Merge remote-tracking branch 'origin/released'

* origin/released:
  fontconfig mesa pixman git update 11 Sep 2013

Conflicts:
	pixman/pixman/pixman-sse2.c

17 files changed, 76 insertions, 141 deletions

diff --git a/fontconfig/doc/fontconfig-devel.sgml b/fontconfig/doc/fontconfig-devel.sgml
index a7b853021..7306d7924 100644
--- a/fontconfig/doc/fontconfig-devel.sgml
+++ b/fontconfig/doc/fontconfig-devel.sgml
@@ -170,7 +170,7 @@ convenience for the application's rendering mechanism.
                                                   the file
     ftface         FC_FT_FACE             FT_Face Use the specified FreeType
                                                   face object
-    rasterizer     FC_RASTERIZER          String  Which rasterizer is in use
+    rasterizer     FC_RASTERIZER          String  Which rasterizer is in use (deprecated)
     outline        FC_OUTLINE             Bool    Whether the glyphs are outlines
     scalable       FC_SCALABLE            Bool    Whether glyphs can be scaled
     scale          FC_SCALE               Double  Scale factor for point->pixel
diff --git a/fontconfig/doc/fontconfig-user.sgml b/fontconfig/doc/fontconfig-user.sgml
index eeff69add..775d1f725 100644
--- a/fontconfig/doc/fontconfig-user.sgml
+++ b/fontconfig/doc/fontconfig-user.sgml
@@ -116,7 +116,7 @@ convenience for the applications' rendering mechanism.
   file            String  The filename holding the font
   index           Int     The index of the font within the file
   ftface          FT_Face Use the specified FreeType face object
-  rasterizer      String  Which rasterizer is in use
+  rasterizer      String  Which rasterizer is in use (deprecated)
   outline         Bool    Whether the glyphs are outlines
   scalable        Bool    Whether glyphs can be scaled
   scale           Double  Scale factor for point->pixel conversions
diff --git a/fontconfig/fontconfig/fontconfig.h b/fontconfig/fontconfig/fontconfig.h
index 079177d4c..58912f572 100644
--- a/fontconfig/fontconfig/fontconfig.h
+++ b/fontconfig/fontconfig/fontconfig.h
@@ -91,7 +91,7 @@ typedef int		FcBool;
 #define FC_FILE		    "file"		/* String */
 #define FC_INDEX	    "index"		/* Int */
 #define FC_FT_FACE	    "ftface"		/* FT_Face */
-#define FC_RASTERIZER	    "rasterizer"	/* String */
+#define FC_RASTERIZER	    "rasterizer"	/* String (deprecated) */
 #define FC_OUTLINE	    "outline"		/* Bool */
 #define FC_SCALABLE	    "scalable"		/* Bool */
 #define FC_SCALE	    "scale"		/* double */
diff --git a/mesalib/docs/autoconf.html b/mesalib/docs/autoconf.html
index a07a3ee14..d4e8a35d8 100644
--- a/mesalib/docs/autoconf.html
+++ b/mesalib/docs/autoconf.html
@@ -123,24 +123,6 @@ directories.</p>
 There are also a few general options for altering the Mesa build:
 </p>
 <dl>
-<dt><code>--with-x</code></dt>
-<dd><p>When the X11 development libraries are
-needed, the <code>pkg-config</code> utility <a href="#pkg-config">will
-be used</a> for locating them. If they cannot be found through
-<code>pkg-config</code> a fallback routing using <code>imake</code> will
-be used. In this case, the <code>--with-x</code>,
-<code>--x-includes</code> and <code>--x-libraries</code> options can
-control the use of X for Mesa.</p>
-</dd>
-
-<dt><code>--enable-gl-osmesa</code></dt>
-<dd><p>The <a href="osmesa.html">OSMesa
-library</a> can be built on top of libGL for drivers that provide it.
-This option controls whether to build libOSMesa. By default, this is
-enabled for the Xlib driver and disabled otherwise. Note that this
-option is different than using OSMesa as the driver.</p>
-</dd>
-
 <dt><code>--enable-debug</code></dt>
 <dd><p>This option will enable compiler
 options and macros to aid in debugging the Mesa libraries.</p>
@@ -155,12 +137,12 @@ assembly will not be used.</p>
 
