[Intel-gfx] Too many memory managers

[Keith Packard]

On Wed, 2008-08-06 at 13:44 +0200, Lukas Hejtmanek wrote:

> wouldn't the pixmap creation be too expensive compared to traditional
> allocation?

A bit of user-mode reuse solves any performance issues.

> is the read penalty still noticable if you read doublequadwords using sse 
> and also explicitly prefetching data?

Yes. Uncached reads must always go to memory for each transaction. Just
compare the speed of x11perf -getimage500 with x11perf -shmput500 to see
how read and write performance differs. (a quick check with sysprof
showed getimage spending >90% of its time in copying data from the frame
buffer).

-- 
keith.packard at intel.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20080806/7649356c/attachment.sig>