fontconfig mesa pixman git update 1 July 2014

pixman           commit 5a2edb3f2c2cfde6b25ac614e2004a9f78583d74
fontconfig       commit dab60e4476ada4ad4639599ea24dd012d4a79584
mesa             commit 1bfc0a11027449ae7ab7c28eb695f26de530eccf

70 files changed, 1080 insertions, 1114 deletions

diff --git a/mesalib/configure.ac b/mesalib/configure.ac
index e4cfb07ee..0d4fb0518 100644
--- a/mesalib/configure.ac
+++ b/mesalib/configure.ac
@@ -1317,6 +1317,7 @@ if test "x$enable_gallium_egl" = xyes; then
 
     GALLIUM_STATE_TRACKERS_DIRS="egl $GALLIUM_STATE_TRACKERS_DIRS"
     GALLIUM_TARGET_DIRS="$GALLIUM_TARGET_DIRS egl-static"
+#    enable_gallium_loader=yes
 fi
 AM_CONDITIONAL(HAVE_GALLIUM_EGL, test "x$enable_gallium_egl" = xyes)
 
@@ -1425,6 +1426,7 @@ AM_CONDITIONAL(HAVE_ST_VDPAU, test "x$enable_vdpau" = xyes)
 if test "x$enable_omx" = xyes; then
     PKG_CHECK_MODULES([OMX], [libomxil-bellagio >= $LIBOMXIL_BELLAGIO_REQUIRED x11-xcb xcb-dri2 >= $XCBDRI2_REQUIRED])
     GALLIUM_STATE_TRACKERS_DIRS="$GALLIUM_STATE_TRACKERS_DIRS omx"
+    enable_gallium_loader=yes
 fi
 AM_CONDITIONAL(HAVE_ST_OMX, test "x$enable_omx" = xyes)
 
@@ -1602,7 +1604,7 @@ fi
 
 AC_ARG_WITH([egl-driver-dir],
     [AS_HELP_STRING([--with-egl-driver-dir=DIR],
-                    [directory for EGL drivers [[default=${libdir}/egl]]])],
+                    [directory for EGL drivers @<:@default=${libdir}/egl@:>@])],
     [EGL_DRIVER_INSTALL_DIR="$withval"],
     [EGL_DRIVER_INSTALL_DIR='${libdir}/egl'])
 AC_SUBST([EGL_DRIVER_INSTALL_DIR])
@@ -1909,14 +1911,14 @@ if test -n "$with_gallium_drivers"; then
             HAVE_GALLIUM_SVGA=yes
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS svga softpipe"
             gallium_require_drm_loader
-            gallium_check_st "svga/drm" "dri-vmwgfx" ""
+            gallium_check_st "svga/drm" "dri-vmwgfx" "xa/vmwgfx"
             ;;
         xi915)
             HAVE_GALLIUM_I915=yes
             PKG_CHECK_MODULES([INTEL], [libdrm_intel >= $LIBDRM_INTEL_REQUIRED])
             gallium_require_drm_loader
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS i915"
-            gallium_check_st "i915/drm" "dri-i915"
+            gallium_check_st "i915/drm" "dri-i915" "xa/i915"
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xilo)
@@ -1924,7 +1926,7 @@ if test -n "$with_gallium_drivers"; then
             PKG_CHECK_MODULES([INTEL], [libdrm_intel >= $LIBDRM_INTEL_REQUIRED])
             gallium_require_drm_loader
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS ilo"
-            gallium_check_st "intel/drm" "dri-ilo"
+            gallium_check_st "intel/drm" "dri-ilo" "xa/ilo"
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xr300)
@@ -1933,7 +1935,7 @@ if test -n "$with_gallium_drivers"; then
             gallium_require_drm_loader
             gallium_require_llvm "Gallium R300"
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS r300"
-            gallium_check_st "radeon/drm" "r300/dri" "" "" ""
+            gallium_check_st "radeon/drm" "r300/dri" "xa/r300" "" ""
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xr600)
@@ -1951,7 +1953,7 @@ if test -n "$with_gallium_drivers"; then
             if test "x$enable_opencl" = xyes; then
                 LLVM_COMPONENTS="${LLVM_COMPONENTS} bitreader asmparser"
             fi
-            gallium_check_st "radeon/drm" "r600/dri" "" "xvmc/r600" "vdpau/r600" "r600/omx"
+            gallium_check_st "radeon/drm" "r600/dri" "xa/r600" "xvmc/r600" "vdpau/r600" "omx/r600"
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xradeonsi)
@@ -1961,7 +1963,7 @@ if test -n "$with_gallium_drivers"; then
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS radeonsi"
             radeon_llvm_check "radeonsi"
             require_egl_drm "radeonsi"
-            gallium_check_st "radeon/drm" "radeonsi/dri" "" "" "vdpau/radeonsi" "radeonsi/omx"
+            gallium_check_st "radeon/drm" "radeonsi/dri" "xa/radeonsi" "" "vdpau/radeonsi" "omx/radeonsi"
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xnouveau)
@@ -1969,7 +1971,7 @@ if test -n "$with_gallium_drivers"; then
             PKG_CHECK_MODULES([NOUVEAU], [libdrm_nouveau >= $LIBDRM_NOUVEAU_REQUIRED])
             gallium_require_drm_loader
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS nouveau"
-            gallium_check_st "nouveau/drm" "dri-nouveau" "" "xvmc/nouveau" "vdpau/nouveau" "omx-nouveau"
+            gallium_check_st "nouveau/drm" "dri-nouveau" "xa/nouveau" "xvmc/nouveau" "vdpau/nouveau" "omx/nouveau"
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xfreedreno)
@@ -1977,7 +1979,7 @@ if test -n "$with_gallium_drivers"; then
             PKG_CHECK_MODULES([FREEDRENO], [libdrm_freedreno >= $LIBDRM_FREEDRENO_REQUIRED])
             gallium_require_drm_loader
             GALLIUM_DRIVERS_DIRS="$GALLIUM_DRIVERS_DIRS freedreno"
-            gallium_check_st "freedreno/drm" "dri-freedreno" "" "" ""
+            gallium_check_st "freedreno/drm" "dri-freedreno" "xa/freedreno" "" ""
             DRICOMMON_NEED_LIBDRM=yes
             ;;
         xswrast)
@@ -2214,16 +2216,14 @@ AC_CONFIG_FILES([Makefile
 		src/gallium/targets/egl-static/Makefile
 		src/gallium/targets/gbm/Makefile
 		src/gallium/targets/libgl-xlib/Makefile
-		src/gallium/targets/omx-nouveau/Makefile
+		src/gallium/targets/omx/Makefile
 		src/gallium/targets/opencl/Makefile
 		src/gallium/targets/osmesa/Makefile
 		src/gallium/targets/osmesa/osmesa.pc
 		src/gallium/targets/pipe-loader/Makefile
 		src/gallium/targets/r300/dri/Makefile
 		src/gallium/targets/r600/dri/Makefile
-		src/gallium/targets/r600/omx/Makefile
 		src/gallium/targets/radeonsi/dri/Makefile
-		src/gallium/targets/radeonsi/omx/Makefile
 		src/gallium/targets/vdpau/Makefile
 		src/gallium/targets/xa/Makefile
 		src/gallium/targets/xa/xatracker.pc
diff --git a/mesalib/docs/GL3.txt b/mesalib/docs/GL3.txt
index 47165fe1e..bc418a908 100644
--- a/mesalib/docs/GL3.txt
+++ b/mesalib/docs/GL3.txt
@@ -101,21 +101,21 @@ GL 4.0:
   GL_ARB_draw_indirect                                 DONE (i965)
   GL_ARB_gpu_shader5                                   started
   - 'precise' qualifier                                DONE
-  - Dynamically uniform sampler array indices          not started
-  - Dynamically uniform UBO array indices              not started
+  - Dynamically uniform sampler array indices          started (Chris)
+  - Dynamically uniform UBO array indices              started (Chris)
   - Implicit signed -> unsigned conversions            DONE
   - Fused multiply-add                                 DONE
   - Packing/bitfield/conversion functions              DONE
   - Enhanced textureGather                             DONE
   - Geometry shader instancing                         DONE
-  - Geometry shader multiple streams                   not started
+  - Geometry shader multiple streams                   DONE (i965)
   - Enhanced per-sample shading                        DONE
-  - Interpolation functions                            started
+  - Interpolation functions                            started (Chris)
   - New overload resolution rules                      DONE
-  GL_ARB_gpu_shader_fp64                               not started
+  GL_ARB_gpu_shader_fp64                               started (Dave)
   GL_ARB_sample_shading                                DONE (i965, nv50, nvc0, radeonsi)
   GL_ARB_shader_subroutine                             not started
-  GL_ARB_tessellation_shader                           not started
+  GL_ARB_tessellation_shader                           started (Fabian)
   GL_ARB_texture_buffer_object_rgb32                   DONE (i965, nvc0, r600, radeonsi, softpipe)
   GL_ARB_texture_cube_map_array                        DONE (i965, nv50, nvc0, r600, radeonsi, softpipe)
   GL_ARB_texture_gather                                DONE (i965, nv50, nvc0, radeonsi)
@@ -130,7 +130,7 @@ GL 4.1:
   GL_ARB_ES2_compatibility                             DONE (i965, nv50, nvc0, r300, r600, radeonsi)
   GL_ARB_get_program_binary                            DONE (0 binary formats)
   GL_ARB_separate_shader_objects                       DONE (all drivers)
-  GL_ARB_shader_precision                              not started
+  GL_ARB_shader_precision                              started (Micah)
   GL_ARB_vertex_attrib_64bit                           not started
   GL_ARB_viewport_array                                DONE (i965, nv50, nvc0, r600)
 
@@ -154,7 +154,7 @@ GL 4.2:
 GL 4.3:
 
   GLSL 4.3                                             not started
-  GL_ARB_arrays_of_arrays                              started
+  GL_ARB_arrays_of_arrays                              started (Timothy)
   GL_ARB_ES3_compatibility                             DONE (i965)
   GL_ARB_clear_buffer_object                           DONE (all drivers)
   GL_ARB_compute_shader                                started (Paul Berry)
diff --git a/mesalib/docs/relnotes/10.3.html b/mesalib/docs/relnotes/10.3.html
index ecc8580f4..47d2c5fff 100644
--- a/mesalib/docs/relnotes/10.3.html
+++ b/mesalib/docs/relnotes/10.3.html
@@ -53,6 +53,7 @@ Note: some of the new features are only available with certain drivers.
 <li>GL_ARB_texture_query_levels on nv50, nvc0, llvmpipe, r600, radeonsi, softpipe</li>
 <li>GL_ARB_texture_query_lod on radeonsi</li>
 <li>GL_ARB_viewport_array on nvc0</li>
+<li>GL_ARB_seamless_cubemap_per_texture on i965, llvmpipe, nvc0, r600, radeonsi, softpipe</li>
 </ul>
 
 
@@ -63,6 +64,7 @@ TBD.
 <h2>Changes</h2>
 
 <ul>
+<li>Removed support for the GL_ATI_envmap_bumpmap extension</li>
 </ul>
 
 </div>
diff --git a/mesalib/include/GL/glx.h b/mesalib/include/GL/glx.h
index 234abc050..78f5052b2 100644
--- a/mesalib/include/GL/glx.h
+++ b/mesalib/include/GL/glx.h
@@ -518,7 +518,7 @@ typedef struct {
     unsigned long serial;	/* # of last request processed by server */
     Bool send_event;		/* true if this came from a SendEvent request */
     Display *display;		/* Display the event was read from */
-    GLXDrawable drawable;	/* drawable on which event was requested in event mask */
+    Drawable drawable;	/* drawable on which event was requested in event mask */
     int event_type;
     int64_t ust;
     int64_t msc;
diff --git a/mesalib/include/GL/internal/dri_interface.h b/mesalib/include/GL/internal/dri_interface.h
index 4d57d0b82..725d622a1 100644
--- a/mesalib/include/GL/internal/dri_interface.h
+++ b/mesalib/include/GL/internal/dri_interface.h
@@ -40,6 +40,7 @@
 #ifndef DRI_INTERFACE_H
 #define DRI_INTERFACE_H
 
+#include <stdbool.h>
 /* For archs with no drm.h */
 #if defined(__APPLE__) || defined(__CYGWIN__) || defined(__GNU__)
 #ifndef __NOT_HAVE_DRM_H
@@ -1005,7 +1006,7 @@ struct __DRIdri2ExtensionRec {
  * extensions.
  */
 #define __DRI_IMAGE "DRI_IMAGE"
-#define __DRI_IMAGE_VERSION 8
+#define __DRI_IMAGE_VERSION 9
 
 /**
  * These formats correspond to the similarly named MESA_FORMAT_*
@@ -1133,6 +1134,13 @@ enum __DRIChromaSiting {
 #define __DRI_IMAGE_ERROR_BAD_PARAMETER 3
 /*@}*/
 
+/**
+ * blitImage flags
+ */
+
+#define __BLIT_FLAG_FLUSH		0x0001
+#define __BLIT_FLAG_FINISH		0x0002
+
 typedef struct __DRIimageRec          __DRIimage;
 typedef struct __DRIimageExtensionRec __DRIimageExtension;
 struct __DRIimageExtensionRec {
@@ -1239,6 +1247,21 @@ struct __DRIimageExtensionRec {
                                          enum __DRIChromaSiting vert_siting,
                                          unsigned *error,
                                          void *loaderPrivate);
+
+   /**
+    * Blit a part of a __DRIimage to another and flushes
+    *
+    * flush_flag:
+    *    0:                  no flush
+    *    __BLIT_FLAG_FLUSH:  flush after the blit operation
+    *    __BLIT_FLAG_FINISH: flush and wait the blit finished
+    *
+    * \since 9
+    */
+   void (*blitImage)(__DRIcontext *context, __DRIimage *dst, __DRIimage *src,
+                     int dstx0, int dsty0, int dstwidth, int dstheight,
+                     int srcx0, int srcy0, int srcwidth, int srcheight,
+                     int flush_flag);
 };
 
 
@@ -1272,9 +1295,9 @@ typedef struct __DRI2configQueryExtensionRec __DRI2configQueryExtension;
 struct __DRI2configQueryExtensionRec {
    __DRIextension base;
 
-   int (*configQueryb)(__DRIscreen *screen, const char *var, GLboolean *val);
-   int (*configQueryi)(__DRIscreen *screen, const char *var, GLint *val);
-   int (*configQueryf)(__DRIscreen *screen, const char *var, GLfloat *val);
+   int (*configQueryb)(__DRIscreen *screen, const char *var, bool *val);
+   int (*configQueryi)(__DRIscreen *screen, const char *var, int *val);
+   int (*configQueryf)(__DRIscreen *screen, const char *var, float *val);
 };
 
 /**
diff --git a/mesalib/src/Makefile.am b/mesalib/src/Makefile.am
index 9d1580f90..d4a7090dc 100644
--- a/mesalib/src/Makefile.am
+++ b/mesalib/src/Makefile.am
@@ -19,12 +19,14 @@
 # FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 # IN THE SOFTWARE.
 
-SUBDIRS = gtest loader mapi
+SUBDIRS = gtest mapi
 
 if NEED_OPENGL_COMMON
 SUBDIRS += glsl mesa
 endif
 
+SUBDIRS += loader
+
 if HAVE_DRI_GLX
 SUBDIRS += glx
 endif
diff --git a/mesalib/src/gallium/Automake.inc b/mesalib/src/gallium/Automake.inc
index 8bf2a12e8..4600b9c5b 100644
--- a/mesalib/src/gallium/Automake.inc
+++ b/mesalib/src/gallium/Automake.inc
@@ -39,7 +39,7 @@ GALLIUM_DRI_CFLAGS = \
 	$(LIBDRM_CFLAGS) \
 	$(VISIBILITY_CFLAGS)
 
-GALLIUM_VIDEO_CFLAGS = \
+GALLIUM_TARGET_CFLAGS = \
 	-I$(top_srcdir)/include \
 	-I$(top_srcdir)/src/loader \
 	-I$(top_srcdir)/src/gallium/include \
@@ -59,20 +59,10 @@ GALLIUM_DRI_LINKER_FLAGS = \
 	-avoid-version \
 	$(GC_SECTIONS)
 
-GALLIUM_OMX_LINKER_FLAGS = \
-	-shared \
-	-module \
-	-no-undefined \
-	-avoid-version \
-	$(GC_SECTIONS) \
-	$(LD_NO_UNDEFINED)
-
 if HAVE_LD_VERSION_SCRIPT
 GALLIUM_DRI_LINKER_FLAGS += \
 	-Wl,--version-script=$(top_srcdir)/src/gallium/targets/dri.sym
 
-GALLIUM_OMX_LINKER_FLAGS += \
-	-Wl,--version-script=$(top_srcdir)/src/gallium/targets/omx.sym
 endif
 
 
@@ -90,12 +80,6 @@ GALLIUM_DRI_LIB_DEPS = \
 	$(EXPAT_LIBS) \
 	$(GALLIUM_COMMON_LIB_DEPS)
 
-GALLIUM_OMX_LIB_DEPS = \
-	$(top_builddir)/src/gallium/auxiliary/libgallium.la \
-	$(top_builddir)/src/gallium/state_trackers/omx/libomxtracker.la \
-	$(OMX_LIBS) \
-	$(GALLIUM_COMMON_LIB_DEPS)
-
 GALLIUM_WINSYS_CFLAGS = \
 	-I$(top_srcdir)/include \
 	-I$(top_srcdir)/src/gallium/include \
@@ -104,13 +88,8 @@ GALLIUM_WINSYS_CFLAGS = \
 	$(VISIBILITY_CFLAGS)
 
 if HAVE_MESA_LLVM
-
 GALLIUM_DRI_LINKER_FLAGS += $(LLVM_LDFLAGS)
-GALLIUM_OMX_LINKER_FLAGS += $(LLVM_LDFLAGS)
-
 GALLIUM_DRI_LIB_DEPS += $(LLVM_LIBS)
-GALLIUM_OMX_LIB_DEPS += $(LLVM_LIBS)
-
 endif
 
 
diff --git a/mesalib/src/glsl/ast.h b/mesalib/src/glsl/ast.h
index 56e7bd86f..15bf086a1 100644
--- a/mesalib/src/glsl/ast.h
+++ b/mesalib/src/glsl/ast.h
@@ -509,6 +509,8 @@ struct ast_type_qualifier {
          /** \name Layout qualifiers for GL_ARB_gpu_shader5 */
          /** \{ */
          unsigned invocations:1;
+         unsigned stream:1; /**< Has stream value assigned  */
+         unsigned explicit_stream:1; /**< stream value assigned explicitly by shader code */
          /** \} */
       }
       /** \brief Set of flags, accessed by name. */
@@ -542,6 +544,9 @@ struct ast_type_qualifier {
    /** Maximum output vertices in GLSL 1.50 geometry shaders. */
    int max_vertices;
 
+   /** Stream in GLSL 1.50 geometry shaders. */
+   unsigned stream;
+
    /** Input or output primitive type in GLSL 1.50 geometry shaders */
    GLenum prim_type;
 
diff --git a/mesalib/src/glsl/ast_to_hir.cpp b/mesalib/src/glsl/ast_to_hir.cpp
index 7ba04a808..042ef243b 100644
--- a/mesalib/src/glsl/ast_to_hir.cpp
+++ b/mesalib/src/glsl/ast_to_hir.cpp
@@ -2461,6 +2461,11 @@ apply_type_qualifier_to_variable(const struct ast_type_qualifier *qual,
    if (qual->flags.q.sample)
       var->data.sample = 1;
 
+   if (state->stage == MESA_SHADER_GEOMETRY &&
+       qual->flags.q.out && qual->flags.q.stream) {
+      var->data.stream = qual->stream;
+   }
+
    if (qual->flags.q.attribute && state->stage != MESA_SHADER_VERTEX) {
       var->type = glsl_type::error_type;
       _mesa_glsl_error(loc, state,
@@ -5100,6 +5105,9 @@ ast_process_structure_or_interface_block(exec_list *instructions,
          fields[i].centroid = qual->flags.q.centroid ? 1 : 0;
          fields[i].sample = qual->flags.q.sample ? 1 : 0;
 
+         /* Only save explicitly defined streams in block's field */
+         fields[i].stream = qual->flags.q.explicit_stream ? qual->stream : -1;
+
          if (qual->flags.q.row_major || qual->flags.q.column_major) {
             if (!qual->flags.q.uniform) {
                _mesa_glsl_error(&loc, state,
@@ -5548,6 +5556,17 @@ ast_interface_block::hir(exec_list *instructions,
          var->data.sample = fields[i].sample;
          var->init_interface_type(block_type);
 
+         if (fields[i].stream != -1 &&
+             ((unsigned)fields[i].stream) != this->layout.stream) {
+            _mesa_glsl_error(&loc, state,
+                             "stream layout qualifier on "
+                             "interface block member `%s' does not match "
+                             "the interface block (%d vs %d)",
+                             var->name, fields[i].stream, this->layout.stream);
+         }
+
+         var->data.stream = this->layout.stream;
+
          if (redeclaring_per_vertex) {
             ir_variable *earlier =
                get_variable_being_redeclared(var, loc, state,
diff --git a/mesalib/src/glsl/ast_type.cpp b/mesalib/src/glsl/ast_type.cpp
index 77053d5b1..017f23d0e 100644
--- a/mesalib/src/glsl/ast_type.cpp
+++ b/mesalib/src/glsl/ast_type.cpp
@@ -125,9 +125,13 @@ ast_type_qualifier::merge_qualifier(YYLTYPE *loc,
    /* Uniform block layout qualifiers get to overwrite each
     * other (rightmost having priority), while all other
     * qualifiers currently don't allow duplicates.
+    *
+    * Geometry shaders can have several layout qualifiers
+    * assigning different stream values.
     */
 
-   if ((this->flags.i & q.flags.i & ~(ubo_mat_mask.flags.i |
+   if ((state->stage != MESA_SHADER_GEOMETRY) &&
+       (this->flags.i & q.flags.i & ~(ubo_mat_mask.flags.i |
 				      ubo_layout_mask.flags.i |
                                       ubo_binding_mask.flags.i)) != 0) {
       _mesa_glsl_error(loc, state,
@@ -154,6 +158,39 @@ ast_type_qualifier::merge_qualifier(YYLTYPE *loc,
       this->max_vertices = q.max_vertices;
    }
 