 <dt><code>--enable-32-bit</code></dt>
 <dt><code>--enable-64-bit</code></dt>
-<dd><p>By default, the
-build will compile code as directed by the environment variables
+<dd><p>By default, the build will compile code as directed by the environment
+variables
 <code>CC</code>, <code>CFLAGS</code>, etc. If the compiler is
 <code>gcc</code>, these options offer a helper to add the compiler flags
 to force 32- or 64-bit code generation as used on the x86 and x86_64
-architectures.</p>
+architectures. Note that these options are mutually exclusive.</p>
 </dd>
 </dl>
 
@@ -171,19 +153,19 @@ architectures.</p>
 There are several different driver modes that Mesa can use. These are
 described in more detail in the <a href="install.html">basic
 installation instructions</a>. The Mesa driver is controlled through the
-configure option --with-driver. There are currently three supported
-options in the configure script.
+configure options <code>--enable-xlib-glx</code>, <code>--enable-osmesa</code>,
+and <code>--enable-dri</code>.
 </p>
 
-<h3 id="xlib">Xlib</h3><p>This is the default mode for building Mesa.
+<h3 id="xlib">Xlib</h3><p>
 It uses Xlib as a software renderer to do all rendering. It corresponds
-to the option <code>--with-driver=xlib</code>. The libX11 and libXext
+to the option <code>--enable-xlib-glx</code>. The libX11 and libXext
 libraries, as well as the X11 development headers, will be need to
 support the Xlib driver.
 
 <h3 id="dri">DRI</h3><p>This mode uses the DRI hardware drivers for
 accelerated OpenGL rendering. Enable the DRI drivers with the option
-<code>--with-driver=dri</code>. See the <a href="install.html">basic
+<code>--enable-dri</code>. See the <a href="install.html">basic
 installation instructions</a> for details on prerequisites for the DRI
 drivers.
 
@@ -223,7 +205,8 @@ and <code>/usr/local/lib</code>, respectively.
 <h3 id="osmesa">OSMesa </h3><p> No libGL is built in this
 mode. Instead, the driver code is built into the Off-Screen Mesa
 (OSMesa) library. See the <a href="osmesa.html">Off-Screen Rendering</a>
-page for more details.
+page for more details.  It corresponds to the option
+<code>--enable-osmesa</code>.
 
 <!-- OSMesa specific options -->
 <dl>
diff --git a/mesalib/src/gallium/auxiliary/util/u_cpu_detect.c b/mesalib/src/gallium/auxiliary/util/u_cpu_detect.c
index 2ff40bb00..d2d1313f8 100644
--- a/mesalib/src/gallium/auxiliary/util/u_cpu_detect.c
+++ b/mesalib/src/gallium/auxiliary/util/u_cpu_detect.c
@@ -283,8 +283,11 @@ static INLINE boolean sse2_has_daz(void)
    fxarea.mxcsr_mask = 0;
 #if (defined(PIPE_CC_GCC) || defined(PIPE_CC_SUNPRO))
    __asm __volatile ("fxsave %0" : "+m" (fxarea));
-#elif (defined(PIPE_CC_MSVC) || defined(PIPE_CC_ICL))
+#elif (defined(PIPE_CC_MSVC) && _MSC_VER >= 1700) || defined(PIPE_CC_ICL)
+   /* 1700 = Visual Studio 2012 */
    _fxsave(&fxarea);
+#else
+   fxarea.mxcsr_mask = 0;
 #endif
    return !!(fxarea.mxcsr_mask & (1 << 6));
 }
diff --git a/mesalib/src/glsl/builtin_functions.cpp b/mesalib/src/glsl/builtin_functions.cpp
index 5d8f171b9..ce78df12e 100644
--- a/mesalib/src/glsl/builtin_functions.cpp
+++ b/mesalib/src/glsl/builtin_functions.cpp
@@ -525,6 +525,9 @@ private:
  *  @{
  */
 builtin_builder::builtin_builder()
+   : shader(NULL),
+     gl_ModelViewProjectionMatrix(NULL),
+     gl_Vertex(NULL)
 {
    mem_ctx = NULL;
 }
@@ -1850,6 +1853,7 @@ builtin_builder::add_function(const char *name, ...)
 
