Re: Hard lockup with timer events

From: Frederic Weisbecker
Date: Wed Nov 11 2009 - 11:34:45 EST


On Wed, Nov 11, 2009 at 02:23:01PM +0100, Peter Zijlstra wrote:
> On Wed, 2009-11-11 at 05:50 +0100, Frederic Weisbecker wrote:
> > On Thu, Oct 29, 2009 at 08:15:03PM +0100, Soeren Sandmann wrote:
> > > Hi,
> > >
> > > This program
> > >
> > > http://www.daimi.au.dk/~sandmann/lockup.c
> > >
> > > locks up hard for me on a Pentium 4 in 32 bit mode, with the hrtimer
> > > patches applied (-tip for example).
> > >
> > > What the program does is spawn a thread that just spins, and then 400
> > > threads that each create a performance counter of type
> > > PERF_COUNT_SW_CPU_CLOCK and a sample period of 1000000 and a
> > > sample_type of
> > >
> > > PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_CALLCHAIN
> > >
> > > These threads then map the event buffer and read it continuously.
> > >
> > > It does not apparently lock up if SAMPLE_CALLCHAIN is removed; also I
> > > cannot get it to lock up on Core 2 Duo in either 32 or 64 bit mode.
> > >
> > >
> > > Soren
> >
> >
> >
> > It locks up hard for me too, no more blinking cursor, nothing.
> > I've attached my config, it's against tip:master from yesterday.
>
> What kind of cpu does your laptop have? The .config seems to suggest a
> Core2. Which would make Soeren's report of not being able to trash a
> 64bit core2 'interesting' ;-)


Nop it's a Pentium Dual Core :)

processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Pentium(R) Dual CPU T2310 @ 1.46GHz
stepping : 13
cpu MHz : 800.000
cache size : 1024 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
apicid : 1
initial apicid : 1
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr
sse sse2 ss ht tm pbe syscall nx lm constant_tsc arch_perfmon pebs bts rep_good aperfmperf pni dtes64 monitor ds_cpl
est tm2 ssse3 cx16 xtpr pdcm lahf_lm
bogomips : 2925.95
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:

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