Re: oprofile: WQ on CPU0, prefer CPU1

[Andrew Morton <akpm@linux-foundation.org>]

(cc's added, from MAINTAINERS)

On Thu, 14 May 2009 20:22:14 +0200 Thomas Meyer <thomas@m3y3r.de> wrote:

> 
> After suspend and resume to/from S3 I see a lot of these messages in  
> the kernel log:
> 
> [11750.054031] WQ on CPU0, prefer CPU1
> [11750.154023] WQ on CPU0, prefer CPU1
> [11750.254022] WQ on CPU0, prefer CPU1
> [11750.354022] WQ on CPU0, prefer CPU1
> [11750.454018] WQ on CPU0, prefer CPU1
> [11750.554022] WQ on CPU0, prefer CPU1
> [11750.654022] WQ on CPU0, prefer CPU1
> [11750.754039] WQ on CPU0, prefer CPU1
> [11750.854020] WQ on CPU0, prefer CPU1
> [11750.954024] WQ on CPU0, prefer CPU1
> [11751.054029] WQ on CPU0, prefer CPU1
> [11751.154028] WQ on CPU0, prefer CPU1
> [11751.254026] WQ on CPU0, prefer CPU1
> [11751.354019] WQ on CPU0, prefer CPU1
> [11751.454022] WQ on CPU0, prefer CPU1
> [11751.554032] WQ on CPU0, prefer CPU1
> [11751.654028] WQ on CPU0, prefer CPU1
> [11751.754028] WQ on CPU0, prefer CPU1
> [11751.854033] WQ on CPU0, prefer CPU1
> [11751.954029] WQ on CPU0, prefer CPU1
> [and so on; seems not to stop]
> 
> This comes from drivers/oprofile/cpu_buffer.c
> 
> Bug or feature?
> 

Don't know.  It needs fixing though - the kernel shouldn't blurt random
printks like this.

> 
> PS: 2.6.30-rc5

Do you know if earlier kernels did the same thing?