       f->add_signature(sig);
    }
+   va_end(ap);
 
    shader->symbols->add_function(f);
 }
diff --git a/mesalib/src/mesa/main/errors.c b/mesalib/src/mesa/main/errors.c
index e1a9fe2f5..28357e0e8 100644
--- a/mesalib/src/mesa/main/errors.c
+++ b/mesalib/src/mesa/main/errors.c
@@ -856,7 +856,7 @@ _mesa_PushDebugGroup(GLenum source, GLuint id, GLsizei length,
       length = strlen(message);
    emptySlot = &ctx->Debug.DebugGroupMsgs[ctx->Debug.GroupStackDepth];
    store_message_details(emptySlot, gl_enum_to_debug_source(source),
-                         gl_enum_to_debug_source(GL_DEBUG_TYPE_PUSH_GROUP),
+                         gl_enum_to_debug_type(GL_DEBUG_TYPE_PUSH_GROUP),
                          id,
                    gl_enum_to_debug_severity(GL_DEBUG_SEVERITY_NOTIFICATION),
                          length, message);
diff --git a/mesalib/src/mesa/main/syncobj.c b/mesalib/src/mesa/main/syncobj.c
index f9736938a..3157491bf 100644
--- a/mesalib/src/mesa/main/syncobj.c
+++ b/mesalib/src/mesa/main/syncobj.c
@@ -71,7 +71,7 @@
 static struct gl_sync_object *
 _mesa_new_sync_object(struct gl_context *ctx, GLenum type)
 {
-   struct gl_sync_object *s = MALLOC_STRUCT(gl_sync_object);
+   struct gl_sync_object *s = CALLOC_STRUCT(gl_sync_object);
    (void) ctx;
    (void) type;
 
diff --git a/mesalib/src/mesa/program/prog_statevars.c b/mesalib/src/mesa/program/prog_statevars.c
index f6073be13..145c07c67 100644
--- a/mesalib/src/mesa/program/prog_statevars.c
+++ b/mesalib/src/mesa/program/prog_statevars.c
@@ -256,8 +256,7 @@ _mesa_fetch_state(struct gl_context *ctx, const gl_state_index state[],
       value[0] = ctx->Fog.Density;
       value[1] = ctx->Fog.Start;
       value[2] = ctx->Fog.End;
-      value[3] = (ctx->Fog.End == ctx->Fog.Start)
-         ? 1.0f : (GLfloat)(1.0 / (ctx->Fog.End - ctx->Fog.Start));
+      value[3] = 1.0f / (ctx->Fog.End - ctx->Fog.Start);
       return;
    case STATE_CLIPPLANE:
       {
diff --git a/mesalib/src/mesa/state_tracker/st_manager.c b/mesalib/src/mesa/state_tracker/st_manager.c
index 098e6c02c..b1fd91a0e 100644
--- a/mesalib/src/mesa/state_tracker/st_manager.c
+++ b/mesalib/src/mesa/state_tracker/st_manager.c
@@ -189,7 +189,7 @@ st_framebuffer_validate(struct st_framebuffer *stfb,
 
    /* validate the fb */
    do {
-      if (!stfb->iface->validate(stfb->iface, stfb->statts,
+      if (!stfb->iface->validate(&st->iface, stfb->iface, stfb->statts,
 				 stfb->num_statts, textures))
 	 return;
 
diff --git a/pixman/pixman/pixman-access.c b/pixman/pixman/pixman-access.c
index b5c8e4017..4f0642d77 100644
--- a/pixman/pixman/pixman-access.c
+++ b/pixman/pixman/pixman-access.c
@@ -294,14 +294,14 @@ convert_pixel (pixman_format_code_t from, pixman_format_code_t to, uint32_t pixe
 }
 
 static force_inline uint32_t
-convert_pixel_to_a8r8g8b8 (pixman_image_t *image,
+convert_pixel_to_a8r8g8b8 (bits_image_t *image,
 			   pixman_format_code_t format,
 			   uint32_t pixel)
 {
     if (PIXMAN_FORMAT_TYPE (format) == PIXMAN_TYPE_GRAY		||
 	PIXMAN_FORMAT_TYPE (format) == PIXMAN_TYPE_COLOR)
     {
-	return image->bits.indexed->rgba[pixel];
+	return image->indexed->rgba[pixel];
     }
     else
     {
@@ -332,7 +332,7 @@ convert_pixel_from_a8r8g8b8 (pixman_image_t *image,
 }
 
 static force_inline uint32_t
-fetch_and_convert_pixel (pixman_image_t	*	image,
+fetch_and_convert_pixel (bits_image_t *		image,
 			 const uint8_t *	bits,
 			 int			offset,
 			 pixman_format_code_t	format)
@@ -417,7 +417,7 @@ convert_and_store_pixel (bits_image_t *		image,
 