+   if (state->stage == MESA_SHADER_GEOMETRY &&
+       state->has_explicit_attrib_stream()) {
+      if (q.flags.q.stream && q.stream >= state->ctx->Const.MaxVertexStreams) {
+         _mesa_glsl_error(loc, state,
+                          "`stream' value is larger than MAX_VERTEX_STREAMS - 1 "
+                          "(%d > %d)",
+                          q.stream, state->ctx->Const.MaxVertexStreams - 1);
+      }
+      if (this->flags.q.explicit_stream &&
+          this->stream >= state->ctx->Const.MaxVertexStreams) {
+         _mesa_glsl_error(loc, state,
+                          "`stream' value is larger than MAX_VERTEX_STREAMS - 1 "
+                          "(%d > %d)",
+                          this->stream, state->ctx->Const.MaxVertexStreams - 1);
+      }
+
+      if (!this->flags.q.explicit_stream) {
+         if (q.flags.q.stream) {
+            this->flags.q.stream = 1;
+            this->stream = q.stream;
+         } else if (!this->flags.q.stream && this->flags.q.out) {
+            /* Assign default global stream value */
+            this->flags.q.stream = 1;
+            this->stream = state->out_qualifier->stream;
+         }
+      } else {
+         if (q.flags.q.explicit_stream) {
+            _mesa_glsl_error(loc, state,
+                             "duplicate layout `stream' qualifier");
+         }
+      }
+   }
+
    if ((q.flags.i & ubo_mat_mask.flags.i) != 0)
       this->flags.i &= ~ubo_mat_mask.flags.i;
    if ((q.flags.i & ubo_layout_mask.flags.i) != 0)
diff --git a/mesalib/src/glsl/builtin_functions.cpp b/mesalib/src/glsl/builtin_functions.cpp
index f9f06862e..94b578107 100644
--- a/mesalib/src/glsl/builtin_functions.cpp
+++ b/mesalib/src/glsl/builtin_functions.cpp
@@ -359,6 +359,12 @@ shader_image_load_store(const _mesa_glsl_parse_state *state)
            state->ARB_shader_image_load_store_enable);
 }
 
+static bool
+gs_streams(const _mesa_glsl_parse_state *state)
+{
+   return gpu_shader5(state) && gs_only(state);
+}
+
 /** @} */
 
 /******************************************************************************/
@@ -594,6 +600,10 @@ private:
 
    B0(EmitVertex)
    B0(EndPrimitive)
+   ir_function_signature *_EmitStreamVertex(builtin_available_predicate avail,
+                                            const glsl_type *stream_type);
+   ir_function_signature *_EndStreamPrimitive(builtin_available_predicate avail,
+                                              const glsl_type *stream_type);
 
    B2(textureQueryLod);
    B1(textureQueryLevels);
@@ -1708,6 +1718,14 @@ builtin_builder::create_builtins()
 
    add_function("EmitVertex",   _EmitVertex(),   NULL);
    add_function("EndPrimitive", _EndPrimitive(), NULL);
+   add_function("EmitStreamVertex",
+                _EmitStreamVertex(gs_streams, glsl_type::uint_type),
+                _EmitStreamVertex(gs_streams, glsl_type::int_type),
+                NULL);
+   add_function("EndStreamPrimitive",
+                _EndStreamPrimitive(gs_streams, glsl_type::uint_type),
+                _EndStreamPrimitive(gs_streams, glsl_type::int_type),
+                NULL);
 
    add_function("textureQueryLOD",
                 _textureQueryLod(glsl_type::sampler1D_type,  glsl_type::float_type),
@@ -3872,7 +3890,28 @@ builtin_builder::_EmitVertex()
 {
    MAKE_SIG(glsl_type::void_type, gs_only, 0);
 
-   body.emit(new(mem_ctx) ir_emit_vertex());
+   ir_rvalue *stream = new(mem_ctx) ir_constant(0, 1);
+   body.emit(new(mem_ctx) ir_emit_vertex(stream));
+
+   return sig;
+}
+
+ir_function_signature *
+builtin_builder::_EmitStreamVertex(builtin_available_predicate avail,
+                                   const glsl_type *stream_type)
+{
+   /* Section 8.12 (Geometry Shader Functions) of the GLSL 4.0 spec says:
+    *
+    *     "Emit the current values of output variables to the current output
+    *     primitive on stream stream. The argument to stream must be a constant
+    *     integral expression."
+    */
+   ir_variable *stream =
+      new(mem_ctx) ir_variable(stream_type, "stream", ir_var_const_in);
+
+   MAKE_SIG(glsl_type::void_type, avail, 1, stream);
+
+   body.emit(new(mem_ctx) ir_emit_vertex(var_ref(stream)));
 
    return sig;
 }
@@ -3882,7 +3921,28 @@ builtin_builder::_EndPrimitive()
 {
    MAKE_SIG(glsl_type::void_type, gs_only, 0);
 
-   body.emit(new(mem_ctx) ir_end_primitive());
+   ir_rvalue *stream = new(mem_ctx) ir_constant(0, 1);
+   body.emit(new(mem_ctx) ir_end_primitive(stream));
+
+   return sig;
+}
+
+ir_function_signature *
+builtin_builder::_EndStreamPrimitive(builtin_available_predicate avail,
+                                     const glsl_type *stream_type)
+{
+   /* Section 8.12 (Geometry Shader Functions) of the GLSL 4.0 spec says:
+    *
+    *     "Completes the current output primitive on stream stream and starts
+    *     a new one. The argument to stream must be a constant integral
+    *     expression."
+    */
+   ir_variable *stream =
+      new(mem_ctx) ir_variable(stream_type, "stream", ir_var_const_in);
+
+   MAKE_SIG(glsl_type::void_type, avail, 1, stream);
+
+   body.emit(new(mem_ctx) ir_end_primitive(var_ref(stream)));
 
    return sig;
 }
diff --git a/mesalib/src/glsl/builtin_variables.cpp b/mesalib/src/glsl/builtin_variables.cpp
index b9c69d23c..9f9571619 100644
--- a/mesalib/src/glsl/builtin_variables.cpp
+++ b/mesalib/src/glsl/builtin_variables.cpp
@@ -160,14 +160,6 @@ static const struct gl_builtin_uniform_element gl_NormalScale_elements[] = {
    {NULL, {STATE_NORMAL_SCALE}, SWIZZLE_XXXX},
 };
 
-static const struct gl_builtin_uniform_element gl_BumpRotMatrix0MESA_elements[] = {
-   {NULL, {STATE_INTERNAL, STATE_ROT_MATRIX_0}, SWIZZLE_XYZW},
-};
-
-static const struct gl_builtin_uniform_element gl_BumpRotMatrix1MESA_elements[] = {
-   {NULL, {STATE_INTERNAL, STATE_ROT_MATRIX_1}, SWIZZLE_XYZW},
-};
-
 static const struct gl_builtin_uniform_element gl_FogParamsOptimizedMESA_elements[] = {
    {NULL, {STATE_INTERNAL, STATE_FOG_PARAMS_OPTIMIZED}, SWIZZLE_XYZW},
 };
@@ -284,8 +276,6 @@ static const struct gl_builtin_uniform_desc _mesa_builtin_uniform_desc[] = {
    STATEVAR(gl_NormalMatrix),
    STATEVAR(gl_NormalScale),
 
-   STATEVAR(gl_BumpRotMatrix0MESA),
-   STATEVAR(gl_BumpRotMatrix1MESA),
    STATEVAR(gl_FogParamsOptimizedMESA),
    STATEVAR(gl_CurrentAttribVertMESA),
    STATEVAR(gl_CurrentAttribFragMESA),
@@ -761,8 +751,6 @@ builtin_variable_generator::generate_uniforms()
       add_uniform(mat4_t, "gl_ModelViewProjectionMatrixInverseTranspose");
       add_uniform(float_t, "gl_NormalScale");
       add_uniform(type("gl_LightModelParameters"), "gl_LightModel");
-      add_uniform(vec2_t, "gl_BumpRotMatrix0MESA");
-      add_uniform(vec2_t, "gl_BumpRotMatrix1MESA");
       add_uniform(vec4_t, "gl_FogParamsOptimizedMESA");
 
       const glsl_type *const mat4_array_type =
diff --git a/mesalib/src/glsl/glsl_parser.yy b/mesalib/src/glsl/glsl_parser.yy
index 240995901..b9897498f 100644
--- a/mesalib/src/glsl/glsl_parser.yy
+++ b/mesalib/src/glsl/glsl_parser.yy
@@ -1395,6 +1395,22 @@ layout_qualifier_id:
          }
       }
 
+      if (state->stage == MESA_SHADER_GEOMETRY) {
+         if (match_layout_qualifier("stream", $1, state) == 0 &&
+             state->check_explicit_attrib_stream_allowed(& @3)) {
+            $$.flags.q.stream = 1;
+
+            if ($3 < 0) {
+               _mesa_glsl_error(& @3, state,
+                                "invalid stream %d specified", $3);
+               YYERROR;
+            } else {
+               $$.flags.q.explicit_stream = 1;
+               $$.stream = $3;
+            }
+         }
+      }
+
       static const char * const local_size_qualifiers[3] = {
          "local_size_x",
          "local_size_y",
@@ -1693,6 +1709,20 @@ storage_qualifier:
    {
       memset(& $$, 0, sizeof($$));
       $$.flags.q.out = 1;
+
+      if (state->stage == MESA_SHADER_GEOMETRY &&
+          state->has_explicit_attrib_stream()) {
+         /* Section 4.3.8.2 (Output Layout Qualifiers) of the GLSL 4.00
+          * spec says:
+          *
+          *     "If the block or variable is declared with the stream
+          *     identifier, it is associated with the specified stream;
+          *     otherwise, it is associated with the current default stream."
+          */
+          $$.flags.q.stream = 1;
+          $$.flags.q.explicit_stream = 0;
+          $$.stream = state->out_qualifier->stream;
+      }
    }
    | UNIFORM
    {
@@ -2361,6 +2391,18 @@ interface_block:
       if (!block->layout.merge_qualifier(& @1, state, $1)) {
          YYERROR;
       }
+
+      foreach_list_typed (ast_declarator_list, member, link, &block->declarations) {
+         ast_type_qualifier& qualifier = member->type->qualifier;
+         if (qualifier.flags.q.stream && qualifier.stream != block->layout.stream) {
+               _mesa_glsl_error(& @1, state,
+                             "stream layout qualifier on "
+                             "interface block member does not match "
+                             "the interface block (%d vs %d)",
+                             qualifier.stream, block->layout.stream);
+               YYERROR;
+         }
+      }
       $$ = block;
    }
    ;
@@ -2434,6 +2476,14 @@ basic_interface_block:
 
       block->layout.flags.i |= block_interface_qualifier;
 
+      if (state->stage == MESA_SHADER_GEOMETRY &&
+          state->has_explicit_attrib_stream()) {
+         /* Assign global layout's stream value. */
+         block->layout.flags.q.stream = 1;
+         block->layout.flags.q.explicit_stream = 0;
+         block->layout.stream = state->out_qualifier->stream;
+      }
+
       foreach_list_typed (ast_declarator_list, member, link, &block->declarations) {
          ast_type_qualifier& qualifier = member->type->qualifier;
          if ((qualifier.flags.i & interface_type_mask) == 0) {
@@ -2576,6 +2626,9 @@ layout_defaults:
          }
          if (!state->out_qualifier->merge_qualifier(& @1, state, $1))
             YYERROR;
+
+         /* Allow future assigments of global out's stream id value */
+         state->out_qualifier->flags.q.explicit_stream = 0;
       }
       $$ = NULL;
    }
diff --git a/mesalib/src/glsl/glsl_parser_extras.h b/mesalib/src/glsl/glsl_parser_extras.h
index aa4a114e4..a0af965fd 100644
--- a/mesalib/src/glsl/glsl_parser_extras.h
+++ b/mesalib/src/glsl/glsl_parser_extras.h
@@ -119,6 +119,19 @@ struct _mesa_glsl_parse_state {
       return check_version(130, 300, locp, "bit-wise operations are forbidden");
    }
 
+   bool check_explicit_attrib_stream_allowed(YYLTYPE *locp)
+   {
+      if (!this->has_explicit_attrib_stream()) {
+         const char *const requirement = "GL_ARB_gpu_shader5 extension or GLSL 400";
+
+         _mesa_glsl_error(locp, this, "explicit stream requires %s",
+                          requirement);
+         return false;
+      }
+
+      return true;
+   }
+
    bool check_explicit_attrib_location_allowed(YYLTYPE *locp,
                                                const ir_variable *var)
    {
@@ -166,6 +179,11 @@ struct _mesa_glsl_parse_state {
       return true;
    }
 
+   bool has_explicit_attrib_stream() const
+   {
+      return ARB_gpu_shader5_enable || is_version(400, 0);
+   }
+
    bool has_explicit_attrib_location() const
    {
       return ARB_explicit_attrib_location_enable || is_version(330, 300);
diff --git a/mesalib/src/glsl/glsl_types.h b/mesalib/src/glsl/glsl_types.h
index f6d4a02ab..0b63d4850 100644
--- a/mesalib/src/glsl/glsl_types.h
+++ b/mesalib/src/glsl/glsl_types.h
@@ -671,6 +671,12 @@ struct glsl_struct_field {
     * in ir_variable::sample). 0 otherwise.
     */
    unsigned sample:1;
+
+   /**
+    * For interface blocks, it has a value if this variable uses multiple vertex
+    * streams (as in ir_variable::stream). -1 otherwise.
+    */
+   int stream;
 };
 
 static inline unsigned int
diff --git a/mesalib/src/glsl/ir.h b/mesalib/src/glsl/ir.h
index b4e52d3d0..d5239d4de 100644
--- a/mesalib/src/glsl/ir.h
+++ b/mesalib/src/glsl/ir.h
@@ -706,6 +706,11 @@ public:
       int location;
 
       /**
+       * Vertex stream output identifier.
+       */
+      unsigned stream;
+
+      /**
        * output index for dual source blending.
        */
       int index;
@@ -2154,9 +2159,11 @@ private:
  */
 class ir_emit_vertex : public ir_instruction {
 public:
-   ir_emit_vertex()
-      : ir_instruction(ir_type_emit_vertex)
+   ir_emit_vertex(ir_rvalue *stream)
+      : ir_instruction(ir_type_emit_vertex),
+        stream(stream)
    {
+      assert(stream);
    }
 
    virtual void accept(ir_visitor *v)
@@ -2164,12 +2171,19 @@ public:
       v->visit(this);
    }
 
-   virtual ir_emit_vertex *clone(void *mem_ctx, struct hash_table *) const
+   virtual ir_emit_vertex *clone(void *mem_ctx, struct hash_table *ht) const
    {
-      return new(mem_ctx) ir_emit_vertex();
+      return new(mem_ctx) ir_emit_vertex(this->stream->clone(mem_ctx, ht));
    }
 
    virtual ir_visitor_status accept(ir_hierarchical_visitor *);
+
+   int stream_id() const
+   {
+      return stream->as_constant()->value.i[0];
+   }
+
+   ir_rvalue *stream;
 };
 
 /**
@@ -2178,9 +2192,11 @@ public:
  */
 class ir_end_primitive : public ir_instruction {
 public:
-   ir_end_primitive()
-      : ir_instruction(ir_type_end_primitive)
+   ir_end_primitive(ir_rvalue *stream)
+      : ir_instruction(ir_type_end_primitive),
+        stream(stream)
    {
+      assert(stream);
    }
 
    virtual void accept(ir_visitor *v)
@@ -2188,12 +2204,19 @@ public:
       v->visit(this);
    }
 
-   virtual ir_end_primitive *clone(void *mem_ctx, struct hash_table *) const
+   virtual ir_end_primitive *clone(void *mem_ctx, struct hash_table *ht) const
    {
-      return new(mem_ctx) ir_end_primitive();
+      return new(mem_ctx) ir_end_primitive(this->stream->clone(mem_ctx, ht));
    }
 
    virtual ir_visitor_status accept(ir_hierarchical_visitor *);
+
+   int stream_id() const
+   {
+      return stream->as_constant()->value.i[0];
+   }
+
+   ir_rvalue *stream;
 };
 
 /*@}*/
diff --git a/mesalib/src/glsl/ir_hierarchical_visitor.cpp b/mesalib/src/glsl/ir_hierarchical_visitor.cpp
index 2e606dda4..d3c00ecdb 100644
--- a/mesalib/src/glsl/ir_hierarchical_visitor.cpp
+++ b/mesalib/src/glsl/ir_hierarchical_visitor.cpp
@@ -69,24 +69,6 @@ ir_hierarchical_visitor::visit(ir_loop_jump *ir)
 }
 
 ir_visitor_status
-ir_hierarchical_visitor::visit(ir_emit_vertex *ir)
-{
-   if (this->callback != NULL)
-      this->callback(ir, this->data);
-
-   return visit_continue;
-}
-
-ir_visitor_status
-ir_hierarchical_visitor::visit(ir_end_primitive *ir)
-{
-   if (this->callback != NULL)
-      this->callback(ir, this->data);
-
-   return visit_continue;
-}
-
-ir_visitor_status
 ir_hierarchical_visitor::visit(ir_dereference_variable *ir)
 {
    if (this->callback != NULL)
@@ -303,6 +285,38 @@ ir_hierarchical_visitor::visit_leave(ir_if *ir)
    return visit_continue;
 }
 
+ir_visitor_status
+ir_hierarchical_visitor::visit_enter(ir_emit_vertex *ir)
+{
+   if (this->callback != NULL)
+      this->callback(ir, this->data);
+
+   return visit_continue;
+}
+
+ir_visitor_status
+ir_hierarchical_visitor::visit_leave(ir_emit_vertex *ir)
+{
+   (void) ir;
+   return visit_continue;
+}
+
+ir_visitor_status
+ir_hierarchical_visitor::visit_enter(ir_end_primitive *ir)
+{
+   if (this->callback != NULL)
+      this->callback(ir, this->data);
+
+   return visit_continue;
+}
+
+ir_visitor_status
+ir_hierarchical_visitor::visit_leave(ir_end_primitive *ir)
+{
+   (void) ir;
+   return visit_continue;
+}
+
 void
 ir_hierarchical_visitor::run(exec_list *instructions)
 {
diff --git a/mesalib/src/glsl/ir_hierarchical_visitor.h b/mesalib/src/glsl/ir_hierarchical_visitor.h
index 647d2e002..bc89a04d8 100644
--- a/mesalib/src/glsl/ir_hierarchical_visitor.h
+++ b/mesalib/src/glsl/ir_hierarchical_visitor.h
@@ -87,8 +87,6 @@ public:
    virtual ir_visitor_status visit(class ir_variable *);
    virtual ir_visitor_status visit(class ir_constant *);
    virtual ir_visitor_status visit(class ir_loop_jump *);
-   virtual ir_visitor_status visit(class ir_emit_vertex *);
-   virtual ir_visitor_status visit(class ir_end_primitive *);
 
    /**
     * ir_dereference_variable isn't technically a leaf, but it is treated as a
@@ -137,6 +135,10 @@ public:
    virtual ir_visitor_status visit_leave(class ir_discard *);
    virtual ir_visitor_status visit_enter(class ir_if *);
    virtual ir_visitor_status visit_leave(class ir_if *);
+   virtual ir_visitor_status visit_enter(class ir_emit_vertex *);
+   virtual ir_visitor_status visit_leave(class ir_emit_vertex *);
+   virtual ir_visitor_status visit_enter(class ir_end_primitive *);
+   virtual ir_visitor_status visit_leave(class ir_end_primitive *);
    /*@}*/
 
 
diff --git a/mesalib/src/glsl/ir_hv_accept.cpp b/mesalib/src/glsl/ir_hv_accept.cpp
index 2a1f70e5b..3ca7a5887 100644
--- a/mesalib/src/glsl/ir_hv_accept.cpp
+++ b/mesalib/src/glsl/ir_hv_accept.cpp
@@ -405,12 +405,28 @@ ir_if::accept(ir_hierarchical_visitor *v)
 ir_visitor_status
 ir_emit_vertex::accept(ir_hierarchical_visitor *v)
 {
-   return v->visit(this);
+   ir_visitor_status s = v->visit_enter(this);
+   if (s != visit_continue)
+      return (s == visit_continue_with_parent) ? visit_continue : s;
+
+   s = this->stream->accept(v);
+   if (s != visit_continue)
+      return (s == visit_continue_with_parent) ? visit_continue : s;
+
+   return (s == visit_stop) ? s : v->visit_leave(this);
 }
 
 
 ir_visitor_status
 ir_end_primitive::accept(ir_hierarchical_visitor *v)
 {
-   return v->visit(this);
+   ir_visitor_status s = v->visit_enter(this);
+   if (s != visit_continue)
+      return (s == visit_continue_with_parent) ? visit_continue : s;
+
+   s = this->stream->accept(v);
+   if (s != visit_continue)
+      return (s == visit_continue_with_parent) ? visit_continue : s;
+
+   return (s == visit_stop) ? s : v->visit_leave(this);
 }
diff --git a/mesalib/src/glsl/ir_print_visitor.cpp b/mesalib/src/glsl/ir_print_visitor.cpp
index c4a6f9c9f..72ad4223f 100644
--- a/mesalib/src/glsl/ir_print_visitor.cpp
+++ b/mesalib/src/glsl/ir_print_visitor.cpp
@@ -169,11 +169,14 @@ void ir_print_visitor::visit(ir_variable *ir)
                                 "in ", "out ", "inout ",
 			        "const_in ", "sys ", "temporary " };
    STATIC_ASSERT(ARRAY_SIZE(mode) == ir_var_mode_count);
+   const char *const stream [] = {"", "stream1 ", "stream2 ", "stream3 "};
    const char *const interp[] = { "", "smooth", "flat", "noperspective" };
    STATIC_ASSERT(ARRAY_SIZE(interp) == INTERP_QUALIFIER_COUNT);
 
