Re: gettimeofday order of magnitude slower with pmtimer, which is default

[Con Kolivas <kernel@kolivas.org>]

On Tue, 21 Mar 2006 01:59 pm, OGAWA Hirofumi wrote:
> Yes. However, if machines uses buggy chip, I guessed TSC/PIT would be
> more proper as time source. 

Oh yes but there has been an epidemic of timer problems (fast/slow, lost ticks 
etc) lately meaning the pm timer is being relied upon more and more.

> But probably you are right, timer_pit.c 
> seems more slow usually (it uses many I/O port).
>
> I'll remove unlikely(), and also will remove "Use other timer source"
> from warning.

Suggesting another timer source is ok in the warning I believe given massive 
amounts of wasted cpu.

> BTW, this patch is still quick hack.

Understood. Perhaps having an indirect function call set to either 
good_pmtmr() or bad_pmtmr() after checking would be preferable to a variable 
that is checked on each function call despite never changing.

> At least, we would need to check the ICH4 which says in comment.
> However, I couldn't find the PM-Timer Errata in ICH4 spec update.
>
> Do you/anyone know about a ICH4 error?

Not personally but my ICH4 pm timer seems to work very well whereas Andi's 
apparently similar chipset exhibits terrible problems.

Cheers,
Con