 #define MAKE_ACCESSORS(format)						\
     static void								\
-    fetch_scanline_ ## format (pixman_image_t *image,			\
+    fetch_scanline_ ## format (bits_image_t *image,			\
 			       int	       x,			\
 			       int             y,			\
 			       int             width,			\
@@ -425,7 +425,7 @@ convert_and_store_pixel (bits_image_t *		image,
 			       const uint32_t *mask)			\
     {									\
 	uint8_t *bits =							\
-	    (uint8_t *)(image->bits.bits + y * image->bits.rowstride);	\
+	    (uint8_t *)(image->bits + y * image->rowstride);		\
 	int i;								\
 									\
 	for (i = 0; i < width; ++i)					\
@@ -461,8 +461,8 @@ convert_and_store_pixel (bits_image_t *		image,
 	uint8_t *bits =							\
 	    (uint8_t *)(image->bits + line * image->rowstride);		\
 									\
-	return fetch_and_convert_pixel ((pixman_image_t *)image,	\
-					bits, offset, PIXMAN_ ## format); \
+	return fetch_and_convert_pixel (				\
+	    image, bits, offset, PIXMAN_ ## format);			\
     }									\
 									\
     static const void *const __dummy__ ## format
@@ -583,14 +583,14 @@ to_srgb (float f)
 }
 
 static void
-fetch_scanline_a8r8g8b8_sRGB_float (pixman_image_t *image,
+fetch_scanline_a8r8g8b8_sRGB_float (bits_image_t *  image,
 				    int             x,
 				    int             y,
 				    int             width,
 				    uint32_t *      b,
 				    const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = bits + x;
     const uint32_t *end = pixel + width;
     argb_t *buffer = (argb_t *)b;
@@ -612,14 +612,14 @@ fetch_scanline_a8r8g8b8_sRGB_float (pixman_image_t *image,
 
 /* Expects a float buffer */
 static void
-fetch_scanline_a2r10g10b10_float (pixman_image_t *image,
+fetch_scanline_a2r10g10b10_float (bits_image_t *  image,
 				  int             x,
 				  int             y,
 				  int             width,
 				  uint32_t *      b,
 				  const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = bits + x;
     const uint32_t *end = pixel + width;
     argb_t *buffer = (argb_t *)b;
@@ -643,14 +643,14 @@ fetch_scanline_a2r10g10b10_float (pixman_image_t *image,
 
 /* Expects a float buffer */
 static void
-fetch_scanline_x2r10g10b10_float (pixman_image_t *image,
+fetch_scanline_x2r10g10b10_float (bits_image_t   *image,
 				  int             x,
 				  int             y,
 				  int             width,
 				  uint32_t *      b,
 				  const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = (uint32_t *)bits + x;
     const uint32_t *end = pixel + width;
     argb_t *buffer = (argb_t *)b;
@@ -673,14 +673,14 @@ fetch_scanline_x2r10g10b10_float (pixman_image_t *image,
 
 /* Expects a float buffer */
 static void
-fetch_scanline_a2b10g10r10_float (pixman_image_t *image,
+fetch_scanline_a2b10g10r10_float (bits_image_t   *image,
 				  int             x,
 				  int             y,
 				  int             width,
 				  uint32_t *      b,
 				  const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = bits + x;
     const uint32_t *end = pixel + width;
     argb_t *buffer = (argb_t *)b;
@@ -704,14 +704,14 @@ fetch_scanline_a2b10g10r10_float (pixman_image_t *image,
 
 /* Expects a float buffer */
 static void
-fetch_scanline_x2b10g10r10_float (pixman_image_t *image,
+fetch_scanline_x2b10g10r10_float (bits_image_t   *image,
 				  int             x,
 				  int             y,
 				  int             width,
 				  uint32_t *      b,
 				  const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = (uint32_t *)bits + x;
     const uint32_t *end = pixel + width;
     argb_t *buffer = (argb_t *)b;
@@ -733,14 +733,14 @@ fetch_scanline_x2b10g10r10_float (pixman_image_t *image,
 }
 
 static void
-fetch_scanline_yuy2 (pixman_image_t *image,
+fetch_scanline_yuy2 (bits_image_t   *image,
                      int             x,
                      int             line,
                      int             width,
                      uint32_t *      buffer,
                      const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + image->bits.rowstride * line;
+    const uint32_t *bits = image->bits + image->rowstride * line;
     int i;
     
