[lkml]   [2008]   [Feb]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [patch] block layer: kmemcheck fixes
On Fri, Feb 08, 2008 at 02:56:09PM -0800, Arjan van de Ven wrote:
> Nick Piggin wrote:
> >>>Maybe cpus these days have so much store bandwith that doing
> >>>things like the above is OK, but I doubt it :-)
> >>on modern x86 cpus the memset may even be faster if the memory isn't in
> >>cache;
> >>the "explicit" method ends up doing Write Allocate on the cache lines
> >>(so read them from memory) even though they then end up being written
> >>entirely.
> >>With memset the CPU is told that the entire range is set to a new value,
> >>and
> >>the WA can be avoided for the whole-cachelines in the range.
> >
> >Don't you have write combining store buffers? Or is it still speculatively
> >issuing the reads even before the whole cacheline is combined?
> x86 memory order model doesn't allow that quite; and you need a "series" of
> at least 64 bytes
> without any other memory accesses in between even if it would....
> not happening in practice.

OK, fair enough... then it will be a very nice test to see if it
helps. I'm sure you could have an arch specific initialisation
function if it makes a significant difference.

 \ /
  Last update: 2008-02-09 01:01    [W:0.070 / U:7.128 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site