head	1.2;
access;
symbols
	pkgsrc-2013Q2:1.2.0.10
	pkgsrc-2013Q2-base:1.2
	pkgsrc-2012Q4:1.2.0.8
	pkgsrc-2012Q4-base:1.2
	pkgsrc-2011Q4:1.2.0.6
	pkgsrc-2011Q4-base:1.2
	pkgsrc-2011Q2:1.2.0.4
	pkgsrc-2011Q2-base:1.2
	pkgsrc-2009Q4:1.2.0.2
	pkgsrc-2009Q4-base:1.2
	pkgsrc-2009Q1:1.1.0.4
	pkgsrc-2009Q1-base:1.1
	pkgsrc-2008Q4:1.1.0.2
	pkgsrc-2008Q4-base:1.1;
locks; strict;
comment	@# @;


1.2
date	2009.04.23.08.30.53;	author sno;	state dead;
branches;
next	1.1;

1.1
date	2008.10.15.00.56.45;	author bjs;	state Exp;
branches;
next	;


desc
@@


1.2
log
@Update to pixman-0.15.2 from 0.12.0

Upstream changes (since 0.12.0):
  - Fix MMX & SSE intrinsics to work with Sun compilers & Solaris
  - More SSE2 optimizations
  - Optimize rectilinear nearest-neighbour scaling
  - Add support for BGRA and BGRx formats
  and a lot of minor bug fixes and optimizations

Oked by wiz@@
@
text
@$NetBSD: patch-ah,v 1.1 2008/10/15 00:56:45 bjs Exp $


1c5de7b05831b3d66821707276b71974a232f5c7

Move _mm_empty() to correct place

The "fbComposeSetupSSE2()" function is guarding most of its code
depending on the capabilities of the CPU, but unfortunately the call
to "_mm_empty()" is not part of this code path but executed
unconditionally.  This results in a "illegal instruction" crash on
non-MMX / non-SSE capable CPUs caused by the the "emms" instruction
(embedded in "_mm_empty()").

Fix bug 17729.

--- pixman/pixman-sse2.c.orig	2008-09-17 09:50:49.000000000 -0400
+++ pixman/pixman-sse2.c
@@@@ -2357,11 +2357,11 @@@@ fbComposeSetupSSE2(void)
         pixman_composeFunctions.combineC[PIXMAN_OP_ADD] = sse2CombineAddC;
 
         pixman_composeFunctions.combineMaskU = sse2CombineMaskU;
+
+	_mm_empty();
     }
 
     initialized = TRUE;
-
-    _mm_empty();
 }
 
 
@


1.1
log
@Update to pixman-0.12.0 from 0.11.10.

In addition to many bug fixes, new features [since approximately 0.10.0]
include:

 	- New image formats with 10 bits per channel
 	- SSE2 optimizations for many operations
 	- Opacity based optimizations for operators
 	- Altivec optimizations for many operations
 	- A 32 bit region implementation
@
text
@d1 1
a1 1
$NetBSD$
@