     for (i = 0; i < width; i++)
@@ -767,7 +767,7 @@ fetch_scanline_yuy2 (pixman_image_t *image,
 }
 
 static void
-fetch_scanline_yv12 (pixman_image_t *image,
+fetch_scanline_yv12 (bits_image_t   *image,
                      int             x,
                      int             line,
                      int             width,
@@ -1121,30 +1121,30 @@ store_scanline_generic_float (bits_image_t *  image,
 }
 
 static void
-fetch_scanline_generic_float (pixman_image_t *image,
+fetch_scanline_generic_float (bits_image_t *  image,
 			      int	      x,
 			      int	      y,
 			      int	      width,
 			      uint32_t *      buffer,
 			      const uint32_t *mask)
 {
-    image->bits.fetch_scanline_32 (image, x, y, width, buffer, NULL);
+    image->fetch_scanline_32 (image, x, y, width, buffer, NULL);
 
-    pixman_expand_to_float ((argb_t *)buffer, buffer, image->bits.format, width);
+    pixman_expand_to_float ((argb_t *)buffer, buffer, image->format, width);
 }
 
 /* The 32_sRGB paths should be deleted after narrow processing
  * is no longer invoked for formats that are considered wide.
  * (Also see fetch_pixel_generic_lossy_32) */
 static void
-fetch_scanline_a8r8g8b8_32_sRGB (pixman_image_t *image,
+fetch_scanline_a8r8g8b8_32_sRGB (bits_image_t   *image,
                                  int             x,
                                  int             y,
                                  int             width,
                                  uint32_t       *buffer,
                                  const uint32_t *mask)
 {
-    const uint32_t *bits = image->bits.bits + y * image->bits.rowstride;
+    const uint32_t *bits = image->bits + y * image->rowstride;
     const uint32_t *pixel = (uint32_t *)bits + x;
     const uint32_t *end = pixel + width;
     uint32_t tmp;
diff --git a/pixman/pixman/pixman-bits-image.c b/pixman/pixman/pixman-bits-image.c
index 75a39a115..7b9a97335 100644
--- a/pixman/pixman/pixman-bits-image.c
+++ b/pixman/pixman/pixman-bits-image.c
@@ -1253,9 +1253,9 @@ bits_image_fetch_untransformed_repeat_none (bits_image_t *image,
 	w = MIN (width, image->width - x);
 
 	if (wide)
-	    image->fetch_scanline_float ((pixman_image_t *)image, x, y, w, buffer, NULL);
+	    image->fetch_scanline_float (image, x, y, w, buffer, NULL);
 	else
-	    image->fetch_scanline_32 ((pixman_image_t *)image, x, y, w, buffer, NULL);
+	    image->fetch_scanline_32 (image, x, y, w, buffer, NULL);
 
 	width -= w;
 	buffer += w * (wide? 4 : 1);
@@ -1301,9 +1301,9 @@ bits_image_fetch_untransformed_repeat_normal (bits_image_t *image,
 	w = MIN (width, image->width - x);
 
 	if (wide)
-	    image->fetch_scanline_float ((pixman_image_t *)image, x, y, w, buffer, NULL);
+	    image->fetch_scanline_float (image, x, y, w, buffer, NULL);
 	else
-	    image->fetch_scanline_32 ((pixman_image_t *)image, x, y, w, buffer, NULL);
+	    image->fetch_scanline_32 (image, x, y, w, buffer, NULL);
 
 	buffer += w * (wide? 4 : 1);
 	x += w;
@@ -1528,7 +1528,7 @@ dest_get_scanline_narrow (pixman_iter_t *iter, const uint32_t *mask)
     int             width  = iter->width;
     uint32_t *	    buffer = iter->buffer;
 