-   fprintf(f, "(%s%s%s%s%s) ",
-	  cent, samp, inv, mode[ir->data.mode], interp[ir->data.interpolation]);
+   fprintf(f, "(%s%s%s%s%s%s) ",
+           cent, samp, inv, mode[ir->data.mode],
+           stream[ir->data.stream],
+           interp[ir->data.interpolation]);
 
    print_type(f, ir->type);
    fprintf(f, " %s)", unique_name(ir));
@@ -560,13 +563,18 @@ ir_print_visitor::visit(ir_loop_jump *ir)
 }
 
 void
-ir_print_visitor::visit(ir_emit_vertex *)
+ir_print_visitor::visit(ir_emit_vertex *ir)
 {
-   fprintf(f, "(emit-vertex)");
+   fprintf(f, "(emit-vertex ");
+   ir->stream->accept(this);
+   fprintf(f, ")\n");
 }
 
 void
-ir_print_visitor::visit(ir_end_primitive *)
+ir_print_visitor::visit(ir_end_primitive *ir)
 {
-   fprintf(f, "(end-primitive)");
+   fprintf(f, "(end-primitive ");
+   ir->stream->accept(this);
+   fprintf(f, ")\n");
+
 }
diff --git a/mesalib/src/glsl/ir_reader.cpp b/mesalib/src/glsl/ir_reader.cpp
index 28923f3b8..a178c82b5 100644
--- a/mesalib/src/glsl/ir_reader.cpp
+++ b/mesalib/src/glsl/ir_reader.cpp
@@ -437,6 +437,12 @@ ir_reader::read_declaration(s_expression *expr)
 	 var->data.mode = ir_var_function_inout;
       } else if (strcmp(qualifier->value(), "temporary") == 0) {
 	 var->data.mode = ir_var_temporary;
+      } else if (strcmp(qualifier->value(), "stream1") == 0) {
+	 var->data.stream = 1;
+      } else if (strcmp(qualifier->value(), "stream2") == 0) {
+	 var->data.stream = 2;
+      } else if (strcmp(qualifier->value(), "stream3") == 0) {
+	 var->data.stream = 3;
       } else if (strcmp(qualifier->value(), "smooth") == 0) {
 	 var->data.interpolation = INTERP_QUALIFIER_SMOOTH;
       } else if (strcmp(qualifier->value(), "flat") == 0) {
@@ -1109,10 +1115,17 @@ ir_reader::read_texture(s_expression *expr)
 ir_emit_vertex *
 ir_reader::read_emit_vertex(s_expression *expr)
 {
-   s_pattern pat[] = { "emit-vertex" };
+   s_expression *s_stream = NULL;
+
+   s_pattern pat[] = { "emit-vertex", s_stream };
 
    if (MATCH(expr, pat)) {
-      return new(mem_ctx) ir_emit_vertex();
+      ir_rvalue *stream = read_dereference(s_stream);
+      if (stream == NULL) {
+         ir_read_error(NULL, "when reading stream info in emit-vertex");
+         return NULL;
+      }
+      return new(mem_ctx) ir_emit_vertex(stream);
    }
    ir_read_error(NULL, "when reading emit-vertex");
    return NULL;
@@ -1121,10 +1134,17 @@ ir_reader::read_emit_vertex(s_expression *expr)
 ir_end_primitive *
 ir_reader::read_end_primitive(s_expression *expr)
 {
-   s_pattern pat[] = { "end-primitive" };
+   s_expression *s_stream = NULL;
+
+   s_pattern pat[] = { "end-primitive", s_stream };
 
    if (MATCH(expr, pat)) {
-      return new(mem_ctx) ir_end_primitive();
+      ir_rvalue *stream = read_dereference(s_stream);
+      if (stream == NULL) {
+         ir_read_error(NULL, "when reading stream info in end-primitive");
+         return NULL;
+      }
+      return new(mem_ctx) ir_end_primitive(stream);
    }
    ir_read_error(NULL, "when reading end-primitive");
    return NULL;
diff --git a/mesalib/src/glsl/ir_rvalue_visitor.cpp b/mesalib/src/glsl/ir_rvalue_visitor.cpp
index fcbe9448d..0370170b3 100644
--- a/mesalib/src/glsl/ir_rvalue_visitor.cpp
+++ b/mesalib/src/glsl/ir_rvalue_visitor.cpp
@@ -149,6 +149,19 @@ ir_rvalue_base_visitor::rvalue_visit(ir_if *ir)
    return visit_continue;
 }
 
+ir_visitor_status
+ir_rvalue_base_visitor::rvalue_visit(ir_emit_vertex *ir)
+{
+   handle_rvalue(&ir->stream);
+   return visit_continue;
+}
+
+ir_visitor_status
+ir_rvalue_base_visitor::rvalue_visit(ir_end_primitive *ir)
+{
+   handle_rvalue(&ir->stream);
+   return visit_continue;
+}
 
 ir_visitor_status
 ir_rvalue_visitor::visit_leave(ir_expression *ir)
@@ -205,6 +218,18 @@ ir_rvalue_visitor::visit_leave(ir_if *ir)
 }
 
 ir_visitor_status
+ir_rvalue_visitor::visit_leave(ir_emit_vertex *ir)
+{
+   return rvalue_visit(ir);
+}
+
+ir_visitor_status
+ir_rvalue_visitor::visit_leave(ir_end_primitive *ir)
+{
+   return rvalue_visit(ir);
+}
+
+ir_visitor_status
 ir_rvalue_enter_visitor::visit_enter(ir_expression *ir)
 {
    return rvalue_visit(ir);
@@ -257,3 +282,15 @@ ir_rvalue_enter_visitor::visit_enter(ir_if *ir)
 {
    return rvalue_visit(ir);
 }
+
+ir_visitor_status
+ir_rvalue_enter_visitor::visit_enter(ir_emit_vertex *ir)
+{
+   return rvalue_visit(ir);
+}
+
+ir_visitor_status
+ir_rvalue_enter_visitor::visit_enter(ir_end_primitive *ir)
+{
+   return rvalue_visit(ir);
+}
diff --git a/mesalib/src/glsl/ir_rvalue_visitor.h b/mesalib/src/glsl/ir_rvalue_visitor.h
index 2179fa5a8..04ec0fa39 100644
--- a/mesalib/src/glsl/ir_rvalue_visitor.h
+++ b/mesalib/src/glsl/ir_rvalue_visitor.h
@@ -41,6 +41,8 @@ public:
    ir_visitor_status rvalue_visit(ir_return *);
    ir_visitor_status rvalue_visit(ir_swizzle *);
    ir_visitor_status rvalue_visit(ir_texture *);
+   ir_visitor_status rvalue_visit(ir_emit_vertex *);
+   ir_visitor_status rvalue_visit(ir_end_primitive *);
 
    virtual void handle_rvalue(ir_rvalue **rvalue) = 0;
 };
@@ -57,6 +59,8 @@ public:
    virtual ir_visitor_status visit_leave(ir_return *);
    virtual ir_visitor_status visit_leave(ir_swizzle *);
    virtual ir_visitor_status visit_leave(ir_texture *);
+   virtual ir_visitor_status visit_leave(ir_emit_vertex *);
+   virtual ir_visitor_status visit_leave(ir_end_primitive *);
 };
 
 class ir_rvalue_enter_visitor : public ir_rvalue_base_visitor {
@@ -71,4 +75,6 @@ public:
    virtual ir_visitor_status visit_enter(ir_return *);
    virtual ir_visitor_status visit_enter(ir_swizzle *);
    virtual ir_visitor_status visit_enter(ir_texture *);
+   virtual ir_visitor_status visit_enter(ir_emit_vertex *);
+   virtual ir_visitor_status visit_enter(ir_end_primitive *);
 };
diff --git a/mesalib/src/glsl/link_atomics.cpp b/mesalib/src/glsl/link_atomics.cpp
index d92cdb117..fbe4e7364 100644
--- a/mesalib/src/glsl/link_atomics.cpp
+++ b/mesalib/src/glsl/link_atomics.cpp
@@ -54,9 +54,18 @@ namespace {
 
       void push_back(unsigned id, ir_variable *var)
       {
-         counters = (active_atomic_counter *)
-            realloc(counters, sizeof(active_atomic_counter) * (num_counters + 1));
+         active_atomic_counter *new_counters;
 
+         new_counters = (active_atomic_counter *)
+            realloc(counters, sizeof(active_atomic_counter) *
+                    (num_counters + 1));
+
+         if (new_counters == NULL) {
+            _mesa_error_no_memory(__func__);
+            return;
+         }
+
+         counters = new_counters;
          counters[num_counters].id = id;
          counters[num_counters].var = var;
          num_counters++;
diff --git a/mesalib/src/glsl/link_uniform_blocks.cpp b/mesalib/src/glsl/link_uniform_blocks.cpp
index 1a0e64318..53a18c934 100644
--- a/mesalib/src/glsl/link_uniform_blocks.cpp
+++ b/mesalib/src/glsl/link_uniform_blocks.cpp
@@ -170,6 +170,12 @@ link_uniform_blocks(void *mem_ctx,
    struct hash_table *block_hash =
       _mesa_hash_table_create(mem_ctx, _mesa_key_string_equal);
 
+   if (block_hash == NULL) {
+      _mesa_error_no_memory(__func__);
+      linker_error(prog, "out of memory\n");
+      return 0;
+   }
+
    /* Determine which uniform blocks are active.
     */
    link_uniform_block_active_visitor v(mem_ctx, block_hash, prog);
diff --git a/mesalib/src/glsl/link_varyings.cpp b/mesalib/src/glsl/link_varyings.cpp
index 686329857..520a51a27 100644
--- a/mesalib/src/glsl/link_varyings.cpp
+++ b/mesalib/src/glsl/link_varyings.cpp
@@ -291,6 +291,7 @@ tfeedback_decl::init(struct gl_context *ctx, const void *mem_ctx,
    this->skip_components = 0;
    this->next_buffer_separator = false;
    this->matched_candidate = NULL;
+   this->stream_id = 0;
 
    if (ctx->Extensions.ARB_transform_feedback3) {
       /* Parse gl_NextBuffer. */
@@ -355,8 +356,8 @@ tfeedback_decl::is_same(const tfeedback_decl &x, const tfeedback_decl &y)
 
 
 /**
- * Assign a location for this tfeedback_decl object based on the transform
- * feedback candidate found by find_candidate.
+ * Assign a location and stream ID for this tfeedback_decl object based on the
+ * transform feedback candidate found by find_candidate.
  *
  * If an error occurs, the error is reported through linker_error() and false
  * is returned.
@@ -437,6 +438,11 @@ tfeedback_decl::assign_location(struct gl_context *ctx,
       return false;
    }
 
+   /* Only transform feedback varyings can be assigned to non-zero streams,
+    * so assign the stream id here.
+    */
+   this->stream_id = this->matched_candidate->toplevel_var->data.stream;
+
    return true;
 }
 
@@ -495,6 +501,7 @@ tfeedback_decl::store(struct gl_context *ctx, struct gl_shader_program *prog,
       info->Outputs[info->NumOutputs].ComponentOffset = location_frac;
       info->Outputs[info->NumOutputs].OutputRegister = location;
       info->Outputs[info->NumOutputs].NumComponents = output_size;
+      info->Outputs[info->NumOutputs].StreamId = stream_id;
       info->Outputs[info->NumOutputs].OutputBuffer = buffer;
       info->Outputs[info->NumOutputs].DstOffset = info->BufferStride[buffer];
       ++info->NumOutputs;
@@ -628,10 +635,27 @@ store_tfeedback_info(struct gl_context *ctx, struct gl_shader_program *prog,
    }
    else {
       /* GL_INVERLEAVED_ATTRIBS */
+      int buffer_stream_id = -1;
       for (unsigned i = 0; i < num_tfeedback_decls; ++i) {
          if (tfeedback_decls[i].is_next_buffer_separator()) {
             num_buffers++;
+            buffer_stream_id = -1;
             continue;
+         } else if (buffer_stream_id == -1)  {
+            /* First varying writing to this buffer: remember its stream */
+            buffer_stream_id = (int) tfeedback_decls[i].get_stream_id();
+         } else if (buffer_stream_id !=
+                    (int) tfeedback_decls[i].get_stream_id()) {
+            /* Varying writes to the same buffer from a different stream */
+            linker_error(prog,
+                         "Transform feedback can't capture varyings belonging "
+                         "to different vertex streams in a single buffer. "
+                         "Varying %s writes to buffer from stream %u, other "
+                         "varyings in the same buffer write from stream %u.",
+                         tfeedback_decls[i].name(),
+                         tfeedback_decls[i].get_stream_id(),
+                         buffer_stream_id);
+            return false;
          }
 
          if (!tfeedback_decls[i].store(ctx, prog,
@@ -1323,6 +1347,11 @@ assign_varying_locations(struct gl_context *ctx,
              (output_var->data.mode != ir_var_shader_out))
             continue;
 
+         /* Only geometry shaders can use non-zero streams */
+         assert(output_var->data.stream == 0 ||
+                (output_var->data.stream < MAX_VERTEX_STREAMS &&
+                 producer->Stage == MESA_SHADER_GEOMETRY));
+
          tfeedback_candidate_generator g(mem_ctx, tfeedback_candidates);
          g.process(output_var);
 
@@ -1338,6 +1367,14 @@ assign_varying_locations(struct gl_context *ctx,
          if (input_var || (prog->SeparateShader && consumer == NULL)) {
             matches.record(output_var, input_var);
          }
+
+         /* Only stream 0 outputs can be consumed in the next stage */
+         if (input_var && output_var->data.stream != 0) {
+            linker_error(prog, "output %s is assigned to stream=%d but "
+                         "is linked to an input, which requires stream=0",
+                         output_var->name, output_var->data.stream);
+            return false;
+         }
       }
    } else {
       /* If there's no producer stage, then this must be a separable program.
diff --git a/mesalib/src/glsl/link_varyings.h b/mesalib/src/glsl/link_varyings.h
index 6fa268176..afc16a8ba 100644
--- a/mesalib/src/glsl/link_varyings.h
+++ b/mesalib/src/glsl/link_varyings.h
@@ -112,6 +112,16 @@ public:
       return !this->next_buffer_separator && !this->skip_components;
    }
 
+   const char *name() const
+   {
+      return this->orig_name;
+   }
+
+   unsigned get_stream_id() const
+   {
+      return this->stream_id;
+   }
+
    /**
     * The total number of varying components taken up by this variable.  Only
     * valid if assign_location() has been called.
@@ -210,6 +220,13 @@ private:
     * data structure that was found.  Otherwise NULL.
     */
    const tfeedback_candidate *matched_candidate;
+
+   /**
+    * StreamId assigned to this varying (defaults to 0). Can only be set to
+    * values other than 0 in geometry shaders that use the stream layout
+    * modifier. Accepted values must be in the range [0, MAX_VERTEX_STREAMS-1].
+    */
+   unsigned stream_id;
 };
 
 
diff --git a/mesalib/src/glsl/linker.cpp b/mesalib/src/glsl/linker.cpp
index 0b6a71679..3036ebcb3 100644
--- a/mesalib/src/glsl/linker.cpp
+++ b/mesalib/src/glsl/linker.cpp
@@ -250,31 +250,100 @@ public:
    }
 };
 
-
 /**
- * Visitor that determines whether or not a shader uses ir_end_primitive.
+ * Visitor that determines the highest stream id to which a (geometry) shader
+ * emits vertices. It also checks whether End{Stream}Primitive is ever called.
  */
-class find_end_primitive_visitor : public ir_hierarchical_visitor {
+class find_emit_vertex_visitor : public ir_hierarchical_visitor {
 public:
-   find_end_primitive_visitor()
-      : found(false)
+   find_emit_vertex_visitor(int max_allowed)
+      : max_stream_allowed(max_allowed),
+        invalid_stream_id(0),
+        invalid_stream_id_from_emit_vertex(false),
+        end_primitive_found(false),
+        uses_non_zero_stream(false)
    {
       /* empty */
    }
 
-   virtual ir_visitor_status visit(ir_end_primitive *)
+   virtual ir_visitor_status visit_leave(ir_emit_vertex *ir)
    {
-      found = true;
-      return visit_stop;
+      int stream_id = ir->stream_id();
+
+      if (stream_id < 0) {
+         invalid_stream_id = stream_id;
+         invalid_stream_id_from_emit_vertex = true;
+         return visit_stop;
+      }
+
+      if (stream_id > max_stream_allowed) {
+         invalid_stream_id = stream_id;
+         invalid_stream_id_from_emit_vertex = true;
+         return visit_stop;
+      }
+
+      if (stream_id != 0)
+         uses_non_zero_stream = true;
+
+      return visit_continue;
    }
 
-   bool end_primitive_found()
+   virtual ir_visitor_status visit_leave(ir_end_primitive *ir)
    {
-      return found;
+      end_primitive_found = true;
+
+      int stream_id = ir->stream_id();
+
+      if (stream_id < 0) {
+         invalid_stream_id = stream_id;
+         invalid_stream_id_from_emit_vertex = false;
+         return visit_stop;
+      }
+
+      if (stream_id > max_stream_allowed) {
+         invalid_stream_id = stream_id;
+         invalid_stream_id_from_emit_vertex = false;
+         return visit_stop;
+      }
+
+      if (stream_id != 0)
+         uses_non_zero_stream = true;
+
+      return visit_continue;
+   }
+
+   bool error()
+   {
+      return invalid_stream_id != 0;
+   }
+
+   const char *error_func()
+   {
+      return invalid_stream_id_from_emit_vertex ?
+         "EmitStreamVertex" : "EndStreamPrimitive";
+   }
+
+   int error_stream()
+   {
+      return invalid_stream_id;
+   }
+
+   bool uses_streams()
+   {
+      return uses_non_zero_stream;
+   }
+
+   bool uses_end_primitive()
+   {
+      return end_primitive_found;
    }
 
 private:
-   bool found;
+   int max_stream_allowed;
+   int invalid_stream_id;
+   bool invalid_stream_id_from_emit_vertex;
+   bool end_primitive_found;
+   bool uses_non_zero_stream;
 };
 
 } /* anonymous namespace */
@@ -551,10 +620,58 @@ validate_geometry_shader_executable(struct gl_shader_program *prog,
 
    analyze_clip_usage(prog, shader, &prog->Geom.UsesClipDistance,
                       &prog->Geom.ClipDistanceArraySize);
+}
+
+/**
+ * Check if geometry shaders emit to non-zero streams and do corresponding
+ * validations.
+ */
+static void
+validate_geometry_shader_emissions(struct gl_context *ctx,
+                                   struct gl_shader_program *prog)
+{
+   if (prog->_LinkedShaders[MESA_SHADER_GEOMETRY] != NULL) {
+      find_emit_vertex_visitor emit_vertex(ctx->Const.MaxVertexStreams - 1);
+      emit_vertex.run(prog->_LinkedShaders[MESA_SHADER_GEOMETRY]->ir);
+      if (emit_vertex.error()) {
+         linker_error(prog, "Invalid call %s(%d). Accepted values for the "
+                      "stream parameter are in the range [0, %d].",
+                      emit_vertex.error_func(),
+                      emit_vertex.error_stream(),
+                      ctx->Const.MaxVertexStreams - 1);
+      }
+      prog->Geom.UsesStreams = emit_vertex.uses_streams();
+      prog->Geom.UsesEndPrimitive = emit_vertex.uses_end_primitive();
 
-   find_end_primitive_visitor end_primitive;
-   end_primitive.run(shader->ir);
-   prog->Geom.UsesEndPrimitive = end_primitive.end_primitive_found();
+      /* From the ARB_gpu_shader5 spec:
+       *
+       *   "Multiple vertex streams are supported only if the output primitive
+       *    type is declared to be "points".  A program will fail to link if it
+       *    contains a geometry shader calling EmitStreamVertex() or
+       *    EndStreamPrimitive() if its output primitive type is not "points".
+       *
+       * However, in the same spec:
+       *
+       *   "The function EmitVertex() is equivalent to calling EmitStreamVertex()
+       *    with <stream> set to zero."
+       *
+       * And:
+       *
+       *   "The function EndPrimitive() is equivalent to calling
+       *    EndStreamPrimitive() with <stream> set to zero."
+       *
+       * Since we can call EmitVertex() and EndPrimitive() when we output
+       * primitives other than points, calling EmitStreamVertex(0) or
+       * EmitEndPrimitive(0) should not produce errors. This it also what Nvidia
+       * does. Currently we only set prog->Geom.UsesStreams to TRUE when
+       * EmitStreamVertex() or EmitEndPrimitive() are called with a non-zero
+       * stream.
+       */
+      if (prog->Geom.UsesStreams && prog->Geom.OutputType != GL_POINTS) {
+         linker_error(prog, "EmitStreamVertex(n) and EndStreamPrimitive(n) "
+                      "with n>0 requires point output");
+      }
+   }
 }
 
 
@@ -1479,6 +1596,8 @@ link_intrastage_shaders(void *mem_ctx,
    const unsigned num_uniform_blocks =
       link_uniform_blocks(mem_ctx, prog, shader_list, num_shaders,
                           &uniform_blocks);
+   if (!prog->LinkStatus)
+      return NULL;
 
    /* Check that there is only a single definition of each function signature
     * across all shaders.
@@ -2556,6 +2675,9 @@ link_shaders(struct gl_context *ctx, struct gl_shader_program *prog)
 	 ;
    }
 
+   /* Check and validate stream emissions in geometry shaders */
+   validate_geometry_shader_emissions(ctx, prog);
+
    /* Mark all generic shader inputs and outputs as unpaired. */
    for (unsigned i = MESA_SHADER_VERTEX; i <= MESA_SHADER_FRAGMENT; i++) {
       if (prog->_LinkedShaders[i] != NULL) {
diff --git a/mesalib/src/glsl/lower_output_reads.cpp b/mesalib/src/glsl/lower_output_reads.cpp
index afe17766b..1ee815d5e 100644
--- a/mesalib/src/glsl/lower_output_reads.cpp
+++ b/mesalib/src/glsl/lower_output_reads.cpp
@@ -52,7 +52,7 @@ public:
    output_read_remover();
    ~output_read_remover();
    virtual ir_visitor_status visit(class ir_dereference_variable *);
-   virtual ir_visitor_status visit(class ir_emit_vertex *);
+   virtual ir_visitor_status visit_leave(class ir_emit_vertex *);
    virtual ir_visitor_status visit_leave(class ir_return *);
    virtual ir_visitor_status visit_leave(class ir_function_signature *);
 };
@@ -148,7 +148,7 @@ output_read_remover::visit_leave(ir_return *ir)
 }
 
 ir_visitor_status
-output_read_remover::visit(ir_emit_vertex *ir)
+output_read_remover::visit_leave(ir_emit_vertex *ir)
 {
    hash_table_call_foreach(replacements, emit_return_copy, ir);
    hash_table_clear(replacements);
diff --git a/mesalib/src/glsl/lower_packed_varyings.cpp b/mesalib/src/glsl/lower_packed_varyings.cpp
index e8654748f..eda56a97b 100644
--- a/mesalib/src/glsl/lower_packed_varyings.cpp
+++ b/mesalib/src/glsl/lower_packed_varyings.cpp
@@ -613,7 +613,7 @@ public:
    explicit lower_packed_varyings_gs_splicer(void *mem_ctx,
                                              const exec_list *instructions);
 
-   virtual ir_visitor_status visit(ir_emit_vertex *ev);
+   virtual ir_visitor_status visit_leave(ir_emit_vertex *ev);
 
 private:
    /**
@@ -637,7 +637,7 @@ lower_packed_varyings_gs_splicer::lower_packed_varyings_gs_splicer(
 
 
 ir_visitor_status
-lower_packed_varyings_gs_splicer::visit(ir_emit_vertex *ev)
+lower_packed_varyings_gs_splicer::visit_leave(ir_emit_vertex *ev)
 {
    foreach_list(node, this->instructions) {
       ir_instruction *ir = (ir_instruction *) node;
diff --git a/mesalib/src/glsl/opt_dead_code_local.cpp b/mesalib/src/glsl/opt_dead_code_local.cpp
index c27c526f9..88895fb0e 100644
--- a/mesalib/src/glsl/opt_dead_code_local.cpp
+++ b/mesalib/src/glsl/opt_dead_code_local.cpp
@@ -114,7 +114,7 @@ public:
       return visit_continue_with_parent;
    }
 
