it doesn't actually shut down, it
just freezes and a hard reset is needed. The same kernels as non-RT work fine.
Not sure if this is really related, but perhaps it helps.
Cheers,
Alex
Aug 10 17:04:52 linux kernel: [ 189.542192] INFO: rcu_preempt detected
stalls on CPUs/tasks: { 3} (detected by 0, t=5252 jiffies, g=3874, c=3873,
q=8201)
Aug 10 17:04:52 linux kernel: [ 189.542194] Task dump for CPU 3:
Aug 10 17:04:52 linux kernel: [ 189.542200] swapper/3 R running task
0 0 1 0x00200000
Aug 10 17:04:52 linux kernel: [ 189.542206] ffffffff8136e2fa
ffffffff8101f9e5 ffff8803333eee00 0000000008004000
Aug 10 17:04:52 linux kernel: [ 189.542209] 0000000000000004
0000000000000004 ffffffff8145baf4 0000000000000000
Aug 10 17:04:52 linux kernel: [ 189.542212] 0000000000000000
ffff88033338f020 ffff8803333eee00 ffffffff8190b640
Aug 10 17:04:52 linux kernel: [ 189.542213] Call Trace:
Aug 10 17:04:52 linux kernel: [ 189.542224] [<ffffffff8136e2fa>] ?
intel_idle+0x12a/0x140
Aug 10 17:04:52 linux kernel: [ 189.542230] [<ffffffff8101f9e5>] ?
read_tsc+0x5/0x10
Aug 10 17:04:52 linux kernel: [ 189.542238] [<ffffffff8145baf4>] ?
cpuidle_enter_state+0x74/0x270
Aug 10 17:04:52 linux kernel: [ 189.542244] [<ffffffff810b9298>] ?
cpu_startup_entry+0x3a8/0x4e0
Aug 10 17:04:52 linux kernel: [ 189.542249] [<ffffffff8104e0d9>] ?
start_secondary+0x169/0x190
Aug 10 17:04:52 linux kernel: [ 189.542254] rcu_preempt kthread starved for
5251 jiffies!
Aug 10 17:05:01 linux CRON[4531]: (root) CMD (command -v debian-sa1 >
/dev/null && debian-sa1 1 1)
Aug 10 17:05:33 linux kernel: [ 210.562994] INFO: rcu_preempt detected
stalls on CPUs/tasks: { 3} (detected by 1, t=5252 jiffies, g=3875, c=3874,
q=14365)
Aug 10 17:05:33 linux kernel: [ 210.562997] Task dump for CPU 3:
Aug 10 17:05:33 linux kernel: [ 210.563003] swapper/3 R running task
0 0 1 0x00200000
Aug 10 17:05:33 linux kernel: [ 210.563081] ffffffff8136e2fa
ffffffff8101f9e5 ffff8803333eee00 0000000008004000
Aug 10 17:05:33 linux kernel: [ 210.563085] 0000000000000004
0000000000000004 ffffffff8145baf4 0000000000000000
Aug 10 17:05:33 linux kernel: [ 210.563090] 0000000000000000
ffff88033338f020 ffff8803333eee00 ffffffff8190b640
Aug 10 17:05:33 linux kernel: [ 210.563091] Call Trace:
Aug 10 17:05:33 linux kernel: [ 210.563103] [<ffffffff8136e2fa>] ?
intel_idle+0x12a/0x140
Aug 10 17:05:33 linux kernel: [ 210.563111] [<ffffffff8101f9e5>] ?
read_tsc+0x5/0x10
Aug 10 17:05:33 linux kernel: [ 210.563121] [<ffffffff8145baf4>] ?
cpuidle_enter_state+0x74/0x270
Aug 10 17:05:33 linux kernel: [ 210.563133] [<ffffffff810b9298>] ?
cpu_startup_entry+0x3a8/0x4e0
Aug 10 17:05:33 linux kernel: [ 210.563139] [<ffffffff8104e0d9>] ?
start_secondary+0x169/0x190
Aug 10 17:05:33 linux kernel: [ 210.563143] rcu_preempt kthread starved for
5252 jiffies!
Aug 10 17:05:34 linux shutdown[4689]: shutting down for system halt
Aug 10 17:05:34 linux init: Switching to runlevel: 0
Aug 10 17:05:34 linux shutdown[4822]: shutting down for system halt
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