-    image->bits.fetch_scanline_32 (image, x, y, width, buffer, mask);
+    image->bits.fetch_scanline_32 (&image->bits, x, y, width, buffer, mask);
     if (image->common.alpha_map)
     {
 	uint32_t *alpha;
@@ -1541,8 +1541,7 @@ dest_get_scanline_narrow (pixman_iter_t *iter, const uint32_t *mask)
 	    y -= image->common.alpha_origin_y;
 
 	    image->common.alpha_map->fetch_scanline_32 (
-		(pixman_image_t *)image->common.alpha_map,
-		x, y, width, alpha, mask);
+		image->common.alpha_map, x, y, width, alpha, mask);
 
 	    for (i = 0; i < width; ++i)
 	    {
@@ -1567,7 +1566,7 @@ dest_get_scanline_wide (pixman_iter_t *iter, const uint32_t *mask)
     argb_t *	    buffer = (argb_t *)iter->buffer;
 
     image->fetch_scanline_float (
-	(pixman_image_t *)image, x, y, width, (uint32_t *)buffer, mask);
+	image, x, y, width, (uint32_t *)buffer, mask);
     if (image->common.alpha_map)
     {
 	argb_t *alpha;
@@ -1580,8 +1579,7 @@ dest_get_scanline_wide (pixman_iter_t *iter, const uint32_t *mask)
 	    y -= image->common.alpha_origin_y;
 
 	    image->common.alpha_map->fetch_scanline_float (
-		(pixman_image_t *)image->common.alpha_map,
-		x, y, width, (uint32_t *)alpha, mask);
+		image->common.alpha_map, x, y, width, (uint32_t *)alpha, mask);
 
 	    for (i = 0; i < width; ++i)
 		buffer[i].a = alpha[i].a;
diff --git a/pixman/pixman/pixman-mmx.c b/pixman/pixman/pixman-mmx.c
index d30d36a6d..1dc2eea5a 100644
--- a/pixman/pixman/pixman-mmx.c
+++ b/pixman/pixman/pixman-mmx.c
@@ -3561,7 +3561,6 @@ mmx_composite_over_reverse_n_8888 (pixman_implementation_t *imp,
 #define BILINEAR_DECLARE_VARIABLES						\
     const __m64 mm_wt = _mm_set_pi16 (wt, wt, wt, wt);				\
     const __m64 mm_wb = _mm_set_pi16 (wb, wb, wb, wb);				\
-    const __m64 mm_BSHIFT = _mm_set_pi16 (BSHIFT, BSHIFT, BSHIFT, BSHIFT);	\
     const __m64 mm_addc7 = _mm_set_pi16 (0, 1, 0, 1);				\
     const __m64 mm_xorc7 = _mm_set_pi16 (0, BMSK, 0, BMSK);			\
     const __m64 mm_ux = _mm_set_pi16 (unit_x, unit_x, unit_x, unit_x);		\
@@ -3581,35 +3580,15 @@ do {										\
     __m64 hi = _mm_add_pi16 (t_hi, b_hi);					\
     __m64 lo = _mm_add_pi16 (t_lo, b_lo);					\
     vx += unit_x;								\
-    if (BILINEAR_INTERPOLATION_BITS < 8)					\
-    {										\
-	/* calculate horizontal weights */					\
-	__m64 mm_wh = _mm_add_pi16 (mm_addc7, _mm_xor_si64 (mm_xorc7,		\
+    /* calculate horizontal weights */						\
+    __m64 mm_wh = _mm_add_pi16 (mm_addc7, _mm_xor_si64 (mm_xorc7,		\
 			  _mm_srli_pi16 (mm_x,					\
 					 16 - BILINEAR_INTERPOLATION_BITS)));	\
-	/* horizontal interpolation */						\
-	__m64 p = _mm_unpacklo_pi16 (lo, hi);					\
-	__m64 q = _mm_unpackhi_pi16 (lo, hi);					\
-	lo = _mm_madd_pi16 (p, mm_wh);						\
-	hi = _mm_madd_pi16 (q, mm_wh);						\
-    }										\
-    else									\
-    {										\
-	/* calculate horizontal weights */					\
-	__m64 mm_wh_lo = _mm_sub_pi16 (mm_BSHIFT, _mm_srli_pi16 (mm_x,		\
-					16 - BILINEAR_INTERPOLATION_BITS));	\
-	__m64 mm_wh_hi = _mm_srli_pi16 (mm_x,					\
-					16 - BILINEAR_INTERPOLATION_BITS);	\
-	/* horizontal interpolation */						\
-	__m64 mm_lo_lo = _mm_mullo_pi16 (lo, mm_wh_lo);				\
-	__m64 mm_lo_hi = _mm_mullo_pi16 (hi, mm_wh_hi);				\
-	__m64 mm_hi_lo = _mm_mulhi_pu16 (lo, mm_wh_lo);				\
-	__m64 mm_hi_hi = _mm_mulhi_pu16 (hi, mm_wh_hi);				\
-	lo = _mm_add_pi32 (_mm_unpacklo_pi16 (mm_lo_lo, mm_hi_lo),		\
-			   _mm_unpacklo_pi16 (mm_lo_hi, mm_hi_hi));		\
-	hi = _mm_add_pi32 (_mm_unpackhi_pi16 (mm_lo_lo, mm_hi_lo),		\
-			   _mm_unpackhi_pi16 (mm_lo_hi, mm_hi_hi));		\
-    }										\
+    /* horizontal interpolation */						\
+    __m64 p = _mm_unpacklo_pi16 (lo, hi);					\
+    __m64 q = _mm_unpackhi_pi16 (lo, hi);					\
+    lo = _mm_madd_pi16 (p, mm_wh);						\
+    hi = _mm_madd_pi16 (q, mm_wh);						\
     mm_x = _mm_add_pi16 (mm_x, mm_ux);						\
     /* shift and pack the result */						\
     hi = _mm_srli_pi32 (hi, BILINEAR_INTERPOLATION_BITS * 2);			\
diff --git a/pixman/pixman/pixman-private.h b/pixman/pixman/pixman-private.h
index 964660508..120196ddf 100644
--- a/pixman/pixman/pixman-private.h
+++ b/pixman/pixman/pixman-private.h
@@ -57,7 +57,7 @@ struct argb_t
     float b;
 };
 