-   virtual ir_visitor_status visit(ir_emit_vertex *)
+   virtual ir_visitor_status visit_leave(ir_emit_vertex *)
    {
       /* For the purpose of dead code elimination, emitting a vertex counts as
        * "reading" all of the currently assigned output variables.
diff --git a/mesalib/src/loader/Makefile.am b/mesalib/src/loader/Makefile.am
index bddf7ac35..ae8a84492 100644
--- a/mesalib/src/loader/Makefile.am
+++ b/mesalib/src/loader/Makefile.am
@@ -29,6 +29,25 @@ libloader_la_CPPFLAGS = \
 	$(VISIBILITY_CFLAGS) \
 	$(LIBUDEV_CFLAGS)
 
+libloader_la_SOURCES = $(LOADER_C_FILES)
+libloader_la_LIBADD =
+
+if NEED_OPENGL_COMMON
+libloader_la_CPPFLAGS += \
+	-I$(top_srcdir)/src/mesa/drivers/dri/common/ \
+	-I$(top_builddir)/src/mesa/drivers/dri/common/ \
+	-I$(top_srcdir)/src/mesa/ \
+	-I$(top_srcdir)/src/mapi/ \
+	-DUSE_DRICONF
+
+libloader_la_SOURCES += \
+	$(top_srcdir)/src/mesa/drivers/dri/common/xmlconfig.c
+
+libloader_la_LIBADD += \
+	-lm \
+	$(EXPAT_LIBS)
+endif
+
 if !HAVE_LIBDRM
 libloader_la_CPPFLAGS += \
 	-D__NOT_HAVE_DRM_H
@@ -36,8 +55,6 @@ else
 libloader_la_CPPFLAGS += \
 	$(LIBDRM_CFLAGS)
 
-libloader_la_LIBADD = \
+libloader_la_LIBADD += \
 	$(LIBDRM_LIBS)
 endif
-
-libloader_la_SOURCES = $(LOADER_C_FILES)
diff --git a/mesalib/src/loader/loader.c b/mesalib/src/loader/loader.c
index 0f262653b..47e1f5874 100644
--- a/mesalib/src/loader/loader.c
+++ b/mesalib/src/loader/loader.c
@@ -70,6 +70,14 @@
 #ifdef HAVE_LIBUDEV
 #include <assert.h>
 #include <dlfcn.h>
+#include <fcntl.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <errno.h>
+#ifdef USE_DRICONF
+#include "xmlconfig.h"
+#include "xmlpool.h"
+#endif
 #endif
 #ifdef HAVE_SYSFS
 #include <sys/stat.h>
@@ -214,6 +222,210 @@ out:
 
    return (*chip_id >= 0);
 }
+
+static char *
+get_render_node_from_id_path_tag(struct udev *udev,
+                                 char *id_path_tag,
+                                 char another_tag)
+{
+   struct udev_device *device;
+   struct udev_enumerate *e;
+   struct udev_list_entry *entry;
+   const char *path, *id_path_tag_tmp;
+   char *path_res;
+   char found = 0;
+   UDEV_SYMBOL(struct udev_enumerate *, udev_enumerate_new,
+               (struct udev *));
+   UDEV_SYMBOL(int, udev_enumerate_add_match_subsystem,
+               (struct udev_enumerate *, const char *));
+   UDEV_SYMBOL(int, udev_enumerate_add_match_sysname,
+               (struct udev_enumerate *, const char *));
+   UDEV_SYMBOL(int, udev_enumerate_scan_devices,
+               (struct udev_enumerate *));
+   UDEV_SYMBOL(struct udev_list_entry *, udev_enumerate_get_list_entry,
+               (struct udev_enumerate *));
+   UDEV_SYMBOL(struct udev_list_entry *, udev_list_entry_get_next,
+               (struct udev_list_entry *));
+   UDEV_SYMBOL(const char *, udev_list_entry_get_name,
+               (struct udev_list_entry *));
+   UDEV_SYMBOL(struct udev_device *, udev_device_new_from_syspath,
+               (struct udev *, const char *));
+   UDEV_SYMBOL(const char *, udev_device_get_property_value,
+               (struct udev_device *, const char *));
+   UDEV_SYMBOL(const char *, udev_device_get_devnode,
+               (struct udev_device *));
+   UDEV_SYMBOL(struct udev_device *, udev_device_unref,
+               (struct udev_device *));
+
+   e = udev_enumerate_new(udev);
+   udev_enumerate_add_match_subsystem(e, "drm");
+   udev_enumerate_add_match_sysname(e, "render*");
+
+   udev_enumerate_scan_devices(e);
+   udev_list_entry_foreach(entry, udev_enumerate_get_list_entry(e)) {
+      path = udev_list_entry_get_name(entry);
+      device = udev_device_new_from_syspath(udev, path);
+      if (!device)
+         continue;
+      id_path_tag_tmp = udev_device_get_property_value(device, "ID_PATH_TAG");
+      if (id_path_tag_tmp) {
+         if ((!another_tag && !strcmp(id_path_tag, id_path_tag_tmp)) ||
+             (another_tag && strcmp(id_path_tag, id_path_tag_tmp))) {
+            found = 1;
+            break;
+         }
+      }
+      udev_device_unref(device);
+   }
+
+   if (found) {
+      path_res = strdup(udev_device_get_devnode(device));
+      udev_device_unref(device);
+      return path_res;
+   }
+   return NULL;
+}
+
+static char *
+get_id_path_tag_from_fd(struct udev *udev, int fd)
+{
+   struct udev_device *device;
+   const char *id_path_tag_tmp;
+   char *id_path_tag;
+   UDEV_SYMBOL(const char *, udev_device_get_property_value,
+               (struct udev_device *, const char *));
+   UDEV_SYMBOL(struct udev_device *, udev_device_unref,
+               (struct udev_device *));
+
+   device = udev_device_new_from_fd(udev, fd);
+   if (!device)
+      return NULL;
+
+   id_path_tag_tmp = udev_device_get_property_value(device, "ID_PATH_TAG");
+   if (!id_path_tag_tmp)
+      return NULL;
+
+   id_path_tag = strdup(id_path_tag_tmp);
+
+   udev_device_unref(device);
+   return id_path_tag;
+}
+
+static int
+drm_open_device(const char *device_name)
+{
+   int fd;
+#ifdef O_CLOEXEC
+   fd = open(device_name, O_RDWR | O_CLOEXEC);
+   if (fd == -1 && errno == EINVAL)
+#endif
+   {
+      fd = open(device_name, O_RDWR);
+      if (fd != -1)
+         fcntl(fd, F_SETFD, fcntl(fd, F_GETFD) | FD_CLOEXEC);
+   }
+   return fd;
+}
+
+#ifdef USE_DRICONF
+const char __driConfigOptionsLoader[] =
+DRI_CONF_BEGIN
+    DRI_CONF_SECTION_INITIALIZATION
+        DRI_CONF_DEVICE_ID_PATH_TAG()
+    DRI_CONF_SECTION_END
+DRI_CONF_END;
+#endif
+
+int loader_get_user_preferred_fd(int default_fd, int *different_device)
+{
+   struct udev *udev;
+#ifdef USE_DRICONF
+   driOptionCache defaultInitOptions;
+   driOptionCache userInitOptions;
+#endif
+   const char *dri_prime = getenv("DRI_PRIME");
+   char *prime = NULL;
+   int is_different_device = 0, fd = default_fd;
+   char *default_device_id_path_tag;
+   char *device_name = NULL;
+   char another_tag = 0;
+   UDEV_SYMBOL(struct udev *, udev_new, (void));
+   UDEV_SYMBOL(struct udev *, udev_unref, (struct udev *));
+
+   if (dri_prime)
+      prime = strdup(dri_prime);
+#ifdef USE_DRICONF
+   else {
+      driParseOptionInfo(&defaultInitOptions, __driConfigOptionsLoader);
+      driParseConfigFiles(&userInitOptions, &defaultInitOptions, 0, "loader");
+      if (driCheckOption(&userInitOptions, "device_id", DRI_STRING))
+         prime = strdup(driQueryOptionstr(&userInitOptions, "device_id"));
+      driDestroyOptionCache(&userInitOptions);
+      driDestroyOptionInfo(&defaultInitOptions);
+   }
+#endif
+
+   if (prime == NULL) {
+      *different_device = 0;
+      return default_fd;
+   }
+
+   udev = udev_new();
+   if (!udev)
+      goto prime_clean;
+
+   default_device_id_path_tag = get_id_path_tag_from_fd(udev, default_fd);
+   if (!default_device_id_path_tag)
+      goto udev_clean;
+
+   is_different_device = 1;
+   /* two format are supported:
+    * "1": choose any other card than the card used by default.
+    * id_path_tag: (for example "pci-0000_02_00_0") choose the card
+    * with this id_path_tag.
+    */
+   if (!strcmp(prime,"1")) {
+      free(prime);
+      prime = strdup(default_device_id_path_tag);
+      /* request a card with a different card than the default card */
+      another_tag = 1;
+   } else if (!strcmp(default_device_id_path_tag, prime))
+      /* we are to get a new fd (render-node) of the same device */
+      is_different_device = 0;
+
+   device_name = get_render_node_from_id_path_tag(udev,
+                                                  prime,
+                                                  another_tag);
+   if (device_name == NULL) {
+      is_different_device = 0;
+      goto default_device_clean;
+   }
+
+   fd = drm_open_device(device_name);
+   if (fd > 0) {
+      close(default_fd);
+   } else {
+      fd = default_fd;
+      is_different_device = 0;
+   }
+   free(device_name);
+
+ default_device_clean:
+   free(default_device_id_path_tag);
+ udev_clean:
+   udev_unref(udev);
+ prime_clean:
+   free(prime);
+
+   *different_device = is_different_device;
+   return fd;
+}
+#else
+int loader_get_user_preferred_fd(int default_fd, int *different_device)
+{
+   *different_device = 0;
+   return default_fd;
+}
 #endif
 
 #if defined(HAVE_SYSFS)
diff --git a/mesalib/src/loader/loader.h b/mesalib/src/loader/loader.h
index dfd77baad..fa57950de 100644
--- a/mesalib/src/loader/loader.h
+++ b/mesalib/src/loader/loader.h
@@ -41,6 +41,13 @@ loader_get_driver_for_fd(int fd, unsigned driver_types);
 char *
 loader_get_device_name_for_fd(int fd);
 
+/* Function to get a different device than the one we are to use by default,
+ * if the user requests so and it is possible. The initial fd will be closed
+ * if neccessary. The returned fd is potentially a render-node.
+ */
+
+int
+loader_get_user_preferred_fd(int default_fd, int *different_device);
 
 /* for logging.. keep this aligned with egllog.h so we can just use
  * _eglLog directly.
diff --git a/mesalib/src/mapi/glapi/gen/gl_API.xml b/mesalib/src/mapi/glapi/gen/gl_API.xml
index 0a486f070..882e66e0b 100644
--- a/mesalib/src/mapi/glapi/gen/gl_API.xml
+++ b/mesalib/src/mapi/glapi/gen/gl_API.xml
@@ -12157,22 +12157,22 @@
         <size name="GetTexEnviv" mode="get"/>
         <size name="GetTexEnvfv" mode="get"/>
     </enum>
-    <function name="TexBumpParameterfvATI" offset="assign" deprecated="3.1">
+    <function name="TexBumpParameterfvATI" offset="assign" deprecated="3.1" exec="skip">
       <param name="pname" type="GLenum"/>
       <param name="param" type="const GLfloat *" variable_param="pname"/>
       <glx ignore="true"/>
     </function>
-    <function name="TexBumpParameterivATI" offset="assign" deprecated="3.1">
+    <function name="TexBumpParameterivATI" offset="assign" deprecated="3.1" exec="skip">
       <param name="pname" type="GLenum"/>
       <param name="param" type="const GLint *" variable_param="pname"/>
       <glx ignore="true"/>
     </function>
-    <function name="GetTexBumpParameterfvATI" offset="assign" deprecated="3.1">
+    <function name="GetTexBumpParameterfvATI" offset="assign" deprecated="3.1" exec="skip">
       <param name="pname" type="GLenum"/>
       <param name="param" type="GLfloat *" variable_param="pname"/>
       <glx ignore="true"/>
     </function>
-    <function name="GetTexBumpParameterivATI" offset="assign" deprecated="3.1">
+    <function name="GetTexBumpParameterivATI" offset="assign" deprecated="3.1" exec="skip">
       <param name="pname" type="GLenum"/>
       <param name="param" type="GLint *" variable_param="pname"/>
       <glx ignore="true"/>
diff --git a/mesalib/src/mesa/drivers/common/meta.c b/mesalib/src/mesa/drivers/common/meta.c
index 1a2e45320..f313f5645 100644
--- a/mesalib/src/mesa/drivers/common/meta.c
+++ b/mesalib/src/mesa/drivers/common/meta.c
@@ -1515,23 +1515,15 @@ static void
 meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
 {
    const char *vs_source =
+      "#extension GL_AMD_vertex_shader_layer : enable\n"
       "attribute vec4 position;\n"
-      "void main()\n"
-      "{\n"
-      "   gl_Position = position;\n"
-      "}\n";
-   const char *gs_source =
-      "#version 150\n"
-      "layout(triangles) in;\n"
-      "layout(triangle_strip, max_vertices = 4) out;\n"
       "uniform int layer;\n"
       "void main()\n"
       "{\n"
-      "  for (int i = 0; i < 3; i++) {\n"
-      "    gl_Layer = layer;\n"
-      "    gl_Position = gl_in[i].gl_Position;\n"
-      "    EmitVertex();\n"
-      "  }\n"
+      "#ifdef GL_AMD_vertex_shader_layer\n"
+      "   gl_Layer = layer;\n"
+      "#endif\n"
+      "   gl_Position = position;\n"
       "}\n";
    const char *fs_source =
       "uniform vec4 color;\n"
@@ -1539,7 +1531,7 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
       "{\n"
       "   gl_FragColor = color;\n"
       "}\n";
-   GLuint vs, gs = 0, fs;
+   GLuint vs, fs;
    bool has_integer_textures;
 
    _mesa_meta_setup_vertex_objects(&clear->VAO, &clear->VBO, true, 3, 0, 0);
@@ -1551,12 +1543,6 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
    _mesa_ShaderSource(vs, 1, &vs_source, NULL);
    _mesa_CompileShader(vs);
 
-   if (_mesa_has_geometry_shaders(ctx)) {
-      gs = _mesa_CreateShader(GL_GEOMETRY_SHADER);
-      _mesa_ShaderSource(gs, 1, &gs_source, NULL);
-      _mesa_CompileShader(gs);
-   }
-
    fs = _mesa_CreateShader(GL_FRAGMENT_SHADER);
    _mesa_ShaderSource(fs, 1, &fs_source, NULL);
    _mesa_CompileShader(fs);
@@ -1564,20 +1550,14 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
    clear->ShaderProg = _mesa_CreateProgram();
    _mesa_AttachShader(clear->ShaderProg, fs);
    _mesa_DeleteShader(fs);
-   if (gs != 0)
-      _mesa_AttachShader(clear->ShaderProg, gs);
    _mesa_AttachShader(clear->ShaderProg, vs);
    _mesa_DeleteShader(vs);
    _mesa_BindAttribLocation(clear->ShaderProg, 0, "position");
    _mesa_ObjectLabel(GL_PROGRAM, clear->ShaderProg, -1, "meta clear");
    _mesa_LinkProgram(clear->ShaderProg);
 
-   clear->ColorLocation = _mesa_GetUniformLocation(clear->ShaderProg,
-						      "color");
-   if (gs != 0) {
-      clear->LayerLocation = _mesa_GetUniformLocation(clear->ShaderProg,
-						      "layer");
-   }
+   clear->ColorLocation = _mesa_GetUniformLocation(clear->ShaderProg, "color");
+   clear->LayerLocation = _mesa_GetUniformLocation(clear->ShaderProg, "layer");
 
    has_integer_textures = _mesa_is_gles3(ctx) ||
       (_mesa_is_desktop_gl(ctx) && ctx->Const.GLSLVersion >= 130);
@@ -1587,9 +1567,14 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
       const char *vs_int_source =
          ralloc_asprintf(shader_source_mem_ctx,
                          "#version 130\n"
+                         "#extension GL_AMD_vertex_shader_layer : enable\n"
                          "in vec4 position;\n"
+                         "uniform int layer;\n"
                          "void main()\n"
                          "{\n"
+                         "#ifdef GL_AMD_vertex_shader_layer\n"
+                         "   gl_Layer = layer;\n"
+                         "#endif\n"
                          "   gl_Position = position;\n"
                          "}\n");
       const char *fs_int_source =
@@ -1612,8 +1597,6 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
       clear->IntegerShaderProg = _mesa_CreateProgram();
       _mesa_AttachShader(clear->IntegerShaderProg, fs);
       _mesa_DeleteShader(fs);
-      if (gs != 0)
-         _mesa_AttachShader(clear->IntegerShaderProg, gs);
       _mesa_AttachShader(clear->IntegerShaderProg, vs);
       _mesa_DeleteShader(vs);
       _mesa_BindAttribLocation(clear->IntegerShaderProg, 0, "position");
@@ -1629,13 +1612,9 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
 
       clear->IntegerColorLocation =
 	 _mesa_GetUniformLocation(clear->IntegerShaderProg, "color");
-      if (gs != 0) {
-         clear->IntegerLayerLocation =
-            _mesa_GetUniformLocation(clear->IntegerShaderProg, "layer");
-      }
+      clear->IntegerLayerLocation =
+         _mesa_GetUniformLocation(clear->IntegerShaderProg, "layer");
    }
-   if (gs != 0)
-      _mesa_DeleteShader(gs);
 }
 
 static void
@@ -1843,7 +1822,7 @@ meta_clear(struct gl_context *ctx, GLbitfield buffers, bool glsl)
    /* draw quad(s) */
    if (fb->MaxNumLayers > 0) {
       unsigned layer;
-      assert(glsl);
+      assert(glsl && clear->LayerLocation != -1);
       for (layer = 0; layer < fb->MaxNumLayers; layer++) {
          if (fb->_IntegerColor)
             _mesa_Uniform1i(clear->IntegerLayerLocation, layer);
diff --git a/mesalib/src/mesa/drivers/dri/common/dri_util.c b/mesalib/src/mesa/drivers/dri/common/dri_util.c
index 471287909..8506f5356 100644
--- a/mesalib/src/mesa/drivers/dri/common/dri_util.c
+++ b/mesalib/src/mesa/drivers/dri/common/dri_util.c
@@ -677,7 +677,7 @@ dri2ReleaseBuffer(__DRIscreen *screen, __DRIbuffer *buffer)
 
 
 static int
-dri2ConfigQueryb(__DRIscreen *screen, const char *var, GLboolean *val)
+dri2ConfigQueryb(__DRIscreen *screen, const char *var, bool *val)
 {
    if (!driCheckOption(&screen->optionCache, var, DRI_BOOL))
       return -1;
@@ -688,7 +688,7 @@ dri2ConfigQueryb(__DRIscreen *screen, const char *var, GLboolean *val)
 }
 
 static int
-dri2ConfigQueryi(__DRIscreen *screen, const char *var, GLint *val)
+dri2ConfigQueryi(__DRIscreen *screen, const char *var, int *val)
 {
    if (!driCheckOption(&screen->optionCache, var, DRI_INT) &&
        !driCheckOption(&screen->optionCache, var, DRI_ENUM))
@@ -700,7 +700,7 @@ dri2ConfigQueryi(__DRIscreen *screen, const char *var, GLint *val)
 }
 
 static int
-dri2ConfigQueryf(__DRIscreen *screen, const char *var, GLfloat *val)
+dri2ConfigQueryf(__DRIscreen *screen, const char *var, float *val)
 {
    if (!driCheckOption(&screen->optionCache, var, DRI_FLOAT))
       return -1;
diff --git a/mesalib/src/mesa/drivers/dri/common/xmlconfig.c b/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
index b95e452f1..5091c115b 100644
--- a/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
+++ b/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
@@ -27,8 +27,6 @@
  * \author Felix Kuehling
  */
 
-#include "main/glheader.h"
-
 #include <string.h>
 #include <assert.h>
 #include <expat.h>
@@ -106,15 +104,15 @@ static const char *__getProgramName () {
 #endif
 
 /** \brief Find an option in an option cache with the name as key */
-static GLuint findOption (const driOptionCache *cache, const char *name) {
-    GLuint len = strlen (name);
-    GLuint size = 1 << cache->tableSize, mask = size - 1;
-    GLuint hash = 0;
-    GLuint i, shift;
+static uint32_t findOption (const driOptionCache *cache, const char *name) {
+    uint32_t len = strlen (name);
+    uint32_t size = 1 << cache->tableSize, mask = size - 1;
+    uint32_t hash = 0;
+    uint32_t i, shift;
 
   /* compute a hash from the variable length name */
     for (i = 0, shift = 0; i < len; ++i, shift = (shift+8) & 31)
-	hash += (GLuint)name[i] << shift;
+	hash += (uint32_t)name[i] << shift;
     hash *= hash;
     hash = (hash >> (16-cache->tableSize/2)) & mask;
 
@@ -134,7 +132,7 @@ static GLuint findOption (const driOptionCache *cache, const char *name) {
 
