Re: mmotm 2009-04-10-02-21 uploaded - forkbombed by work_for_cpu

[Valdis . Kletnieks]

On Mon, 13 Apr 2009 10:27:49 PDT, Andrew Morton said:

> From: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> 
> Atttempting to rid us of the problematic work_on_cpu().  Just use
> smp_call_fuction_single() here.
> 
> This repairs a 10% sysbench(oltp)+mysql regression which Mike reported,
> due to 
> 
> commit 6b44003e5ca66a3fffeb5bc90f40ada2c4340896
> Author: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Date:   Thu Apr 9 09:50:37 2009 -0600
> 
>     work_on_cpu(): rewrite it to create a kernel thread on demand
> 
> It seems that the kernel calls these acpi-cpufreq functions at a quite
> high frequency.

I put the one reverted commit back on (making it back into what was stock
mmotm-0410 that had the problem) and then applied this patch.  I'm seeing
near-zero forks/second again,  the CPU is at 1Ghz when idling (as it should
be), if I kick off 2 'for(;;)' cycle-suckers it goes to 2Ghz,  And it even
puts 1 CPU at 1Ghz and the other at 2Ghz if there's only one running - I didn't
know a Core2 Duo could do that. ;)

So life is good again. ;)

Attachment: pgp00000.pgp
Description: PGP signature