Adam Jackson writes:
> Formerly we'd zero the image data and then pull out a plane at a time.
> It's faster to apply the planemask after the fact, since that turns the
> GetImage into a memcpy:
>
> 10.0 101000.0 (1.010) (copy 0x) ShmGetImage 10x10 square
>
Adam Jackson writes:
> I have a plan to deal with 24bpp, very harshly indeed.
You know how much time I spent writing that code. You're deleting some
of the "finest" bitblt in the server.
> You do appear to be correct though, xfree86 still has paths by which
> you can ask for a
On Wed, 2016-09-28 at 10:19 -0700, Keith Packard wrote:
> You'll need to deal with 24bpp separately...
I have a plan to deal with 24bpp, very harshly indeed.
You do appear to be correct though, xfree86 still has paths by which
you can ask for a 24bpp pixmap/wire format, so I can't assume we'll
Adam Jackson writes:
> diff --git a/fb/fbimage.c b/fb/fbimage.c
> index 59daa21..8f5f3dc 100644
> --- a/fb/fbimage.c
> +++ b/fb/fbimage.c
> @@ -250,13 +250,16 @@ fbGetImage(DrawablePtr pDrawable,
>
> pm = fbReplicatePixel(planeMask, srcBpp);
> dstStride =
Formerly we'd zero the image data and then pull out a plane at a time.
It's faster to apply the planemask after the fact, since that turns the
GetImage into a memcpy:
10.0 101000.0 (1.010) (copy 0x) ShmGetImage 10x10 square
42400.0 59400.0 (1.401) (copy 0x) ShmGetImage