 /** \brief Like strdup but using malloc and with error checking. */
 #define XSTRDUP(dest,source) do { \
-    GLuint len = strlen (source); \
+    uint32_t len = strlen (source); \
     if (!(dest = malloc(len+1))) { \
 	fprintf (stderr, "%s: %d: out of memory.\n", __FILE__, __LINE__); \
 	abort(); \
@@ -146,8 +144,8 @@ static int compare (const void *a, const void *b) {
     return strcmp (*(char *const*)a, *(char *const*)b);
 }
 /** \brief Binary search in a string array. */
-static GLuint bsearchStr (const XML_Char *name,
-			  const XML_Char *elems[], GLuint count) {
+static uint32_t bsearchStr (const XML_Char *name,
+			  const XML_Char *elems[], uint32_t count) {
     const XML_Char **found;
     found = bsearch (&name, elems, count, sizeof (XML_Char *), compare);
     if (found)
@@ -165,11 +163,11 @@ static GLuint bsearchStr (const XML_Char *name,
  * returning tail points to the first character that is not part of
  * the integer number. If no number was found then tail points to the
  * start of the input string. */
-static GLint strToI (const XML_Char *string, const XML_Char **tail, int base) {
-    GLint radix = base == 0 ? 10 : base;
-    GLint result = 0;
-    GLint sign = 1;
-    GLboolean numberFound = GL_FALSE;
+static int strToI (const XML_Char *string, const XML_Char **tail, int base) {
+    int radix = base == 0 ? 10 : base;
+    int result = 0;
+    int sign = 1;
+    bool numberFound = false;
     const XML_Char *start = string;
 
     assert (radix >= 2 && radix <= 36);
@@ -180,7 +178,7 @@ static GLint strToI (const XML_Char *string, const XML_Char **tail, int base) {
     } else if (*string == '+')
 	string++;
     if (base == 0 && *string == '0') {
-	numberFound = GL_TRUE; 
+	numberFound = true;
 	if (*(string+1) == 'x' || *(string+1) == 'X') {
 	    radix = 16;
 	    string += 2;
@@ -190,7 +188,7 @@ static GLint strToI (const XML_Char *string, const XML_Char **tail, int base) {
 	}
     }
     do {
-	GLint digit = -1;
+	int digit = -1;
 	if (radix <= 10) {
 	    if (*string >= '0' && *string < '0' + radix)
 		digit = *string - '0';
@@ -203,12 +201,12 @@ static GLint strToI (const XML_Char *string, const XML_Char **tail, int base) {
 		digit = *string - 'A' + 10;
 	}
 	if (digit != -1) {
-	    numberFound = GL_TRUE;
+	    numberFound = true;
 	    result = radix*result + digit;
 	    string++;
 	} else
 	    break;
-    } while (GL_TRUE);
+    } while (true);
     *tail = numberFound ? string : start;
     return sign * result;
 }
@@ -225,9 +223,9 @@ static GLint strToI (const XML_Char *string, const XML_Char **tail, int base) {
  * to the start of the input string.
  *
  * Uses two passes for maximum accuracy. */
-static GLfloat strToF (const XML_Char *string, const XML_Char **tail) {
-    GLint nDigits = 0, pointPos, exponent;
-    GLfloat sign = 1.0f, result = 0.0f, scale;
+static float strToF (const XML_Char *string, const XML_Char **tail) {
+    int nDigits = 0, pointPos, exponent;
+    float sign = 1.0f, result = 0.0f, scale;
     const XML_Char *start = string, *numStart;
 
     /* sign */
@@ -270,13 +268,13 @@ static GLfloat strToF (const XML_Char *string, const XML_Char **tail) {
     string = numStart;
 
     /* scale of the first digit */
-    scale = sign * (GLfloat)pow (10.0, (GLdouble)(pointPos-1 + exponent));
+    scale = sign * (float)pow (10.0, (double)(pointPos-1 + exponent));
 
     /* second pass: parse digits */
     do {
 	if (*string != '.') {
 	    assert (*string >= '0' && *string <= '9');
-	    result += scale * (GLfloat)(*string - '0');
+	    result += scale * (float)(*string - '0');
 	    scale *= 0.1f;
 	    nDigits--;
 	}
@@ -287,7 +285,7 @@ static GLfloat strToF (const XML_Char *string, const XML_Char **tail) {
 }
 
 /** \brief Parse a value of a given type. */
-static GLboolean parseValue (driOptionValue *v, driOptionType type,
+static bool parseValue (driOptionValue *v, driOptionType type,
 			     const XML_Char *string) {
     const XML_Char *tail = NULL;
   /* skip leading white-space */
@@ -295,14 +293,14 @@ static GLboolean parseValue (driOptionValue *v, driOptionType type,
     switch (type) {
       case DRI_BOOL:
 	if (!strcmp (string, "false")) {
-	    v->_bool = GL_FALSE;
+	    v->_bool = false;
 	    tail = string + 5;
 	} else if (!strcmp (string, "true")) {
-	    v->_bool = GL_TRUE;
+	    v->_bool = true;
 	    tail = string + 4;
 	}
 	else
-	    return GL_FALSE;
+	    return false;
 	break;
       case DRI_ENUM: /* enum is just a special integer */
       case DRI_INT:
@@ -311,23 +309,28 @@ static GLboolean parseValue (driOptionValue *v, driOptionType type,
       case DRI_FLOAT:
 	v->_float = strToF (string, &tail);
 	break;
+      case DRI_STRING:
+	if (v->_string)
+	    free (v->_string);
+	v->_string = strndup(string, STRING_CONF_MAXLEN);
+	return GL_TRUE;
     }
 
     if (tail == string)
-	return GL_FALSE; /* empty string (or containing only white-space) */
+	return false; /* empty string (or containing only white-space) */
   /* skip trailing white space */
     if (*tail)
 	tail += strspn (tail, " \f\n\r\t\v");
     if (*tail)
-	return GL_FALSE; /* something left over that is not part of value */
+	return false; /* something left over that is not part of value */
 
-    return GL_TRUE;
+    return true;
 }
 
 /** \brief Parse a list of ranges of type info->type. */
-static GLboolean parseRanges (driOptionInfo *info, const XML_Char *string) {
+static bool parseRanges (driOptionInfo *info, const XML_Char *string) {
     XML_Char *cp, *range;
-    GLuint nRanges, i;
+    uint32_t nRanges, i;
     driOptionRange *ranges;
 
     XSTRDUP (cp, string);
@@ -375,39 +378,41 @@ static GLboolean parseRanges (driOptionInfo *info, const XML_Char *string) {
     free(cp);
     if (i < nRanges) {
 	free(ranges);
-	return GL_FALSE;
+	return false;
     } else
 	assert (range == NULL);
 
     info->nRanges = nRanges;
     info->ranges = ranges;
-    return GL_TRUE;
+    return true;
 }
 
 /** \brief Check if a value is in one of info->ranges. */
-static GLboolean checkValue (const driOptionValue *v, const driOptionInfo *info) {
-    GLuint i;
+static bool checkValue (const driOptionValue *v, const driOptionInfo *info) {
+    uint32_t i;
     assert (info->type != DRI_BOOL); /* should be caught by the parser */
     if (info->nRanges == 0)
-	return GL_TRUE;
+	return true;
     switch (info->type) {
       case DRI_ENUM: /* enum is just a special integer */
       case DRI_INT:
 	for (i = 0; i < info->nRanges; ++i)
 	    if (v->_int >= info->ranges[i].start._int &&
 		v->_int <= info->ranges[i].end._int)
-		return GL_TRUE;
+		return true;
 	break;
       case DRI_FLOAT:
 	for (i = 0; i < info->nRanges; ++i)
 	    if (v->_float >= info->ranges[i].start._float &&
 		v->_float <= info->ranges[i].end._float)
-		return GL_TRUE;
+		return true;
+	break;
+      case DRI_STRING:
 	break;
       default:
 	assert (0); /* should never happen */
     }
-    return GL_FALSE;
+    return false;
 }
 
 /**
@@ -478,11 +483,11 @@ struct OptInfoData {
     const char *name;
     XML_Parser parser;
     driOptionCache *cache;
-    GLboolean inDriInfo;
-    GLboolean inSection;
-    GLboolean inDesc;
-    GLboolean inOption;
-    GLboolean inEnum;
+    bool inDriInfo;
+    bool inSection;
+    bool inDesc;
+    bool inOption;
+    bool inEnum;
     int curOption;
 };
 
@@ -500,10 +505,10 @@ static const XML_Char *OptInfoElems[] = {
  * for external configuration tools.
  */
 static void parseEnumAttr (struct OptInfoData *data, const XML_Char **attr) {
-    GLuint i;
+    uint32_t i;
     const XML_Char *value = NULL, *text = NULL;
     driOptionValue v;
-    GLuint opt = data->curOption;
+    uint32_t opt = data->curOption;
     for (i = 0; attr[i]; i += 2) {
 	if (!strcmp (attr[i], "value")) value = attr[i+1];
 	else if (!strcmp (attr[i], "text")) text = attr[i+1];
@@ -523,7 +528,7 @@ static void parseEnumAttr (struct OptInfoData *data, const XML_Char **attr) {
  * for external configuration tools.
  */
 static void parseDescAttr (struct OptInfoData *data, const XML_Char **attr) {
-    GLuint i;
+    uint32_t i;
     const XML_Char *lang = NULL, *text = NULL;
     for (i = 0; attr[i]; i += 2) {
 	if (!strcmp (attr[i], "lang")) lang = attr[i+1];
@@ -541,9 +546,9 @@ static void parseOptInfoAttr (struct OptInfoData *data, const XML_Char **attr) {
     const XML_Char *attrVal[OA_COUNT] = {NULL, NULL, NULL, NULL};
     const char *defaultVal;
     driOptionCache *cache = data->cache;
-    GLuint opt, i;
+    uint32_t opt, i;
     for (i = 0; attr[i]; i += 2) {
-	GLuint attrName = bsearchStr (attr[i], optAttr, OA_COUNT);
+	uint32_t attrName = bsearchStr (attr[i], optAttr, OA_COUNT);
 	if (attrName >= OA_COUNT)
 	    XML_FATAL ("illegal option attribute: %s", attr[i]);
 	attrVal[attrName] = attr[i+1];
@@ -567,6 +572,8 @@ static void parseOptInfoAttr (struct OptInfoData *data, const XML_Char **attr) {
 	cache->info[opt].type = DRI_INT;
     else if (!strcmp (attrVal[OA_TYPE], "float"))
 	cache->info[opt].type = DRI_FLOAT;
+    else if (!strcmp (attrVal[OA_TYPE], "string"))
+	cache->info[opt].type = DRI_STRING;
     else
 	XML_FATAL ("illegal type in option: %s.", attrVal[OA_TYPE]);
 
@@ -608,7 +615,7 @@ static void optInfoStartElem (void *userData, const XML_Char *name,
 	    XML_FATAL1 ("nested <driinfo> elements.");
 	if (attr[0])
 	    XML_FATAL1 ("attributes specified on <driinfo> element.");
-	data->inDriInfo = GL_TRUE;
+	data->inDriInfo = true;
 	break;
       case OI_SECTION:
 	if (!data->inDriInfo)
@@ -617,14 +624,14 @@ static void optInfoStartElem (void *userData, const XML_Char *name,
 	    XML_FATAL1 ("nested <section> elements.");
 	if (attr[0])
 	    XML_FATAL1 ("attributes specified on <section> element.");
-	data->inSection = GL_TRUE;
+	data->inSection = true;
 	break;
       case OI_DESCRIPTION:
 	if (!data->inSection && !data->inOption)
 	    XML_FATAL1 ("<description> must be inside <description> or <option.");
 	if (data->inDesc)
 	    XML_FATAL1 ("nested <description> elements.");
-	data->inDesc = GL_TRUE;
+	data->inDesc = true;
 	parseDescAttr (data, attr);
 	break;
       case OI_OPTION:
@@ -634,7 +641,7 @@ static void optInfoStartElem (void *userData, const XML_Char *name,
 	    XML_FATAL1 ("<option> nested in <description> element.");
 	if (data->inOption)
 	    XML_FATAL1 ("nested <option> elements.");
-	data->inOption = GL_TRUE;
+	data->inOption = true;
 	parseOptInfoAttr (data, attr);
 	break;
       case OI_ENUM:
@@ -642,7 +649,7 @@ static void optInfoStartElem (void *userData, const XML_Char *name,
 	    XML_FATAL1 ("<enum> must be inside <option> and <description>.");
 	if (data->inEnum)
 	    XML_FATAL1 ("nested <enum> elements.");
-	data->inEnum = GL_TRUE;
+	data->inEnum = true;
 	parseEnumAttr (data, attr);
 	break;
       default:
@@ -656,19 +663,19 @@ static void optInfoEndElem (void *userData, const XML_Char *name) {
     enum OptInfoElem elem = bsearchStr (name, OptInfoElems, OI_COUNT);
     switch (elem) {
       case OI_DRIINFO:
-	data->inDriInfo = GL_FALSE;
+	data->inDriInfo = false;
 	break;
       case OI_SECTION:
-	data->inSection = GL_FALSE;
+	data->inSection = false;
 	break;
       case OI_DESCRIPTION:
-	data->inDesc = GL_FALSE;
+	data->inDesc = false;
 	break;
       case OI_OPTION:
-	data->inOption = GL_FALSE;
+	data->inOption = false;
 	break;
       case OI_ENUM:
-	data->inEnum = GL_FALSE;
+	data->inEnum = false;
 	break;
       default:
 	assert (0); /* should have been caught by StartElem */
@@ -699,11 +706,11 @@ void driParseOptionInfo (driOptionCache *info, const char *configOptions) {
     userData.name = "__driConfigOptions";
     userData.parser = p;
     userData.cache = info;
-    userData.inDriInfo = GL_FALSE;
-    userData.inSection = GL_FALSE;
-    userData.inDesc = GL_FALSE;
-    userData.inOption = GL_FALSE;
-    userData.inEnum = GL_FALSE;
+    userData.inDriInfo = false;
+    userData.inSection = false;
+    userData.inDesc = false;
+    userData.inOption = false;
+    userData.inEnum = false;
     userData.curOption = -1;
 
     status = XML_Parse (p, configOptions, strlen (configOptions), 1);
@@ -718,14 +725,14 @@ struct OptConfData {
     const char *name;
     XML_Parser parser;
     driOptionCache *cache;
-    GLint screenNum;
+    int screenNum;
     const char *driverName, *execName;
-    GLuint ignoringDevice;
-    GLuint ignoringApp;
-    GLuint inDriConf;
-    GLuint inDevice;
-    GLuint inApp;
-    GLuint inOption;
+    uint32_t ignoringDevice;
+    uint32_t ignoringApp;
+    uint32_t inDriConf;
+    uint32_t inDevice;
+    uint32_t inApp;
+    uint32_t inOption;
 };
 
 /** \brief Elements in configuration files. */
@@ -738,7 +745,7 @@ static const XML_Char *OptConfElems[] = {
 
 /** \brief Parse attributes of a device element. */
 static void parseDeviceAttr (struct OptConfData *data, const XML_Char **attr) {
-    GLuint i;
+    uint32_t i;
     const XML_Char *driver = NULL, *screen = NULL;
     for (i = 0; attr[i]; i += 2) {
 	if (!strcmp (attr[i], "driver")) driver = attr[i+1];
@@ -758,7 +765,7 @@ static void parseDeviceAttr (struct OptConfData *data, const XML_Char **attr) {
 
 /** \brief Parse attributes of an application element. */
 static void parseAppAttr (struct OptConfData *data, const XML_Char **attr) {
-    GLuint i;
+    uint32_t i;
     const XML_Char *exec = NULL;
     for (i = 0; attr[i]; i += 2) {
 	if (!strcmp (attr[i], "name")) /* not needed here */;
@@ -771,7 +778,7 @@ static void parseAppAttr (struct OptConfData *data, const XML_Char **attr) {
 
 /** \brief Parse attributes of an option element. */
 static void parseOptConfAttr (struct OptConfData *data, const XML_Char **attr) {
-    GLuint i;
+    uint32_t i;
     const XML_Char *name = NULL, *value = NULL;
     for (i = 0; attr[i]; i += 2) {
 	if (!strcmp (attr[i], "name")) name = attr[i+1];
@@ -782,7 +789,7 @@ static void parseOptConfAttr (struct OptConfData *data, const XML_Char **attr) {
     if (!value) XML_WARNING1 ("value attribute missing in option.");
     if (name && value) {
 	driOptionCache *cache = data->cache;
-	GLuint opt = findOption (cache, name);
+	uint32_t opt = findOption (cache, name);
 	if (cache->info[opt].name == NULL)
             /* don't use XML_WARNING, drirc defines options for all drivers,
              * but not all drivers support them */
@@ -867,6 +874,7 @@ static void optConfEndElem (void *userData, const XML_Char *name) {
 
 /** \brief Initialize an option cache based on info */
 static void initOptionCache (driOptionCache *cache, const driOptionCache *info) {
+    GLuint i, size = 1 << info->tableSize;
     cache->info = info->info;
     cache->tableSize = info->tableSize;
     cache->values = malloc((1<<info->tableSize) * sizeof (driOptionValue));
@@ -876,6 +884,10 @@ static void initOptionCache (driOptionCache *cache, const driOptionCache *info)
     }
     memcpy (cache->values, info->values,
 	    (1<<info->tableSize) * sizeof (driOptionValue));
+    for (i = 0; i < size; ++i) {
+	if (cache->info[i].type == DRI_STRING)
+	    XSTRDUP(cache->values[i]._string, info->values[i]._string);
+    }
 }
 
 /** \brief Parse the named configuration file */
@@ -918,10 +930,10 @@ static void parseOneConfigFile (XML_Parser p) {
 }
 
 void driParseConfigFiles (driOptionCache *cache, const driOptionCache *info,
-			  GLint screenNum, const char *driverName) {
+			  int screenNum, const char *driverName) {
     char *filenames[2] = {"/etc/drirc", NULL};
     char *home;
-    GLuint i;
+    uint32_t i;
     struct OptConfData userData;
 
     initOptionCache (cache, info);
@@ -932,7 +944,7 @@ void driParseConfigFiles (driOptionCache *cache, const driOptionCache *info,
     userData.execName = GET_PROGRAM_NAME();
 
     if ((home = getenv ("HOME"))) {
-	GLuint len = strlen (home);
+	uint32_t len = strlen (home);
 	filenames[1] = malloc(len + 7+1);
 	if (filenames[1] == NULL)
 	    __driUtilMessage ("Can't allocate memory for %s/.drirc.", home);
@@ -969,7 +981,7 @@ void driParseConfigFiles (driOptionCache *cache, const driOptionCache *info,
 void driDestroyOptionInfo (driOptionCache *info) {
     driDestroyOptionCache (info);
     if (info->info) {
-	GLuint i, size = 1 << info->tableSize;
+	uint32_t i, size = 1 << info->tableSize;
 	for (i = 0; i < size; ++i) {
 	    if (info->info[i].name) {
 		free(info->info[i].name);
@@ -981,35 +993,50 @@ void driDestroyOptionInfo (driOptionCache *info) {
 }
 
 void driDestroyOptionCache (driOptionCache *cache) {
+    if (cache->info) {
+	GLuint i, size = 1 << cache->tableSize;
+	for (i = 0; i < size; ++i) {
+	    if (cache->info[i].type == DRI_STRING)
+		free(cache->values[i]._string);
+	}
+    }
     free(cache->values);
 }
 
-GLboolean driCheckOption (const driOptionCache *cache, const char *name,
+bool driCheckOption (const driOptionCache *cache, const char *name,
 			  driOptionType type) {
-    GLuint i = findOption (cache, name);
+    uint32_t i = findOption (cache, name);
     return cache->info[i].name != NULL && cache->info[i].type == type;
 }
 
-GLboolean driQueryOptionb (const driOptionCache *cache, const char *name) {
-    GLuint i = findOption (cache, name);
+bool driQueryOptionb (const driOptionCache *cache, const char *name) {
+    uint32_t i = findOption (cache, name);
   /* make sure the option is defined and has the correct type */
     assert (cache->info[i].name != NULL);
     assert (cache->info[i].type == DRI_BOOL);
     return cache->values[i]._bool;
 }
 
-GLint driQueryOptioni (const driOptionCache *cache, const char *name) {
-    GLuint i = findOption (cache, name);
+int driQueryOptioni (const driOptionCache *cache, const char *name) {
+    uint32_t i = findOption (cache, name);
   /* make sure the option is defined and has the correct type */
     assert (cache->info[i].name != NULL);
     assert (cache->info[i].type == DRI_INT || cache->info[i].type == DRI_ENUM);
     return cache->values[i]._int;
 }
 
-GLfloat driQueryOptionf (const driOptionCache *cache, const char *name) {
-    GLuint i = findOption (cache, name);
+float driQueryOptionf (const driOptionCache *cache, const char *name) {
+    uint32_t i = findOption (cache, name);
   /* make sure the option is defined and has the correct type */
     assert (cache->info[i].name != NULL);
     assert (cache->info[i].type == DRI_FLOAT);
     return cache->values[i]._float;
 }
+
+char *driQueryOptionstr (const driOptionCache *cache, const char *name) {
+    GLuint i = findOption (cache, name);
+  /* make sure the option is defined and has the correct type */
+    assert (cache->info[i].name != NULL);
+    assert (cache->info[i].type == DRI_STRING);
+    return cache->values[i]._string;
+}
diff --git a/mesalib/src/mesa/drivers/dri/common/xmlconfig.h b/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
index d0ad42c19..386ddf19d 100644
--- a/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
+++ b/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
@@ -30,16 +30,21 @@
 #ifndef __XMLCONFIG_H
 #define __XMLCONFIG_H
 
+#include <stdbool.h>
+
+#define STRING_CONF_MAXLEN 25
+
 /** \brief Option data types */
 typedef enum driOptionType {
-    DRI_BOOL, DRI_ENUM, DRI_INT, DRI_FLOAT
+    DRI_BOOL, DRI_ENUM, DRI_INT, DRI_FLOAT, DRI_STRING
 } driOptionType;
 
 /** \brief Option value */
 typedef union driOptionValue {
-    GLboolean _bool; /**< \brief Boolean */
-    GLint _int;      /**< \brief Integer or Enum */
-    GLfloat _float;  /**< \brief Floating-point */
+    bool _bool; /**< \brief Boolean */
+    int _int;      /**< \brief Integer or Enum */
+    float _float;  /**< \brief Floating-point */
+    char *_string;   /**< \brief String */
 } driOptionValue;
 
 /** \brief Single range of valid values
@@ -55,7 +60,7 @@ typedef struct driOptionInfo {
     char *name;             /**< \brief Name */
     driOptionType type;     /**< \brief Type */
     driOptionRange *ranges; /**< \brief Array of ranges */
-    GLuint nRanges;         /**< \brief Number of ranges */
+    uint nRanges;         /**< \brief Number of ranges */
 } driOptionInfo;
 
 /** \brief Option cache
@@ -73,7 +78,7 @@ typedef struct driOptionCache {
    * \li Default values in screen
    * \li Actual values in contexts 
    */
-    GLuint tableSize;
+    uint tableSize;
   /**< \brief Size of the arrays
    *
    * In the current implementation it's not actually a size but log2(size).
@@ -98,7 +103,7 @@ void driParseOptionInfo (driOptionCache *info,
  * To be called in <driver>CreateContext. screenNum and driverName select
  * device sections. */
 void driParseConfigFiles (driOptionCache *cache, const driOptionCache *info,
-			  GLint screenNum, const char *driverName);
+			  int screenNum, const char *driverName);
 /** \brief Destroy option info
  *
  * To be called in <driver>DestroyScreen */
@@ -109,14 +114,16 @@ void driDestroyOptionInfo (driOptionCache *info);
 void driDestroyOptionCache (driOptionCache *cache);
 
 /** \brief Check if there exists a certain option */
-GLboolean driCheckOption (const driOptionCache *cache, const char *name,
+bool driCheckOption (const driOptionCache *cache, const char *name,
 			  driOptionType type);
 
