lkml.org 
[lkml]   [2004]   [Aug]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [patch] voluntary-preempt-2.6.8.1-P1
From
Date
Ingo Molnar wrote :
> * Lee Revell <rlrevell@joe-job.com> wrote:
>
> > Anyway, the change to sched.c fixes the mlockall bug, it works
> > perfectly now. Thanks!
>
> great! This fix also means that we've got one more lock-break in the
> ext3 journalling code and one more lock-break in dcache.c. I've released
> -P1 with the fix included:
>
> http://redhat.com/~mingo/voluntary-preempt/voluntary-preempt-2.6.8.1-P1
>
> Ingo

Tested P1. The biggest offender on my system is still reiserfs'
search_by_key. The trace is made of a very long loop of :
0.016ms (+0.000ms): reiserfs_in_journal (scan_bitmap_block)
0.016ms (+0.000ms): find_next_zero_bit (scan_bitmap_block)
(...)
0.977ms (+0.000ms): reiserfs_in_journal (scan_bitmap_block)
0.977ms (+0.000ms): find_next_zero_bit (scan_bitmap_block)
with interrupts showing up in the trace from time to time.
Do you have plans to fix this, or should I switch to ext3 ?

Another point was the ACPI induced latency I was experiencing. I think
that P1 could help track this down. After a fresh boot on an ACPI
enabled kernel, once the initialization-related latency spikes are over,
I reset /proc/sys/kernel/preempt_max_latency to 100, and then I get a
100us latency_trace attributed to swapper/1:

preemption latency trace v1.0
-----------------------------
latency: 100 us, entries: 4000 (14253)
process: swapper/1, uid: 0
nice: 0, policy: 0, rt_priority: 0
=======>
0.000ms (+0.000ms): startup_8259A_irq (probe_irq_on)
0.000ms (+0.000ms): enable_8259A_irq (startup_8259A_irq)
0.001ms (+0.001ms): startup_8259A_irq (probe_irq_on)
0.001ms (+0.000ms): enable_8259A_irq (startup_8259A_irq)
0.002ms (+0.001ms): startup_8259A_irq (probe_irq_on)
0.002ms (+0.000ms): enable_8259A_irq (startup_8259A_irq)
(...)

but the weird thing is that the latency sum goes way above those 100us,
the culprit being do_IRQ, regularly chewing up to 1ms !

(...)
0.009ms (+0.000ms): enable_8259A_irq (startup_8259A_irq)
0.011ms (+0.001ms): startup_8259A_irq (probe_irq_on)
0.011ms (+0.000ms): enable_8259A_irq (startup_8259A_irq)
0.954ms (+0.943ms): do_IRQ (common_interrupt)
-------------^
0.954ms (+0.000ms): mask_and_ack_8259A (do_IRQ)
0.956ms (+0.002ms): generic_redirect_hardirq (do_IRQ)
0.956ms (+0.000ms): generic_handle_IRQ_event (do_IRQ)
0.957ms (+0.000ms): timer_interrupt (generic_handle_IRQ_event)
0.957ms (+0.000ms): mark_offset_tsc (timer_interrupt)
0.962ms (+0.005ms): do_timer (timer_interrupt)
(...)
119.027ms (+0.000ms): preempt_schedule (timer_interrupt)
119.027ms (+0.000ms): preempt_schedule (timer_interrupt)
119.027ms (+0.000ms): generic_note_interrupt (do_IRQ)
119.027ms (+0.000ms): end_8259A_irq (do_IRQ)
119.028ms (+0.000ms): enable_8259A_irq (do_IRQ)
119.028ms (+0.000ms): preempt_schedule (do_IRQ)
119.029ms (+0.000ms): preempt_schedule (do_IRQ)
119.029ms (+0.000ms): do_softirq (do_IRQ)
119.029ms (+0.000ms): __do_softirq (do_softirq)
120.017ms (+0.987ms): do_IRQ (common_interrupt)
---------------^
120.017ms (+0.000ms): mask_and_ack_8259A (do_IRQ)
120.019ms (+0.002ms): preempt_schedule (do_IRQ)
120.019ms (+0.000ms): generic_redirect_hardirq (do_IRQ)
120.019ms (+0.000ms): preempt_schedule (do_IRQ)
120.020ms (+0.000ms): generic_handle_IRQ_event (do_IRQ)
(...)

You can find the full trace here :
http://www.undata.org/~thomas/swapper.trace

Do you have an idea of what's happening here ?

Thomas


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 14:05    [W:0.496 / U:0.456 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site