-typedef void (*fetch_scanline_t) (pixman_image_t *image,
+typedef void (*fetch_scanline_t) (bits_image_t   *image,
 				  int             x,
 				  int             y,
 				  int             width,
diff --git a/pixman/pixman/pixman-sse2.c b/pixman/pixman/pixman-sse2.c
index 932bfd8ee..29ae623c5 100644
--- a/pixman/pixman/pixman-sse2.c
+++ b/pixman/pixman/pixman-sse2.c
@@ -5554,7 +5554,6 @@ FAST_NEAREST_MAINLOOP_COMMON (sse2_8888_n_8888_normal_OVER,
 			      scaled_nearest_scanline_sse2_8888_n_8888_OVER,
 			      uint32_t, uint32_t, uint32_t, NORMAL, TRUE, TRUE)
 
-#if BILINEAR_INTERPOLATION_BITS < 8
 # define BILINEAR_DECLARE_VARIABLES						\
     const __m128i xmm_wt = _mm_set_epi16 (wt, wt, wt, wt, wt, wt, wt, wt);	\
     const __m128i xmm_wb = _mm_set_epi16 (wb, wb, wb, wb, wb, wb, wb, wb);	\
@@ -5564,21 +5563,10 @@ FAST_NEAREST_MAINLOOP_COMMON (sse2_8888_n_8888_normal_OVER,
     const __m128i xmm_zero = _mm_setzero_si128 ();				\
     __m128i xmm_x = _mm_set_epi16 (vx&0xffff, (-(vx + 1))&0xffff, vx&0xffff, (-(vx + 1))&0xffff,		\
 				   vx&0xffff, (-(vx + 1))&0xffff, vx&0xffff, (-(vx + 1))&0xffff)
-#else
-# define BILINEAR_DECLARE_VARIABLES						\
-    const __m128i xmm_wt = _mm_set_epi16 (wt, wt, wt, wt, wt, wt, wt, wt);	\
-    const __m128i xmm_wb = _mm_set_epi16 (wb, wb, wb, wb, wb, wb, wb, wb);	\
-    const __m128i xmm_addc = _mm_set_epi16 (0, 0, 0, 0, 1, 1, 1, 1);		\
-    const __m128i xmm_ux = _mm_set_epi16 (unit_x&0xffff, unit_x&0xffff, unit_x&0xffff, unit_x&0xffff,	\
-					  (-unit_x)&0xffff, (-unit_x)&0xffff, (-unit_x)&0xffff, (-unit_x)&0xffff);	\
-    const __m128i xmm_zero = _mm_setzero_si128 ();				\
-    __m128i xmm_x = _mm_set_epi16 (vx&0xffff, vx&0xffff, vx&0xffff, vx&0xffff,				\
-				   (-(vx + 1))&0xffff, (-(vx + 1))&0xffff, (-(vx + 1))&0xffff, (-(vx + 1))&0xffff)
-#endif
 