 /** \brief Query a boolean option value */
-GLboolean driQueryOptionb (const driOptionCache *cache, const char *name);
+bool driQueryOptionb (const driOptionCache *cache, const char *name);
 /** \brief Query an integer option value */
-GLint driQueryOptioni (const driOptionCache *cache, const char *name);
+int driQueryOptioni (const driOptionCache *cache, const char *name);
 /** \brief Query a floating-point option value */
-GLfloat driQueryOptionf (const driOptionCache *cache, const char *name);
+float driQueryOptionf (const driOptionCache *cache, const char *name);
+/** \brief Query a string option value */
+char *driQueryOptionstr (const driOptionCache *cache, const char *name);
 
 #endif
diff --git a/mesalib/src/mesa/drivers/dri/common/xmlpool/t_options.h b/mesalib/src/mesa/drivers/dri/common/xmlpool/t_options.h
index 3bf804a17..fc9e10461 100644
--- a/mesalib/src/mesa/drivers/dri/common/xmlpool/t_options.h
+++ b/mesalib/src/mesa/drivers/dri/common/xmlpool/t_options.h
@@ -321,3 +321,17 @@ DRI_CONF_SECTION_BEGIN \
 DRI_CONF_OPT_BEGIN_B(always_have_depth_buffer, def) \
         DRI_CONF_DESC(en,gettext("Create all visuals with a depth buffer")) \
 DRI_CONF_OPT_END
+
+
+
+/**
+ * \brief Initialization configuration options
+ */
+#define DRI_CONF_SECTION_INITIALIZATION \
+DRI_CONF_SECTION_BEGIN \
+        DRI_CONF_DESC(en,gettext("Initialization"))
+
+#define DRI_CONF_DEVICE_ID_PATH_TAG(def) \
+DRI_CONF_OPT_BEGIN(device_id, string, def) \
+        DRI_CONF_DESC(en,gettext("Define the graphic device to use if possible")) \
+DRI_CONF_OPT_END
diff --git a/mesalib/src/mesa/main/context.c b/mesalib/src/mesa/main/context.c
index 244e63ede..b08215950 100644
--- a/mesalib/src/mesa/main/context.c
+++ b/mesalib/src/mesa/main/context.c
@@ -656,9 +656,6 @@ _mesa_init_constants(struct gl_context *ctx)
    /* GL_ARB_sync */
    ctx->Const.MaxServerWaitTimeout = 0x1fff7fffffffULL;
 
-   /* GL_ATI_envmap_bumpmap */
-   ctx->Const.SupportedBumpUnits = SUPPORTED_ATI_BUMP_UNITS;
-
    /* GL_EXT_provoking_vertex */
    ctx->Const.QuadsFollowProvokingVertexConvention = GL_TRUE;
 
diff --git a/mesalib/src/mesa/main/dlist.c b/mesalib/src/mesa/main/dlist.c
index 5874b99f0..5c7160d05 100644
--- a/mesalib/src/mesa/main/dlist.c
+++ b/mesalib/src/mesa/main/dlist.c
@@ -317,8 +317,6 @@ typedef enum
    /* GL_ARB_draw_buffers */
    OPCODE_DRAW_BUFFERS_ARB,
    /* GL_ATI_fragment_shader */
-   OPCODE_TEX_BUMP_PARAMETER_ATI,
-   /* GL_ATI_fragment_shader */
    OPCODE_BIND_FRAGMENT_SHADER_ATI,
    OPCODE_SET_FRAGMENT_SHADER_CONSTANTS_ATI,
    /* OpenGL 2.0 */
@@ -4977,36 +4975,6 @@ save_DrawBuffersARB(GLsizei count, const GLenum * buffers)
 }
 
 static void GLAPIENTRY
-save_TexBumpParameterfvATI(GLenum pname, const GLfloat *param)
-{
-   GET_CURRENT_CONTEXT(ctx);
-   Node *n;
-
-   n = alloc_instruction(ctx, OPCODE_TEX_BUMP_PARAMETER_ATI, 5);
-   if (n) {
-      n[1].ui = pname;
-      n[2].f = param[0];
-      n[3].f = param[1];
-      n[4].f = param[2];
-      n[5].f = param[3];
-   }
-   if (ctx->ExecuteFlag) {
-      CALL_TexBumpParameterfvATI(ctx->Exec, (pname, param));
-   }
-}
-
-static void GLAPIENTRY
-save_TexBumpParameterivATI(GLenum pname, const GLint *param)
-{
-   GLfloat p[4];
-   p[0] = INT_TO_FLOAT(param[0]);
-   p[1] = INT_TO_FLOAT(param[1]);
-   p[2] = INT_TO_FLOAT(param[2]);
-   p[3] = INT_TO_FLOAT(param[3]);
-   save_TexBumpParameterfvATI(pname, p);
-}
-
-static void GLAPIENTRY
 save_BindFragmentShaderATI(GLuint id)
 {
    GET_CURRENT_CONTEXT(ctx);
@@ -8653,16 +8621,6 @@ execute_list(struct gl_context *ctx, GLuint list)
             CALL_ClampColor(ctx->Exec, (n[1].e, n[2].e));
             break;
 
-         case OPCODE_TEX_BUMP_PARAMETER_ATI:
-            {
-               GLfloat values[4];
-               GLuint i, pname = n[1].ui;
-
-               for (i = 0; i < 4; i++)
-                  values[i] = n[1 + i].f;
-               CALL_TexBumpParameterfvATI(ctx->Exec, (pname, values));
-            }
-            break;
          case OPCODE_BIND_FRAGMENT_SHADER_ATI:
             CALL_BindFragmentShaderATI(ctx->Exec, (n[1].i));
             break;
@@ -9488,10 +9446,6 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
     */
    SET_BindProgramARB(table, save_BindProgramNV);
 
-   /* 244. GL_ATI_envmap_bumpmap */
-   SET_TexBumpParameterivATI(table, save_TexBumpParameterivATI);
-   SET_TexBumpParameterfvATI(table, save_TexBumpParameterfvATI);
-
    /* 245. GL_ATI_fragment_shader */
    SET_BindFragmentShaderATI(table, save_BindFragmentShaderATI);
    SET_SetFragmentShaderConstantATI(table, save_SetFragmentShaderConstantATI);
diff --git a/mesalib/src/mesa/main/extensions.c b/mesalib/src/mesa/main/extensions.c
index 25e3dab4f..6f2536170 100644
--- a/mesalib/src/mesa/main/extensions.c
+++ b/mesalib/src/mesa/main/extensions.c
@@ -136,6 +136,7 @@ static const struct extension extension_table[] = {
    { "GL_ARB_sample_shading",                      o(ARB_sample_shading),                      GL,             2009 },
    { "GL_ARB_sampler_objects",                     o(dummy_true),                              GL,             2009 },
    { "GL_ARB_seamless_cube_map",                   o(ARB_seamless_cube_map),                   GL,             2009 },
+   { "GL_ARB_seamless_cubemap_per_texture",        o(AMD_seamless_cubemap_per_texture),        GL,             2013 },
    { "GL_ARB_separate_shader_objects",             o(dummy_true),                              GL,             2010 },
    { "GL_ARB_shader_atomic_counters",              o(ARB_shader_atomic_counters),              GL,             2011 },
    { "GL_ARB_shader_bit_encoding",                 o(ARB_shader_bit_encoding),                 GL,             2010 },
@@ -328,7 +329,6 @@ static const struct extension extension_table[] = {
    { "GL_APPLE_vertex_array_object",               o(dummy_true),                              GLL,            2002 },
    { "GL_ATI_blend_equation_separate",             o(EXT_blend_equation_separate),             GL,             2003 },
    { "GL_ATI_draw_buffers",                        o(dummy_true),                              GLL,            2002 },
-   { "GL_ATI_envmap_bumpmap",                      o(ATI_envmap_bumpmap),                      GLL,            2001 },
    { "GL_ATI_fragment_shader",                     o(ATI_fragment_shader),                     GLL,            2001 },
    { "GL_ATI_separate_stencil",                    o(ATI_separate_stencil),                    GLL,            2006 },
    { "GL_ATI_texture_compression_3dc",             o(ATI_texture_compression_3dc),             GL,             2004 },
@@ -461,7 +461,6 @@ _mesa_enable_sw_extensions(struct gl_context *ctx)
    ctx->Extensions.ARB_vertex_shader = GL_TRUE;
    ctx->Extensions.ARB_sync = GL_TRUE;
    ctx->Extensions.APPLE_object_purgeable = GL_TRUE;
-   ctx->Extensions.ATI_envmap_bumpmap = GL_TRUE;
    ctx->Extensions.ATI_fragment_shader = GL_TRUE;
    ctx->Extensions.ATI_texture_compression_3dc = GL_TRUE;
    ctx->Extensions.ATI_texture_env_combine3 = GL_TRUE;
diff --git a/mesalib/src/mesa/main/ff_fragment_shader.cpp b/mesalib/src/mesa/main/ff_fragment_shader.cpp
index 211583738..2c4f3d7df 100644
--- a/mesalib/src/mesa/main/ff_fragment_shader.cpp
+++ b/mesalib/src/mesa/main/ff_fragment_shader.cpp
@@ -220,7 +220,6 @@ static GLuint translate_source( GLenum src )
 #define MODE_MODULATE_SUBTRACT_ATI      12  /* r = a0 * a2 - a1 */
 #define MODE_ADD_PRODUCTS               13  /* r = a0 * a1 + a2 * a3 */
 #define MODE_ADD_PRODUCTS_SIGNED        14  /* r = a0 * a1 + a2 * a3 - 0.5 */
-#define MODE_BUMP_ENVMAP_ATI            15  /* special */
 #define MODE_UNKNOWN                    16
 
 /**
@@ -250,7 +249,6 @@ static GLuint translate_mode( GLenum envMode, GLenum mode )
    case GL_MODULATE_ADD_ATI: return MODE_MODULATE_ADD_ATI;
    case GL_MODULATE_SIGNED_ADD_ATI: return MODE_MODULATE_SIGNED_ADD_ATI;
    case GL_MODULATE_SUBTRACT_ATI: return MODE_MODULATE_SUBTRACT_ATI;
-   case GL_BUMP_ENVMAP_ATI: return MODE_BUMP_ENVMAP_ATI;
    default:
       assert(0);
       return MODE_UNKNOWN;
@@ -283,7 +281,6 @@ need_saturate( GLuint mode )
    case MODE_MODULATE_SUBTRACT_ATI:
    case MODE_ADD_PRODUCTS:
    case MODE_ADD_PRODUCTS_SIGNED:
-   case MODE_BUMP_ENVMAP_ATI:
       return GL_TRUE;
    default:
       assert(0);
@@ -455,16 +452,6 @@ static GLuint make_state_key( struct gl_context *ctx,  struct state_key *key )
          key->unit[i].OptRGB[j].Source = translate_source(comb->SourceRGB[j]);
          key->unit[i].OptA[j].Source = translate_source(comb->SourceA[j]);
       }
-
-      if (key->unit[i].ModeRGB == MODE_BUMP_ENVMAP_ATI) {
-         /* requires some special translation */
-         key->unit[i].NumArgsRGB = 2;
-         key->unit[i].ScaleShiftRGB = 0;
-         key->unit[i].OptRGB[0].Operand = OPR_SRC_COLOR;
-         key->unit[i].OptRGB[0].Source = SRC_TEXTURE;
-         key->unit[i].OptRGB[1].Operand = OPR_SRC_COLOR;
-         key->unit[i].OptRGB[1].Source = texUnit->BumpTarget - GL_TEXTURE0 + SRC_TEXTURE0;
-       }
    }
 
    /* _NEW_LIGHT | _NEW_FOG */
@@ -753,11 +740,6 @@ emit_combine(texenv_fragment_program *p,
    case MODE_ADD_PRODUCTS_SIGNED:
       return add(add(mul(src[0], src[1]), mul(src[2], src[3])),
 		 new(p->mem_ctx) ir_constant(-0.5f));
-
-   case MODE_BUMP_ENVMAP_ATI:
-      /* special - not handled here */
-      assert(0);
-      return src[0];
    default: 
       assert(0);
       return src[0];
@@ -777,10 +759,6 @@ emit_texenv(texenv_fragment_program *p, GLuint unit)
    if (!key->unit[unit].enabled) {
       return get_source(p, SRC_PREVIOUS, 0);
    }
-   if (key->unit[unit].ModeRGB == MODE_BUMP_ENVMAP_ATI) {
-      /* this isn't really a env stage delivering a color and handled elsewhere */
-      return get_source(p, SRC_PREVIOUS, 0);
-   }
    
    switch (key->unit[unit].ModeRGB) {
    case MODE_DOT3_RGB_EXT:
@@ -1075,56 +1053,6 @@ load_texunit_sources( texenv_fragment_program *p, GLuint unit )
 }
 
 /**
- * Generate instructions for loading bump map textures.
- */
-static void
-load_texunit_bumpmap( texenv_fragment_program *p, GLuint unit )
-{
-   const struct state_key *key = p->state;
-   GLuint bumpedUnitNr = key->unit[unit].OptRGB[1].Source - SRC_TEXTURE0;
-   ir_rvalue *bump;
-   ir_rvalue *texcoord;
-   ir_variable *rot_mat_0, *rot_mat_1;
-
-   rot_mat_0 = p->shader->symbols->get_variable("gl_BumpRotMatrix0MESA");
-   assert(rot_mat_0);
-   rot_mat_1 = p->shader->symbols->get_variable("gl_BumpRotMatrix1MESA");
-   assert(rot_mat_1);
-
-   ir_variable *tc_array = p->shader->symbols->get_variable("gl_TexCoord");
-   assert(tc_array);
-   texcoord = new(p->mem_ctx) ir_dereference_variable(tc_array);
-   ir_rvalue *index = new(p->mem_ctx) ir_constant(bumpedUnitNr);
-   texcoord = new(p->mem_ctx) ir_dereference_array(texcoord, index);
-   tc_array->data.max_array_access = MAX2(tc_array->data.max_array_access, unit);
-
-   load_texenv_source( p, unit + SRC_TEXTURE0, unit );
-
-   /* Apply rot matrix and add coords to be available in next phase.
-    * dest = Arg1 + (Arg0.xx * rotMat0) + (Arg0.yy * rotMat1)
-    * note only 2 coords are affected the rest are left unchanged (mul by 0)
-    */
-   ir_rvalue *bump_x, *bump_y;
-
-   texcoord = smear(p, texcoord);
-
-   /* bump_texcoord = texcoord */
-   ir_variable *bumped = p->make_temp(texcoord->type, "bump_texcoord");
-   p->emit(bumped);
-   p->emit(assign(bumped, texcoord));
-
-   /* bump_texcoord.xy += arg0.x * rotmat0 + arg0.y * rotmat1 */
-   bump = get_source(p, key->unit[unit].OptRGB[0].Source, unit);
-   bump_x = mul(swizzle_x(bump), rot_mat_0);
-   bump_y = mul(swizzle_y(bump->clone(p->mem_ctx, NULL)), rot_mat_1);
-
-   p->emit(assign(bumped, add(swizzle_xy(bumped), add(bump_x, bump_y)),
-		  WRITEMASK_XY));
-
-   p->texcoord_tex[bumpedUnitNr] = bumped;
-}
-
-/**
  * Applies the fog calculations.
  *
  * This is basically like the ARB_fragment_prorgam fog options.  Note
@@ -1214,14 +1142,6 @@ emit_instructions(texenv_fragment_program *p)
    GLuint unit;
 
    if (key->enabled_units) {
-      /* Zeroth pass - bump map textures first */
-      for (unit = 0; unit < key->nr_enabled_units; unit++) {
-	 if (key->unit[unit].enabled &&
-             key->unit[unit].ModeRGB == MODE_BUMP_ENVMAP_ATI) {
-	    load_texunit_bumpmap(p, unit);
-	 }
-      }
-
       /* First pass - to support texture_env_crossbar, first identify
        * all referenced texture sources and emit texld instructions
        * for each:
diff --git a/mesalib/src/mesa/main/format_pack.c b/mesalib/src/mesa/main/format_pack.c
index e3cbfff7e..6b28592a6 100644
--- a/mesalib/src/mesa/main/format_pack.c
+++ b/mesalib/src/mesa/main/format_pack.c
@@ -1403,27 +1403,6 @@ pack_float_RG_FLOAT16(const GLfloat src[4], void *dst)
 }
 
 
-/* MESA_FORMAT_DUDV8 */
-
-static void
-pack_ubyte_DUDV8(const GLubyte src[4], void *dst)
-{
-   /* XXX is this ever used? */
-   GLushort *d = ((GLushort *) dst);
-   *d = PACK_COLOR_88(src[0], src[1]);
-}
-
-static void
-pack_float_DUDV8(const GLfloat src[4], void *dst)
-{
-   GLushort *d = ((GLushort *) dst);
-   GLbyte du, dv;
-   du = FLOAT_TO_BYTE(CLAMP(src[0], 0.0F, 1.0F));
-   dv = FLOAT_TO_BYTE(CLAMP(src[1], 0.0F, 1.0F));
-   *d = PACK_COLOR_88(du, dv);
-}
-
-
 /* MESA_FORMAT_RGBA_UNORM16 */
 
 static void
@@ -2027,8 +2006,6 @@ _mesa_get_pack_ubyte_rgba_function(mesa_format format)
       table[MESA_FORMAT_RGBA_UINT16] = NULL; /* pack_ubyte_RGBA_UINT16 */
       table[MESA_FORMAT_RGBA_UINT32] = NULL; /* pack_ubyte_RGBA_UINT32 */
 
-      table[MESA_FORMAT_DUDV8] = pack_ubyte_DUDV8;
-
       table[MESA_FORMAT_RGBA_UNORM16] = pack_ubyte_RGBA_16;
 
       /* n/a */
@@ -2194,8 +2171,6 @@ _mesa_get_pack_float_rgba_function(mesa_format format)
       table[MESA_FORMAT_RGBA_UINT16] = NULL;
       table[MESA_FORMAT_RGBA_UINT32] = NULL;
 
-      table[MESA_FORMAT_DUDV8] = pack_float_DUDV8;
-
       table[MESA_FORMAT_RGBA_UNORM16] = pack_float_RGBA_16;
 
       table[MESA_FORMAT_R_SNORM8] = pack_float_R_SNORM8;
diff --git a/mesalib/src/mesa/main/format_unpack.c b/mesalib/src/mesa/main/format_unpack.c
index 9cc8f4dba..ad5ea4cd1 100644
--- a/mesalib/src/mesa/main/format_unpack.c
+++ b/mesalib/src/mesa/main/format_unpack.c
@@ -1708,19 +1708,6 @@ unpack_RGBA_UINT32(const void *src, GLfloat dst[][4], GLuint n)
 }
 
 static void
-unpack_DUDV8(const void *src, GLfloat dst[][4], GLuint n)
-{
-   const GLbyte *s = (const GLbyte *) src;
-   GLuint i;
-   for (i = 0; i < n; i++) {
-      dst[i][RCOMP] = BYTE_TO_FLOAT(s[i*2+0]);
-      dst[i][GCOMP] = BYTE_TO_FLOAT(s[i*2+1]);
-      dst[i][BCOMP] = 0;
-      dst[i][ACOMP] = 0;
-   }
-}
-
-static void
 unpack_R_SNORM8(const void *src, GLfloat dst[][4], GLuint n)
 {
    const GLbyte *s = ((const GLbyte *) src);
@@ -2486,7 +2473,6 @@ get_unpack_rgba_function(mesa_format format)
       table[MESA_FORMAT_RGB_UINT32] = unpack_RGB_UINT32;
       table[MESA_FORMAT_RGBA_UINT32] = unpack_RGBA_UINT32;
 
-      table[MESA_FORMAT_DUDV8] = unpack_DUDV8;
       table[MESA_FORMAT_R_SNORM8] = unpack_R_SNORM8;
       table[MESA_FORMAT_R8G8_SNORM] = unpack_R8G8_SNORM;
       table[MESA_FORMAT_X8B8G8R8_SNORM] = unpack_X8B8G8R8_SNORM;
diff --git a/mesalib/src/mesa/main/formats.c b/mesalib/src/mesa/main/formats.c
index 5c670115e..1f20a9a6a 100644
--- a/mesalib/src/mesa/main/formats.c
+++ b/mesalib/src/mesa/main/formats.c
@@ -43,7 +43,7 @@ struct gl_format_info
    /**
     * Base format is one of GL_RED, GL_RG, GL_RGB, GL_RGBA, GL_ALPHA,
     * GL_LUMINANCE, GL_LUMINANCE_ALPHA, GL_INTENSITY, GL_YCBCR_MESA,
-    * GL_DEPTH_COMPONENT, GL_STENCIL_INDEX, GL_DEPTH_STENCIL, GL_DUDV_ATI.
+    * GL_DEPTH_COMPONENT, GL_STENCIL_INDEX, GL_DEPTH_STENCIL.
     */
    GLenum BaseFormat;
 
@@ -415,15 +415,6 @@ static struct gl_format_info format_info[MESA_FORMAT_COUNT] =
 
    /* Array unorm formats */
    {
-      MESA_FORMAT_DUDV8,
-      "MESA_FORMAT_DUDV8",
-      GL_DUDV_ATI,
-      GL_SIGNED_NORMALIZED,
-      0, 0, 0, 0,
-      0, 0, 0, 0, 0,
-      1, 1, 2
-   },
-   {
       MESA_FORMAT_A_UNORM8,        /* Name */
       "MESA_FORMAT_A_UNORM8",      /* StrName */
       GL_ALPHA,                    /* BaseFormat */
@@ -2576,11 +2567,6 @@ _mesa_format_to_type_and_comps(mesa_format format,
       *comps = 1;
       return;
 
-   case MESA_FORMAT_DUDV8:
-      *datatype = GL_BYTE;
-      *comps = 2;
-      return;
-
    case MESA_FORMAT_R_SNORM8:
    case MESA_FORMAT_A_SNORM8:
    case MESA_FORMAT_L_SNORM8:
@@ -3387,10 +3373,6 @@ _mesa_format_matches_format_and_type(mesa_format mesa_format,
    case MESA_FORMAT_RGBA_UINT32:
       return format == GL_RGBA_INTEGER && type == GL_UNSIGNED_INT && !swapBytes;
 
-   case MESA_FORMAT_DUDV8:
-      return (format == GL_DU8DV8_ATI || format == GL_DUDV_ATI) &&
-             type == GL_BYTE && littleEndian && !swapBytes;
-
    case MESA_FORMAT_R_SNORM8:
       return format == GL_RED && type == GL_BYTE;
    case MESA_FORMAT_R8G8_SNORM:
diff --git a/mesalib/src/mesa/main/formats.h b/mesalib/src/mesa/main/formats.h
index 185010e02..dc50bc830 100644
--- a/mesalib/src/mesa/main/formats.h
+++ b/mesalib/src/mesa/main/formats.h
@@ -211,8 +211,6 @@ typedef enum
    MESA_FORMAT_YCBCR,            /*                     YYYY YYYY UorV UorV */
    MESA_FORMAT_YCBCR_REV,        /*                     UorV UorV YYYY YYYY */
 
