6.1-STABLE hangs, ddb shows 'acpi_timer_read'?

[Dominic Marks]

On Fri, 20 Oct 2006 11:56:33 +0100
Karl Pielorz <kpielorz at tdx.co.uk> wrote:

> 
> Hi All,
> 
> We've got an HP DL380 server, stacked out with drives running Sendmail. The 
> machine is quite busy (LA rarely below 4 - and it's three 'spindle' sets of 
> RAID drives are always busy). It's probably constantly running 200-300 
> copies of sendmail, plus an assortment of other processes (mostly admin 
> scripts that kind of thing).
> 
> It's got a Xeon 3.2Ghz CPU (HT disabled), w/2Gb of RAM running a generic 
> kernel, w/out IPv6 support (and with DDB/KDB included obviously)
> 
> This machine just 'hangs' every couple of days. I have DDB/KDB compiled in 
> - and if I throw it into DDB I get the following:
> 
> [first couple of lines missed courtesy of screen-dump]
> db> bt
> Tracing pid 38 tid 100027 td 0xc6495180
> acpi_timer_read(c0aba3c0,c09035c0,e6a05bbc,c0662ef3,c0aba3c0) at 
> acpi_timer_read+0x13
> acpi_timer_get_timecout_safe(c0aba3c0) at acpi_timer_get_timecount_safe+0xa
> binuptime(e6a05be8) at binuptime+0x43
> mi_switch(6,c6395900,c639a54,c639500,e6a05c34) at mi_swtich+0x33
> maybe_preempt(c6395900) at maybe_preempt+0xc4
> sched_add(c639500,4,c6495180,c6395900,c637f280) at sched_add+0x27
> setrunqueue(c6395900,4) at setrunqueue+0x63
> intr_even_schedule_thread(c637f280) at intr_event_schedule_thread+0xb5
> intr_exectue_handlers(c638d6e0,e6a05cac,13,46a05cf4,c08189f3) at 
> intr_execute_handlers+0x118
> ...
> 
> I have a crash dump from it - which I've saved (I'm moderately familiar 
> with working with dumps, but this one is split into two?)
> 
> If anyone has any pointers, or can do some hand holding to get more info 
> from the dump, or what to do next time it happens?

Have you tried changing kern.timecounter.hardware to something
else? TSC? 

Dominic