10-50% CPU used by xorg?

Eric Anholt eric at anholt.net
Wed Feb 27 11:45:24 PST 2008


On Wed, 2008-02-20 at 09:51 +1100, Daniel Kasak wrote:
> On Tue, 2008-02-19 at 14:47 +0100, Ben E. Hard wrote:
> 
> > 113       0.6208  libexa.so                libexa.so                
> 
> EXA on Intel 830 systems is bound to be fantastically slow at the
> moment, particularly without a composite manager. We're waiting for LOTS
> of experimental stuff to land: the batchbuffer branch of the Intel
> driver, TTM, DRI2, etc, etc. Last word was that this wasn't happening
> any time soon, so in the meantime, force X to use XAA:
> 
> Option      "AccelMethod" "xaa"

EXA is bound to be slow on 965 until we land the render fixes (currently
sitting on intel-batchbuffer, but we could probably just cherry-pick
them out since they're orthogonal to the batchbuffer work).  But if
you're seeing issues on 915 or earlier, the only improvement we're
expecting to see is from TTM-backed buffers, which might help if you're
limited by migration (or might not).

If you're seeing performance issues on 915 or earlier, profile now
because there's a good chance that what you're seeing doesn't have a fix
in the pipeline.

-- 
Eric Anholt                             anholt at FreeBSD.org
eric at anholt.net                         eric.anholt at intel.com

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 187 bytes
Desc: This is a digitally signed message part
URL: <http://lists.x.org/archives/xorg/attachments/20080227/947e102e/attachment.pgp>


More information about the xorg mailing list