-   MESA_FORMAT_DUDV8,            /*                     DUDU DUDU DVDV DVDV */
-
    /* Array unorm formats */
    MESA_FORMAT_A_UNORM8,      /* ubyte[i] = A */
    MESA_FORMAT_A_UNORM16,     /* ushort[i] = A */
diff --git a/mesalib/src/mesa/main/glformats.c b/mesalib/src/mesa/main/glformats.c
index aee336e12..304d45298 100644
--- a/mesalib/src/mesa/main/glformats.c
+++ b/mesalib/src/mesa/main/glformats.c
@@ -189,8 +189,6 @@ _mesa_components_in_format(GLenum format)
    case GL_RG:
    case GL_YCBCR_MESA:
    case GL_DEPTH_STENCIL_EXT:
-   case GL_DUDV_ATI:
-   case GL_DU8DV8_ATI:
    case GL_RG_INTEGER:
       return 2;
 
@@ -817,22 +815,6 @@ _mesa_is_depth_or_stencil_format(GLenum format)
 
 
 /**
- * Test if the given image format is a dudv format.
- */
-GLboolean
-_mesa_is_dudv_format(GLenum format)
-{
-   switch (format) {
-      case GL_DUDV_ATI:
-      case GL_DU8DV8_ATI:
-         return GL_TRUE;
-      default:
-         return GL_FALSE;
-   }
-}
-
-
-/**
  * Test if an image format is a supported compressed format.
  * \param format the internal format token provided by the user.
  * \return GL_TRUE if compressed, GL_FALSE if uncompressed
@@ -1446,23 +1428,6 @@ _mesa_error_check_format_and_type(const struct gl_context *ctx,
          else
             return GL_INVALID_ENUM;
 
-      case GL_DUDV_ATI:
-      case GL_DU8DV8_ATI:
-         if (!ctx->Extensions.ATI_envmap_bumpmap)
-            return GL_INVALID_ENUM;
-         switch (type) {
-            case GL_BYTE:
-            case GL_UNSIGNED_BYTE:
-            case GL_SHORT:
-            case GL_UNSIGNED_SHORT:
-            case GL_INT:
-            case GL_UNSIGNED_INT:
-            case GL_FLOAT:
-               return GL_NO_ERROR;
-            default:
-               return GL_INVALID_ENUM;
-         }
-
       /* integer-valued formats */
       case GL_RED_INTEGER_EXT:
       case GL_GREEN_INTEGER_EXT:
diff --git a/mesalib/src/mesa/main/glformats.h b/mesalib/src/mesa/main/glformats.h
index ccbce2dab..d64114ec9 100644
--- a/mesalib/src/mesa/main/glformats.h
+++ b/mesalib/src/mesa/main/glformats.h
@@ -84,9 +84,6 @@ extern GLboolean
 _mesa_is_depth_or_stencil_format(GLenum format);
 
 extern GLboolean
-_mesa_is_dudv_format(GLenum format);
-
-extern GLboolean
 _mesa_is_compressed_format(struct gl_context *ctx, GLenum format);
 
 extern GLenum
diff --git a/mesalib/src/mesa/main/mtypes.h b/mesalib/src/mesa/main/mtypes.h
index 4762d9616..eaf377610 100644
--- a/mesalib/src/mesa/main/mtypes.h
+++ b/mesalib/src/mesa/main/mtypes.h
@@ -1342,8 +1342,6 @@ struct gl_texture_unit
    GLbitfield _GenFlags;	/**< Bitwise-OR of Gen[STRQ]._ModeBit */
 
    GLfloat LodBias;		/**< for biasing mipmap levels */
-   GLenum BumpTarget;
-   GLfloat RotMatrix[4]; /* 2x2 matrix */
 
    /** Current sampler object (GL_ARB_sampler_objects) */
    struct gl_sampler_object *Sampler;
@@ -1791,6 +1789,7 @@ struct gl_transform_feedback_output
    unsigned OutputRegister;
    unsigned OutputBuffer;
    unsigned NumComponents;
+   unsigned StreamId;
 
    /** offset (in DWORDs) of this output within the interleaved structure */
    unsigned DstOffset;
@@ -2177,6 +2176,7 @@ struct gl_geometry_program
                            GL_TRIANGLES, or GL_TRIANGLES_ADJACENCY_ARB */
    GLenum OutputType; /**< GL_POINTS, GL_LINE_STRIP or GL_TRIANGLE_STRIP */
    bool UsesEndPrimitive;
+   bool UsesStreams;
 };
 
 
@@ -2680,6 +2680,7 @@ struct gl_shader_program
       GLuint ClipDistanceArraySize; /**< Size of the gl_ClipDistance array, or
                                          0 if not present. */
       bool UsesEndPrimitive;
+      bool UsesStreams;
    } Geom;
 
    /** Vertex shader state */
@@ -2892,6 +2893,7 @@ struct gl_query_object
    GLboolean Active;   /**< inside Begin/EndQuery */
    GLboolean Ready;    /**< result is ready? */
    GLboolean EverBound;/**< has query object ever been bound */
+   GLuint Stream;      /**< The stream */
 };
 
 
@@ -2908,8 +2910,8 @@ struct gl_query_state
    struct gl_query_object *CondRenderQuery;
 
    /** GL_EXT_transform_feedback */
-   struct gl_query_object *PrimitivesGenerated;
-   struct gl_query_object *PrimitivesWritten;
+   struct gl_query_object *PrimitivesGenerated[MAX_VERTEX_STREAMS];
+   struct gl_query_object *PrimitivesWritten[MAX_VERTEX_STREAMS];
 
    /** GL_ARB_timer_query */
    struct gl_query_object *TimeElapsed;
@@ -3358,9 +3360,6 @@ struct gl_constants
     */
    GLuint UniformBooleanTrue;
 
-   /** Which texture units support GL_ATI_envmap_bumpmap as targets */
-   GLbitfield SupportedBumpUnits;
-
    /**
     * Maximum amount of time, measured in nanseconds, that the server can wait.
     */
@@ -3618,7 +3617,6 @@ struct gl_extensions
    GLboolean AMD_seamless_cubemap_per_texture;
    GLboolean AMD_vertex_shader_layer;
    GLboolean APPLE_object_purgeable;
-   GLboolean ATI_envmap_bumpmap;
    GLboolean ATI_texture_compression_3dc;
    GLboolean ATI_texture_mirror_once;
    GLboolean ATI_texture_env_combine3;
diff --git a/mesalib/src/mesa/main/pack.c b/mesalib/src/mesa/main/pack.c
index 1df656832..5ebaaf6e5 100644
--- a/mesalib/src/mesa/main/pack.c
+++ b/mesalib/src/mesa/main/pack.c
@@ -420,11 +420,6 @@ get_component_indexes(GLenum format,
       *blueIndex = 1;
       *alphaIndex = 0;
       break;
-   case GL_DU8DV8_ATI:
-   case GL_DUDV_ATI:
-      *redIndex = 0;
-      *greenIndex = 1;
-      break;
    default:
       assert(0 && "bad format in get_component_indexes()");
    }
@@ -1471,13 +1466,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                                          rgba[i][BCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_UBYTE(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_UBYTE(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -1628,13 +1616,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                                         rgba[i][BCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_BYTE_TEX(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_BYTE_TEX(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -1785,13 +1766,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                                           rgba[i][BCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_USHORT(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_USHORT(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -1942,13 +1916,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                                          rgba[i][BCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_SHORT_TEX(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_SHORT_TEX(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -2099,13 +2066,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                                         rgba[i][BCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_UINT(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_UINT(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -2185,13 +2145,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                      dst[i*4+3] = FLOAT_TO_INT(rgba[i][RCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = FLOAT_TO_INT(rgba[i][RCOMP]);
-                     dst[i*2+1] = FLOAT_TO_INT(rgba[i][GCOMP]);
-                  }
-                  break;
                case GL_RED_INTEGER_EXT:
                   for (i=0;i<n;i++) {
                      dst[i] = (GLint) rgba[i][RCOMP];
@@ -2342,13 +2295,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                      dst[i*4+3] = rgba[i][RCOMP];
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = rgba[i][RCOMP];
-                     dst[i*2+1] = rgba[i][GCOMP];
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -2428,13 +2374,6 @@ _mesa_pack_rgba_span_float(struct gl_context *ctx, GLuint n, GLfloat rgba[][4],
                      dst[i*4+3] = _mesa_float_to_half(rgba[i][RCOMP]);
                   }
                   break;
-               case GL_DUDV_ATI:
-               case GL_DU8DV8_ATI:
-                  for (i=0;i<n;i++) {
-                     dst[i*2+0] = _mesa_float_to_half(rgba[i][RCOMP]);
-                     dst[i*2+1] = _mesa_float_to_half(rgba[i][GCOMP]);
-                  }
-                  break;
                default:
                   _mesa_problem(ctx, "bad format in _mesa_pack_rgba_span\n");
             }
@@ -3095,13 +3034,6 @@ get_component_mapping(GLenum format,
       *bDst = 1;
       *aDst = 0;
       break;
-   case GL_DU8DV8_ATI:
-   case GL_DUDV_ATI:
-      *rSrc = 0;
-      *gSrc = 1;
-      *bSrc = -1;
-      *aSrc = -1;
-      break;
    default:
       _mesa_problem(NULL, "bad srcFormat %s in get_component_mapping",
                     _mesa_lookup_enum_by_nr(format));
@@ -3151,8 +3083,6 @@ extract_float_rgba(GLuint n, GLfloat rgba[][4],
           srcFormat == GL_RGBA ||
           srcFormat == GL_BGRA ||
           srcFormat == GL_ABGR_EXT ||
-          srcFormat == GL_DU8DV8_ATI ||
-          srcFormat == GL_DUDV_ATI ||
           srcFormat == GL_RED_INTEGER_EXT ||
           srcFormat == GL_GREEN_INTEGER_EXT ||
           srcFormat == GL_BLUE_INTEGER_EXT ||
@@ -3768,8 +3698,6 @@ extract_uint_rgba(GLuint n, GLuint rgba[][4],
           srcFormat == GL_RGBA ||
           srcFormat == GL_BGRA ||
           srcFormat == GL_ABGR_EXT ||
-          srcFormat == GL_DU8DV8_ATI ||
-          srcFormat == GL_DUDV_ATI ||
           srcFormat == GL_RED_INTEGER_EXT ||
           srcFormat == GL_RG_INTEGER ||
           srcFormat == GL_GREEN_INTEGER_EXT ||
@@ -4846,71 +4774,6 @@ _mesa_unpack_color_span_uint(struct gl_context *ctx,
 }
 
 
-
-/**
- * Similar to _mesa_unpack_color_span_float(), but for dudv data instead of rgba,
- * directly return GLbyte data, no transfer ops apply.
- */
-void
-_mesa_unpack_dudv_span_byte( struct gl_context *ctx,
-                             GLuint n, GLenum dstFormat, GLbyte dest[],
-                             GLenum srcFormat, GLenum srcType,
-                             const GLvoid *source,
-                             const struct gl_pixelstore_attrib *srcPacking,
-                             GLbitfield transferOps )
-{
-   ASSERT(dstFormat == GL_DUDV_ATI);
-   ASSERT(srcFormat == GL_DUDV_ATI ||
-	  srcFormat == GL_DU8DV8_ATI);
-
-   ASSERT(srcType == GL_UNSIGNED_BYTE ||
-          srcType == GL_BYTE ||
-          srcType == GL_UNSIGNED_SHORT ||
-          srcType == GL_SHORT ||
-          srcType == GL_UNSIGNED_INT ||
-          srcType == GL_INT ||
-          srcType == GL_HALF_FLOAT_ARB ||
-          srcType == GL_FLOAT);
-
-   /* general solution */
-   {
-      GLint dstComponents;
-      GLbyte *dst = dest;
-      GLuint i;
-      GLfloat (*rgba)[4] = malloc(4 * n * sizeof(GLfloat));
-
-      if (!rgba) {
-         _mesa_error(ctx, GL_OUT_OF_MEMORY, "pixel unpacking");
-         return;
-      }
-
-      dstComponents = _mesa_components_in_format( dstFormat );
-      /* source & dest image formats should have been error checked by now */
-      assert(dstComponents > 0);
-
-      /*
-       * Extract image data and convert to RGBA floats
-       */
-      extract_float_rgba(n, rgba, srcFormat, srcType, source,
-                         srcPacking->SwapBytes);
-
-
-      /* Now determine which color channels we need to produce.
-       * And determine the dest index (offset) within each color tuple.
-       */
-
-      /* Now pack results in the requested dstFormat */
-      for (i = 0; i < n; i++) {
-         /* not sure - need clamp[-1,1] here? */
-         dst[0] = FLOAT_TO_BYTE(rgba[i][RCOMP]);
-         dst[1] = FLOAT_TO_BYTE(rgba[i][GCOMP]);
-         dst += dstComponents;
-      }
-
-      free(rgba);
-   }
-}
-
 /*
  * Unpack a row of color index data from a client buffer according to
  * the pixel unpacking parameters.
diff --git a/mesalib/src/mesa/main/pack.h b/mesalib/src/mesa/main/pack.h
index ad357bae1..2173b652e 100644
--- a/mesalib/src/mesa/main/pack.h
+++ b/mesalib/src/mesa/main/pack.h
@@ -83,14 +83,6 @@ _mesa_unpack_color_span_uint(struct gl_context *ctx,
                              const struct gl_pixelstore_attrib *srcPacking);
 
 extern void
-_mesa_unpack_dudv_span_byte(struct gl_context *ctx,
-                            GLuint n, GLenum dstFormat, GLbyte dest[],
-                            GLenum srcFormat, GLenum srcType,
-                            const GLvoid *source,
-                            const struct gl_pixelstore_attrib *srcPacking,
-                            GLbitfield transferOps);
-
-extern void
 _mesa_unpack_index_span(struct gl_context *ctx, GLuint n,
                         GLenum dstType, GLvoid *dest,
                         GLenum srcType, const GLvoid *source,
diff --git a/mesalib/src/mesa/main/queryobj.c b/mesalib/src/mesa/main/queryobj.c
index 86e7c3ad0..932359c4e 100644
--- a/mesalib/src/mesa/main/queryobj.c
+++ b/mesalib/src/mesa/main/queryobj.c
@@ -144,11 +144,12 @@ _mesa_init_query_object_functions(struct dd_function_table *driver)
 
 
 /**
- * Return pointer to the query object binding point for the given target.
+ * Return pointer to the query object binding point for the given target and
+ * index.
  * \return NULL if invalid target, else the address of binding point
  */
 static struct gl_query_object **
-get_query_binding_point(struct gl_context *ctx, GLenum target)
+get_query_binding_point(struct gl_context *ctx, GLenum target, GLuint index)
 {
    switch (target) {
    case GL_SAMPLES_PASSED_ARB:
@@ -174,12 +175,12 @@ get_query_binding_point(struct gl_context *ctx, GLenum target)
          return NULL;
    case GL_PRIMITIVES_GENERATED:
       if (ctx->Extensions.EXT_transform_feedback)
-         return &ctx->Query.PrimitivesGenerated;
+         return &ctx->Query.PrimitivesGenerated[index];
       else
          return NULL;
    case GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN:
       if (ctx->Extensions.EXT_transform_feedback)
-         return &ctx->Query.PrimitivesWritten;
+         return &ctx->Query.PrimitivesWritten[index];
       else
          return NULL;
    default:
@@ -240,7 +241,7 @@ _mesa_DeleteQueries(GLsizei n, const GLuint *ids)
          if (q) {
             if (q->Active) {
                struct gl_query_object **bindpt;
-               bindpt = get_query_binding_point(ctx, q->Target);
+               bindpt = get_query_binding_point(ctx, q->Target, q->Stream);
                assert(bindpt); /* Should be non-null for active q. */
                if (bindpt) {
                   *bindpt = NULL;
@@ -313,7 +314,7 @@ _mesa_BeginQueryIndexed(GLenum target, GLuint index, GLuint id)
 
    FLUSH_VERTICES(ctx, 0);
 
-   bindpt = get_query_binding_point(ctx, target);
+   bindpt = get_query_binding_point(ctx, target, index);
    if (!bindpt) {
       _mesa_error(ctx, GL_INVALID_ENUM, "glBeginQuery{Indexed}(target)");
       return;
@@ -367,6 +368,7 @@ _mesa_BeginQueryIndexed(GLenum target, GLuint index, GLuint id)
    q->Result = 0;
    q->Ready = GL_FALSE;
    q->EverBound = GL_TRUE;
+   q->Stream = index;
 
    /* XXX should probably refcount query objects */
    *bindpt = q;
@@ -390,7 +392,7 @@ _mesa_EndQueryIndexed(GLenum target, GLuint index)
 
    FLUSH_VERTICES(ctx, 0);
 
-   bindpt = get_query_binding_point(ctx, target);
+   bindpt = get_query_binding_point(ctx, target, index);
    if (!bindpt) {
       _mesa_error(ctx, GL_INVALID_ENUM, "glEndQuery{Indexed}(target)");
       return;
@@ -517,7 +519,7 @@ _mesa_GetQueryIndexediv(GLenum target, GLuint index, GLenum pname,
       }
    }
    else {
-      bindpt = get_query_binding_point(ctx, target);
+      bindpt = get_query_binding_point(ctx, target, index);
       if (!bindpt) {
          _mesa_error(ctx, GL_INVALID_ENUM, "glGetQuery{Indexed}iv(target)");
          return;
diff --git a/mesalib/src/mesa/main/shaderapi.c b/mesalib/src/mesa/main/shaderapi.c
index 2ec2444da..2bbef35d3 100644
--- a/mesalib/src/mesa/main/shaderapi.c
+++ b/mesalib/src/mesa/main/shaderapi.c
@@ -1888,6 +1888,7 @@ _mesa_copy_linked_program_data(gl_shader_stage type,
       dst_gp->OutputType = src->Geom.OutputType;
       dst->UsesClipDistanceOut = src->Geom.UsesClipDistance;
       dst_gp->UsesEndPrimitive = src->Geom.UsesEndPrimitive;
+      dst_gp->UsesStreams = src->Geom.UsesStreams;
    }
       break;
    case MESA_SHADER_FRAGMENT: {
diff --git a/mesalib/src/mesa/main/shaderobj.c b/mesalib/src/mesa/main/shaderobj.c
index c6e852c87..b9feff4a4 100644
--- a/mesalib/src/mesa/main/shaderobj.c
+++ b/mesalib/src/mesa/main/shaderobj.c
@@ -249,6 +249,7 @@ _mesa_init_shader_program(struct gl_context *ctx, struct gl_shader_program *prog
    prog->Geom.InputType = GL_TRIANGLES;
    prog->Geom.OutputType = GL_TRIANGLE_STRIP;
    prog->Geom.UsesEndPrimitive = false;
+   prog->Geom.UsesStreams = false;
 
    prog->TransformFeedback.BufferMode = GL_INTERLEAVED_ATTRIBS;
 
diff --git a/mesalib/src/mesa/main/texenv.c b/mesalib/src/mesa/main/texenv.c
index 4848208f5..ec521e6c6 100644
--- a/mesalib/src/mesa/main/texenv.c
+++ b/mesalib/src/mesa/main/texenv.c
@@ -138,11 +138,6 @@ set_combiner_mode(struct gl_context *ctx,
       legal = (ctx->API == API_OPENGL_COMPAT &&
                ctx->Extensions.ATI_texture_env_combine3);
       break;
-   case GL_BUMP_ENVMAP_ATI:
-      legal = (ctx->API == API_OPENGL_COMPAT &&
-               ctx->Extensions.ATI_envmap_bumpmap &&
-               pname == GL_COMBINE_RGB);
-      break;
    default:
       legal = GL_FALSE;
    }
@@ -440,26 +435,6 @@ _mesa_TexEnvfv( GLenum target, GLenum pname, const GLfloat *param )
       case GL_ALPHA_SCALE:
          set_combiner_scale(ctx, texUnit, pname, param[0]);
 	 break;
-      case GL_BUMP_TARGET_ATI:
-         if (ctx->API != API_OPENGL_COMPAT || !ctx->Extensions.ATI_envmap_bumpmap) {
-	    _mesa_error( ctx, GL_INVALID_ENUM, "glTexEnv(pname=0x%x)", pname );
-	    return;
-	 }
-	 if ((iparam0 < GL_TEXTURE0) ||
-             (iparam0 > GL_TEXTURE31)) {
-	    /* spec doesn't say this but it seems logical */
-	    _mesa_error( ctx, GL_INVALID_ENUM, "glTexEnv(param=0x%x)", iparam0);
-	    return;
-	 }
-	 if (!((1 << (iparam0 - GL_TEXTURE0)) & ctx->Const.SupportedBumpUnits)) {
-	    _mesa_error( ctx, GL_INVALID_VALUE, "glTexEnv(param=0x%x)", iparam0);
-	    return;
-	 }
-	 else {
-	    FLUSH_VERTICES(ctx, _NEW_TEXTURE);
-	    texUnit->BumpTarget = iparam0;
-	 }
-	 break;
       default:
 	 _mesa_error( ctx, GL_INVALID_ENUM, "glTexEnv(pname)" );
 	 return;
@@ -641,16 +616,6 @@ get_texenvi(struct gl_context *ctx, const struct gl_texture_unit *texUnit,
       return 1 << texUnit->Combine.ScaleShiftRGB;
    case GL_ALPHA_SCALE:
       return 1 << texUnit->Combine.ScaleShiftA;
-   case GL_BUMP_TARGET_ATI:
-      /* spec doesn't say so, but I think this should be queryable */
-      if (ctx->API == API_OPENGL_COMPAT && ctx->Extensions.ATI_envmap_bumpmap) {
-         return texUnit->BumpTarget;
-      }
-      else {
-         _mesa_error(ctx, GL_INVALID_ENUM, "glGetTexEnvfv(pname)");
-      }
-      break;
-
    default:
       _mesa_error(ctx, GL_INVALID_ENUM, "glGetTexEnvfv(pname)");
       break;
@@ -784,166 +749,3 @@ _mesa_GetTexEnviv( GLenum target, GLenum pname, GLint *params )
    }
 }
 