 #define BILINEAR_INTERPOLATE_ONE_PIXEL(pix)					\
 do {										\
-    __m128i xmm_wh, xmm_lo, xmm_hi, a;						\
+    __m128i xmm_wh, a;								\
     /* fetch 2x2 pixel block into sse2 registers */				\
     __m128i tltr = _mm_loadl_epi64 (						\
 			    (__m128i *)&src_top[pixman_fixed_to_int (vx)]);	\
@@ -5590,28 +5578,13 @@ do {										\
 					xmm_wt),				\
 		       _mm_mullo_epi16 (_mm_unpacklo_epi8 (blbr, xmm_zero),	\
 					xmm_wb));				\
-    if (BILINEAR_INTERPOLATION_BITS < 8)					\
-    {										\
-	/* calculate horizontal weights */					\
-	xmm_wh = _mm_add_epi16 (xmm_addc, _mm_srli_epi16 (xmm_x,		\
-					16 - BILINEAR_INTERPOLATION_BITS));	\
-	xmm_x = _mm_add_epi16 (xmm_x, xmm_ux);					\
-	/* horizontal interpolation */						\
-	a = _mm_madd_epi16 (_mm_unpackhi_epi16 (_mm_shuffle_epi32 (		\
-		a, _MM_SHUFFLE (1, 0, 3, 2)), a), xmm_wh);			\
-    }										\
-    else									\
-    {										\
-	/* calculate horizontal weights */					\
-	xmm_wh = _mm_add_epi16 (xmm_addc, _mm_srli_epi16 (xmm_x,		\
-					16 - BILINEAR_INTERPOLATION_BITS));	\
-	xmm_x = _mm_add_epi16 (xmm_x, xmm_ux);					\
-	/* horizontal interpolation */						\
-	xmm_lo = _mm_mullo_epi16 (a, xmm_wh);					\
-	xmm_hi = _mm_mulhi_epu16 (a, xmm_wh);					\
-	a = _mm_add_epi32 (_mm_unpacklo_epi16 (xmm_lo, xmm_hi),			\
-			   _mm_unpackhi_epi16 (xmm_lo, xmm_hi));		\
-    }										\
+    /* calculate horizontal weights */						\
+    xmm_wh = _mm_add_epi16 (xmm_addc, _mm_srli_epi16 (xmm_x,			\
+			      16 - BILINEAR_INTERPOLATION_BITS));		\
+    xmm_x = _mm_add_epi16 (xmm_x, xmm_ux);					\
+    /* horizontal interpolation */						\
+    a = _mm_madd_epi16 (_mm_unpackhi_epi16 (_mm_shuffle_epi32 (			\
+			       a, _MM_SHUFFLE (1, 0, 3, 2)), a), xmm_wh);	\
     /* shift and pack the result */						\
     a = _mm_srli_epi32 (a, BILINEAR_INTERPOLATION_BITS * 2);			\
     a = _mm_packs_epi32 (a, a);							\
diff --git a/pixman/test/affine-test.c b/pixman/test/affine-test.c
index 2506250db..c1649ed5e 100644
--- a/pixman/test/affine-test.c
+++ b/pixman/test/affine-test.c
@@ -306,9 +306,7 @@ test_composite (int      testnum,
     return crc32;
 }
 
-#if BILINEAR_INTERPOLATION_BITS == 8
-#define CHECKSUM 0x2CDF1F07
-#elif BILINEAR_INTERPOLATION_BITS == 7
+#if BILINEAR_INTERPOLATION_BITS == 7
 #define CHECKSUM 0xBC00B1DF
 #elif BILINEAR_INTERPOLATION_BITS == 4
 #define CHECKSUM 0xA227306B
diff --git a/pixman/test/scaling-test.c b/pixman/test/scaling-test.c
index a8cb4c47b..b4142a796 100644
--- a/pixman/test/scaling-test.c
+++ b/pixman/test/scaling-test.c
@@ -375,9 +375,7 @@ test_composite (int      testnum,
     return crc32;
 }
 
-#if BILINEAR_INTERPOLATION_BITS == 8
-#define CHECKSUM 0x9096E6B6
-#elif BILINEAR_INTERPOLATION_BITS == 7
+#if BILINEAR_INTERPOLATION_BITS == 7
 #define CHECKSUM 0xCE8EC6BA
 #elif BILINEAR_INTERPOLATION_BITS == 4
 #define CHECKSUM 0xAB1D39BE