[Pixman] [PATCH] sse2: faster bilinear interpolation (get rid of XOR instruction)

2013-01-27 Thread Siarhei Siamashka
The old code was calculating horizontal weights for right pixels in the following way (for simplicity assume 8-bit interpolation precision): Start with "x = vx" and do increment "x += ux" after each pixel. In this case right pixel weight for interpolation can be calculated as "((x >> 8) ^ 0x

[Pixman] [PATCH] Change default GPGKEY to 3892336E, which is soren.sandm...@gmail.com

2013-01-27 Thread Søren Sandmann
From: Søren Sandmann Pedersen The old one belongs to sandm...@daimi.au.dk which doesn't work anyore. Also use gpg to get the name and address for the "(Signed by ...)" line since that works more reliably for me than using git. --- Makefile.am | 4 ++-- 1 file changed, 2 insertions(+), 2 deletio

Re: [Pixman] 0.29.2

2013-01-27 Thread Søren Sandmann
Siarhei Siamashka writes: >> It's about time to get a 0.29.2 development snapshot out, > > Hi, that's a really good idea. Any preliminary ETA? Or are we going to > know that this development snapshot is out only after you create the > tag and post an announcement? Thanks for pushing the patches.

Re: [Pixman] 0.29.2

2013-01-27 Thread Matt Turner
On Sun, Jan 27, 2013 at 11:43 AM, Siarhei Siamashka wrote: > Still, I'm not very happy about the code duplication. We already have > similar iterators (fetch only, no writeback) in "pixman-mmx.c": > > > http://cgit.freedesktop.org/pixman/tree/pixman/pixman-mmx.c?id=pixman-0.28.2#n3904 > > Ide

Re: [Pixman] 0.29.2

2013-01-27 Thread Jeff Muizelaar
On 2013-01-27, at 2:43 PM, Siarhei Siamashka wrote: > I was a bit worried about the projective transforms. The accuracy is > improved, but the new code uses a rather naive and slow implementation > for long division. So the performance is going to be worse. But as > almost nobody seems to be usin

Re: [Pixman] 0.29.2

2013-01-27 Thread Siarhei Siamashka
On Sat, 19 Jan 2013 01:15:51 +0100 sandm...@cs.au.dk (Søren Sandmann) wrote: > Hi, > > It's about time to get a 0.29.2 development snapshot out, Hi, that's a really good idea. Any preliminary ETA? Or are we going to know that this development snapshot is out only after you create the tag and pos

Re: [Pixman] [PATCH 2/4] general: Fetch the source directly into the destination

2013-01-27 Thread Chris Wilson
On Sun, Jan 27, 2013 at 05:04:37PM +0100, Søren Sandmann wrote: > Chris Wilson writes: > > > This is principally used by external renderers to convert a procedural > > pixman_image_t (SourcePict) into an image that they can handle, using a > > simple pixman_image_composite(SRC, source, NULL, dst)

Re: [Pixman] [PATCH 2/4] general: Fetch the source directly into the destination

2013-01-27 Thread Søren Sandmann
Chris Wilson writes: > This is principally used by external renderers to convert a procedural > pixman_image_t (SourcePict) into an image that they can handle, using a > simple pixman_image_composite(SRC, source, NULL, dst). For these cases > we typicall hit the general_composite_rect and so retr

[Pixman] [PATCH 3/4] sse2: affine bilinear fetcher

2013-01-27 Thread Chris Wilson
On an SNB i5-2500 using cairo-image: firefox-canvas17.8 -> 10.3: 1.72x speedup firefox-tron 46.3 -> 28.4: 1.63x speedup swfdec-youtube 1.7 -> 1.4: 1.22x speedup firefox-fishbowl 64.6 -> 53.7: 1.20x speedup firefox-paintball 40.8 -> 36.8: 1.11x speedup firefo

[Pixman] [PATCH 4/4] sse2: Implement simple bilinear scaling for x8r8g8b8 to a8r8g8b8

2013-01-27 Thread Chris Wilson
Once upon a time this improved firefon-tron on a IVB i7-3720qm from 68.6s to 45.2s. However, we have accomplished the same goal with earlier tuning... Signed-off-by: Chris Wilson --- pixman/pixman-sse2.c | 113 -- 1 file changed, 110 insertions(+)

[Pixman] [PATCH 2/4] general: Fetch the source directly into the destination

2013-01-27 Thread Chris Wilson
This is principally used by external renderers to convert a procedural pixman_image_t (SourcePict) into an image that they can handle, using a simple pixman_image_composite(SRC, source, NULL, dst). For these cases we typicall hit the general_composite_rect and so retrieve the source into a temporar

[Pixman] [PATCH 1/4] bits: Implement PAD support in the simple fetcher

2013-01-27 Thread Chris Wilson
SNB i5-2500s: firefox-chalkboard 25.9s -> 19.6s: 1.32x speedup --- pixman/pixman-bits-image.c | 66 +++- 1 file changed, 65 insertions(+), 1 deletion(-) diff --git a/pixman/pixman-bits-image.c b/pixman/pixman-bits-image.c index 75a39a1..f25d576 100644 --