-
-/**
- * Why does ATI_envmap_bumpmap require new entrypoints? Should just
- * reuse TexEnv ones...
- */
-void GLAPIENTRY
-_mesa_TexBumpParameterivATI( GLenum pname, const GLint *param )
-{
-   GLfloat p[4];
-   GET_CURRENT_CONTEXT(ctx);
-
-   if (!ctx->Extensions.ATI_envmap_bumpmap) {
-      /* This isn't an "official" error case, but let's tell the user
-       * that something's wrong.
-       */
-      _mesa_error(ctx, GL_INVALID_OPERATION, "glTexBumpParameterivATI");
-      return;
-   }
-
-   if (pname == GL_BUMP_ROT_MATRIX_ATI) {
-      /* hope that conversion is correct here */
-      p[0] = INT_TO_FLOAT( param[0] );
-      p[1] = INT_TO_FLOAT( param[1] );
-      p[2] = INT_TO_FLOAT( param[2] );
-      p[3] = INT_TO_FLOAT( param[3] );
-   }
-   else {
-      p[0] = (GLfloat) param[0];
-      p[1] = p[2] = p[3] = 0.0F;  /* init to zero, just to be safe */
-   }
-   _mesa_TexBumpParameterfvATI( pname, p );
-}
-
-
-void GLAPIENTRY
-_mesa_TexBumpParameterfvATI( GLenum pname, const GLfloat *param )
-{
-   struct gl_texture_unit *texUnit;
-   GET_CURRENT_CONTEXT(ctx);
-
-   if (!ctx->Extensions.ATI_envmap_bumpmap) {
-      _mesa_error(ctx, GL_INVALID_OPERATION, "glTexBumpParameterfvATI");
-      return;
-   }
-
-   texUnit = _mesa_get_current_tex_unit(ctx);
-
-   if (pname == GL_BUMP_ROT_MATRIX_ATI) {
-      if (TEST_EQ_4V(param, texUnit->RotMatrix))
-         return;
-      FLUSH_VERTICES(ctx, _NEW_TEXTURE);
-      COPY_4FV(texUnit->RotMatrix, param);
-   }
-   else {
-      _mesa_error( ctx, GL_INVALID_ENUM, "glTexBumpParameter(pname)" );
-      return;
-   }
-   /* Drivers might want to know about this, instead of dedicated function
-      just shove it into TexEnv where it really belongs anyway */
-   if (ctx->Driver.TexEnv) {
-      (*ctx->Driver.TexEnv)( ctx, 0, pname, param );
-   }
-}
-
-
-void GLAPIENTRY
-_mesa_GetTexBumpParameterivATI( GLenum pname, GLint *param )
-{
-   const struct gl_texture_unit *texUnit;
-   GLuint i;
-   GET_CURRENT_CONTEXT(ctx);
-
-   if (!ctx->Extensions.ATI_envmap_bumpmap) {
-      _mesa_error(ctx, GL_INVALID_OPERATION, "glGetTexBumpParameterivATI");
-      return;
-   }
-
-   texUnit = _mesa_get_current_tex_unit(ctx);
-
-   if (pname == GL_BUMP_ROT_MATRIX_SIZE_ATI) {
-      /* spec leaves open to support larger matrices.
-         Don't think anyone would ever want to use it
-         (and apps almost certainly would not understand it and
-         thus fail to submit matrices correctly) so hardcode this. */
-      *param = 4;
-   }
-   else if (pname == GL_BUMP_ROT_MATRIX_ATI) {
-      /* hope that conversion is correct here */
-      param[0] = FLOAT_TO_INT(texUnit->RotMatrix[0]);
-      param[1] = FLOAT_TO_INT(texUnit->RotMatrix[1]);
-      param[2] = FLOAT_TO_INT(texUnit->RotMatrix[2]);
-      param[3] = FLOAT_TO_INT(texUnit->RotMatrix[3]);
-   }
-   else if (pname == GL_BUMP_NUM_TEX_UNITS_ATI) {
-      GLint count = 0;
-      for (i = 0; i < ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits; i++) {
-         if (ctx->Const.SupportedBumpUnits & (1 << i)) {
-            count++;
-         }
-      }
-      *param = count;
-   }
-   else if (pname == GL_BUMP_TEX_UNITS_ATI) {
-      for (i = 0; i < ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits; i++) {
-         if (ctx->Const.SupportedBumpUnits & (1 << i)) {
-            *param++ = i + GL_TEXTURE0;
-         }
-      }
-   }
-   else {
-      _mesa_error( ctx, GL_INVALID_ENUM, "glGetTexBumpParameter(pname)" );
-      return;
-   }
-}
-
-
-void GLAPIENTRY
-_mesa_GetTexBumpParameterfvATI( GLenum pname, GLfloat *param )
-{
-   const struct gl_texture_unit *texUnit;
-   GLuint i;
-   GET_CURRENT_CONTEXT(ctx);
-
-   if (!ctx->Extensions.ATI_envmap_bumpmap) {
-      _mesa_error(ctx, GL_INVALID_OPERATION, "glGetTexBumpParameterfvATI");
-      return;
-   }
-
-   texUnit = _mesa_get_current_tex_unit(ctx);
-
-   if (pname == GL_BUMP_ROT_MATRIX_SIZE_ATI) {
-      /* spec leaves open to support larger matrices.
-         Don't think anyone would ever want to use it
-         (and apps might not understand it) so hardcode this. */
-      *param = 4.0F;
-   }
-   else if (pname == GL_BUMP_ROT_MATRIX_ATI) {
-      param[0] = texUnit->RotMatrix[0];
-      param[1] = texUnit->RotMatrix[1];
-      param[2] = texUnit->RotMatrix[2];
-      param[3] = texUnit->RotMatrix[3];
-   }
-   else if (pname == GL_BUMP_NUM_TEX_UNITS_ATI) {
-      GLint count = 0;
-      for (i = 0; i < ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits; i++) {
-         if (ctx->Const.SupportedBumpUnits & (1 << i)) {
-            count++;
-         }
-      }
-      *param = (GLfloat) count;
-   }
-   else if (pname == GL_BUMP_TEX_UNITS_ATI) {
-      for (i = 0; i < ctx->Const.Program[MESA_SHADER_FRAGMENT].MaxTextureImageUnits; i++) {
-         if (ctx->Const.SupportedBumpUnits & (1 << i)) {
-            *param++ = (GLfloat) (i + GL_TEXTURE0);
-         }
-      }
-   }
-   else {
-      _mesa_error( ctx, GL_INVALID_ENUM, "glGetTexBumpParameter(pname)" );
-      return;
-   }
-}
diff --git a/mesalib/src/mesa/main/texenv.h b/mesalib/src/mesa/main/texenv.h
index 46a7ec4e2..4aa3cb179 100644
--- a/mesalib/src/mesa/main/texenv.h
+++ b/mesalib/src/mesa/main/texenv.h
@@ -48,16 +48,4 @@ _mesa_TexEnvi( GLenum target, GLenum pname, GLint param );
 extern void GLAPIENTRY
 _mesa_TexEnviv( GLenum target, GLenum pname, const GLint *param );
 
-extern void GLAPIENTRY
-_mesa_TexBumpParameterivATI( GLenum pname, const GLint *param );
-
-extern void GLAPIENTRY
-_mesa_TexBumpParameterfvATI( GLenum pname, const GLfloat *param );
-
-extern void GLAPIENTRY
-_mesa_GetTexBumpParameterivATI( GLenum pname, GLint *param );
-
-extern void GLAPIENTRY
-_mesa_GetTexBumpParameterfvATI( GLenum pname, GLfloat *param );
-
 #endif /* TEXENV_H */
diff --git a/mesalib/src/mesa/main/texformat.c b/mesalib/src/mesa/main/texformat.c
index d43480482..c61a74859 100644
--- a/mesalib/src/mesa/main/texformat.c
+++ b/mesalib/src/mesa/main/texformat.c
@@ -440,11 +440,6 @@ _mesa_choose_tex_format(struct gl_context *ctx, GLenum target,
       ASSERT(ctx->TextureFormatSupported[MESA_FORMAT_Z32_FLOAT_S8X24_UINT]);
       return MESA_FORMAT_Z32_FLOAT_S8X24_UINT;
 
-   case GL_DUDV_ATI:
-   case GL_DU8DV8_ATI:
-      RETURN_IF_SUPPORTED(MESA_FORMAT_DUDV8);
-      break;
-
    case GL_RED_SNORM:
    case GL_R8_SNORM:
       RETURN_IF_SUPPORTED(MESA_FORMAT_R_SNORM8);
diff --git a/mesalib/src/mesa/main/texgetimage.c b/mesalib/src/mesa/main/texgetimage.c
index 8c0d3a18e..f1e09c986 100644
--- a/mesalib/src/mesa/main/texgetimage.c
+++ b/mesalib/src/mesa/main/texgetimage.c
@@ -859,11 +859,6 @@ getteximage_error_check(struct gl_context *ctx, GLenum target, GLint level,
       _mesa_error(ctx, GL_INVALID_OPERATION, "glGetTexImage(format mismatch)");
       return GL_TRUE;
    }
-   else if (_mesa_is_dudv_format(format)
-            && !_mesa_is_dudv_format(baseFormat)) {
-      _mesa_error(ctx, GL_INVALID_OPERATION, "glGetTexImage(format mismatch)");
-      return GL_TRUE;
-   }
    else if (_mesa_is_enum_format_integer(format) !=
             _mesa_is_format_integer(texImage->TexFormat)) {
       _mesa_error(ctx, GL_INVALID_OPERATION, "glGetTexImage(format mismatch)");
diff --git a/mesalib/src/mesa/main/teximage.c b/mesalib/src/mesa/main/teximage.c
index cab29c35d..a06959463 100644
--- a/mesalib/src/mesa/main/teximage.c
+++ b/mesalib/src/mesa/main/teximage.c
@@ -255,16 +255,6 @@ _mesa_base_tex_format( struct gl_context *ctx, GLint internalFormat )
       }
    }
 
-   if (ctx->Extensions.ATI_envmap_bumpmap) {
-      switch (internalFormat) {
-         case GL_DUDV_ATI:
-         case GL_DU8DV8_ATI:
-            return GL_DUDV_ATI;
-         default:
-            ; /* fallthrough */
-      }
-   }
-
    if (ctx->Extensions.EXT_texture_snorm) {
       switch (internalFormat) {
          case GL_RED_SNORM:
@@ -2157,8 +2147,7 @@ texture_error_check( struct gl_context *ctx,
    colorFormat = _mesa_is_color_format(format);
    if ((_mesa_is_color_format(internalFormat) && !colorFormat && !indexFormat) ||
        (is_internalFormat_depth_or_depthstencil != is_format_depth_or_depthstencil) ||
-       (_mesa_is_ycbcr_format(internalFormat) != _mesa_is_ycbcr_format(format)) ||
-       (_mesa_is_dudv_format(internalFormat) != _mesa_is_dudv_format(format))) {
+       (_mesa_is_ycbcr_format(internalFormat) != _mesa_is_ycbcr_format(format))) {
       _mesa_error(ctx, GL_INVALID_OPERATION,
                   "glTexImage%dD(incompatible internalFormat = %s, format = %s)",
                   dimensions, _mesa_lookup_enum_by_nr(internalFormat),
diff --git a/mesalib/src/mesa/main/texstate.c b/mesalib/src/mesa/main/texstate.c
index 19508cf39..e0f085218 100644
--- a/mesalib/src/mesa/main/texstate.c
+++ b/mesalib/src/mesa/main/texstate.c
@@ -91,10 +91,6 @@ _mesa_copy_texture_state( const struct gl_context *src, struct gl_context *dst )
       /* GL_EXT_texture_env_combine */
       dst->Texture.Unit[u].Combine = src->Texture.Unit[u].Combine;
 
-      /* GL_ATI_envmap_bumpmap - need this? */
-      dst->Texture.Unit[u].BumpTarget = src->Texture.Unit[u].BumpTarget;
-      COPY_4V(dst->Texture.Unit[u].RotMatrix, src->Texture.Unit[u].RotMatrix);
-
       /*
        * XXX strictly speaking, we should compare texture names/ids and
        * bind textures in the dest context according to id.  For now, only
@@ -188,7 +184,6 @@ calculate_derived_texenv( struct gl_tex_env_combine_state *state,
    case GL_RG:
    case GL_RGB:
    case GL_YCBCR_MESA:
-   case GL_DUDV_ATI:
       state->SourceA[0] = GL_PREVIOUS;
       break;
       
@@ -230,7 +225,6 @@ calculate_derived_texenv( struct gl_tex_env_combine_state *state,
       case GL_RG:
       case GL_RGB:
       case GL_YCBCR_MESA:
-      case GL_DUDV_ATI:
 	 mode_rgb = GL_REPLACE;
 	 break;
       case GL_RGBA:
@@ -259,7 +253,6 @@ calculate_derived_texenv( struct gl_tex_env_combine_state *state,
       case GL_LUMINANCE_ALPHA:
       case GL_RGBA:
       case GL_YCBCR_MESA:
-      case GL_DUDV_ATI:
 	 state->SourceRGB[2] = GL_TEXTURE;
 	 state->SourceA[2]   = GL_TEXTURE;
 	 state->SourceRGB[0] = GL_CONSTANT;
@@ -443,10 +436,6 @@ update_tex_combine(struct gl_context *ctx, struct gl_texture_unit *texUnit)
    case GL_MODULATE_SUBTRACT_ATI:
       combine->_NumArgsRGB = 3;
       break;
-   case GL_BUMP_ENVMAP_ATI:
-      /* no real arguments for this case */
-      combine->_NumArgsRGB = 0;
-      break;
    default:
       combine->_NumArgsRGB = 0;
       _mesa_problem(ctx, "invalid RGB combine mode in update_texture_state");
@@ -848,7 +837,6 @@ init_texture_unit( struct gl_context *ctx, GLuint unit )
    texUnit->Combine = default_combine_state;
    texUnit->_EnvMode = default_combine_state;
    texUnit->_CurrentCombine = & texUnit->_EnvMode;
-   texUnit->BumpTarget = GL_TEXTURE0;
 
    texUnit->TexGenEnabled = 0x0;
    texUnit->GenS.Mode = GL_EYE_LINEAR;
@@ -870,9 +858,6 @@ init_texture_unit( struct gl_context *ctx, GLuint unit )
    ASSIGN_4V( texUnit->GenR.EyePlane, 0.0, 0.0, 0.0, 0.0 );
    ASSIGN_4V( texUnit->GenQ.EyePlane, 0.0, 0.0, 0.0, 0.0 );
 
-   /* no mention of this in spec, but maybe id matrix expected? */
-   ASSIGN_4V( texUnit->RotMatrix, 1.0, 0.0, 0.0, 1.0 );
-
    /* initialize current texture object ptrs to the shared default objects */
    for (tex = 0; tex < NUM_TEXTURE_TARGETS; tex++) {
       _mesa_reference_texobj(&texUnit->CurrentTex[tex],
diff --git a/mesalib/src/mesa/main/texstore.c b/mesalib/src/mesa/main/texstore.c
index cb81f3fde..d363f9faa 100644
--- a/mesalib/src/mesa/main/texstore.c
+++ b/mesalib/src/mesa/main/texstore.c
@@ -2061,83 +2061,6 @@ _mesa_texstore_ycbcr(TEXSTORE_PARAMS)
    return GL_TRUE;
 }
 
-static GLboolean
-_mesa_texstore_dudv8(TEXSTORE_PARAMS)
-{
-   const GLboolean littleEndian = _mesa_little_endian();
-   const GLuint texelBytes = _mesa_get_format_bytes(dstFormat);
-
-   ASSERT(dstFormat == MESA_FORMAT_DUDV8);
-   ASSERT(texelBytes == 2);
-   ASSERT(ctx->Extensions.ATI_envmap_bumpmap);
-   ASSERT((srcFormat == GL_DU8DV8_ATI) ||
-	  (srcFormat == GL_DUDV_ATI));
-   ASSERT(baseInternalFormat == GL_DUDV_ATI);
-
-   if (srcType == GL_BYTE) {
-      GLubyte dstmap[4];
-
-      /* dstmap - how to swizzle from RGBA to dst format:
-       */
-      if (littleEndian) {
-	 dstmap[0] = 0;
-	 dstmap[1] = 3;
-      }
-      else {
-	 dstmap[0] = 3;
-	 dstmap[1] = 0;
-      }
-      dstmap[2] = ZERO;		/* ? */
-      dstmap[3] = ONE;		/* ? */
-      
-      _mesa_swizzle_ubyte_image(ctx, dims,
-				GL_LUMINANCE_ALPHA, /* hack */
-				GL_UNSIGNED_BYTE, /* hack */
-				GL_LUMINANCE_ALPHA, /* hack */
-				dstmap, 2,
-				dstRowStride, dstSlices,
-				srcWidth, srcHeight, srcDepth, srcAddr,
-				srcPacking);      
-   }   
-   else {
-      /* general path - note this is defined for 2d textures only */
-      const GLint components = _mesa_components_in_format(baseInternalFormat);
-      const GLint srcStride = _mesa_image_row_stride(srcPacking, srcWidth,
-                                                     srcFormat, srcType);
-      GLbyte *tempImage, *dst, *src;
-      GLint row;
-
-      tempImage = malloc(srcWidth * srcHeight * srcDepth
-                                          * components * sizeof(GLbyte));
-      if (!tempImage)
-         return GL_FALSE;
-
-      src = (GLbyte *) _mesa_image_address(dims, srcPacking, srcAddr,
-                                           srcWidth, srcHeight,
-                                           srcFormat, srcType,
-                                           0, 0, 0);
-
-      dst = tempImage;
-      for (row = 0; row < srcHeight; row++) {
-         _mesa_unpack_dudv_span_byte(ctx, srcWidth, baseInternalFormat,
-                                     dst, srcFormat, srcType, src,
-                                     srcPacking, 0);
-         dst += srcWidth * components;
-         src += srcStride;
-      }
- 
-      src = tempImage;
-      dst = (GLbyte *) dstSlices[0];
-      for (row = 0; row < srcHeight; row++) {
-         memcpy(dst, src, srcWidth * texelBytes);
-         dst += dstRowStride;
-         src += srcWidth * texelBytes;
-      }
-      free((void *) tempImage);
-   }
-   return GL_TRUE;
-}
-
 
 /**
  * Store a texture in a signed normalized 8-bit format.
@@ -3723,7 +3646,6 @@ _mesa_get_texstore_func(mesa_format format)
       table[MESA_FORMAT_R_FLOAT16] = _mesa_texstore_rgba_float16;
       table[MESA_FORMAT_RG_FLOAT32] = _mesa_texstore_rgba_float32;
       table[MESA_FORMAT_RG_FLOAT16] = _mesa_texstore_rgba_float16;
-      table[MESA_FORMAT_DUDV8] = _mesa_texstore_dudv8;
       table[MESA_FORMAT_R_SNORM8] = _mesa_texstore_snorm8;
       table[MESA_FORMAT_R8G8_SNORM] = _mesa_texstore_snorm88;
       table[MESA_FORMAT_X8B8G8R8_SNORM] = _mesa_texstore_signed_rgbx8888;
diff --git a/mesalib/src/mesa/main/vdpau.c b/mesalib/src/mesa/main/vdpau.c
index d97459393..975b812cd 100644
--- a/mesalib/src/mesa/main/vdpau.c
+++ b/mesalib/src/mesa/main/vdpau.c
@@ -132,6 +132,11 @@ register_surface(struct gl_context *ctx, GLboolean isOutput,
    }
 
    surf = CALLOC_STRUCT( vdp_surface );
+   if (surf == NULL) {
+      _mesa_error_no_memory("VDPAURegisterSurfaceNV");
+      return (GLintptr)NULL;
+   }
+
    surf->vdpSurface = vdpSurface;
    surf->target = target;
    surf->access = GL_READ_WRITE;
@@ -140,6 +145,12 @@ register_surface(struct gl_context *ctx, GLboolean isOutput,
    for (i = 0; i < numTextureNames; ++i) {
       struct gl_texture_object *tex;
       tex  = _mesa_lookup_texture(ctx, textureNames[i]);
+      if (tex == NULL) {
+         free(surf);
+         _mesa_error(ctx, GL_INVALID_OPERATION,
+                     "VDPAURegisterSurfaceNV(texture ID not found)");
+         return (GLintptr)NULL;
+      }
 
       _mesa_lock_texture(ctx, tex);
 
diff --git a/mesalib/src/mesa/program/prog_statevars.c b/mesalib/src/mesa/program/prog_statevars.c
index 5dda8e28d..be5ddb106 100644
--- a/mesalib/src/mesa/program/prog_statevars.c
+++ b/mesalib/src/mesa/program/prog_statevars.c
@@ -593,28 +593,6 @@ _mesa_fetch_state(struct gl_context *ctx, const gl_state_index state[],
          }
          return;
 
-      case STATE_ROT_MATRIX_0:
-         {
-            const int unit = (int) state[2];
-            GLfloat *rotMat22 = ctx->Texture.Unit[unit].RotMatrix;
-            value[0] = rotMat22[0]; 
-            value[1] = rotMat22[2];
-            value[2] = 0.0;
-            value[3] = 0.0;
-         }
-         return;
-
-      case STATE_ROT_MATRIX_1:
-         {
-            const int unit = (int) state[2];
-            GLfloat *rotMat22 = ctx->Texture.Unit[unit].RotMatrix;
-            value[0] = rotMat22[1];
-            value[1] = rotMat22[3];
-            value[2] = 0.0;
-            value[3] = 0.0;
-         }
-         return;
-
       /* XXX: make sure new tokens added here are also handled in the 
        * _mesa_program_state_flags() switch, below.
        */
@@ -706,8 +684,6 @@ _mesa_program_state_flags(const gl_state_index state[STATE_LENGTH])
          return _NEW_MODELVIEW;
 
       case STATE_TEXRECT_SCALE:
-      case STATE_ROT_MATRIX_0:
-      case STATE_ROT_MATRIX_1:
 	 return _NEW_TEXTURE;
       case STATE_FOG_PARAMS_OPTIMIZED:
 	 return _NEW_FOG;
@@ -933,12 +909,6 @@ append_token(char *dst, gl_state_index k)
    case STATE_FB_WPOS_Y_TRANSFORM:
       append(dst, "FbWposYTransform");
       break;
-   case STATE_ROT_MATRIX_0:
-      append(dst, "rotMatrixRow0");
-      break;
-   case STATE_ROT_MATRIX_1:
-      append(dst, "rotMatrixRow1");
-      break;
    default:
       /* probably STATE_INTERNAL_DRIVER+i (driver private state) */
       append(dst, "driverState");
diff --git a/mesalib/src/mesa/program/prog_statevars.h b/mesalib/src/mesa/program/prog_statevars.h
index 23a9f48c3..6333e6328 100644
--- a/mesalib/src/mesa/program/prog_statevars.h
+++ b/mesalib/src/mesa/program/prog_statevars.h
@@ -128,8 +128,6 @@ typedef enum gl_state_index_ {
    STATE_PT_BIAS,               /**< Pixel transfer RGBA bias */
    STATE_FB_SIZE,               /**< (width-1, height-1, 0, 0) */
    STATE_FB_WPOS_Y_TRANSFORM,   /**< (1, 0, -1, height) if a FBO is bound, (-1, height, 1, 0) otherwise */
-   STATE_ROT_MATRIX_0,          /**< ATI_envmap_bumpmap, rot matrix row 0 */
-   STATE_ROT_MATRIX_1,          /**< ATI_envmap_bumpmap, rot matrix row 1 */
    STATE_INTERNAL_DRIVER	/* first available state index for drivers (must be last) */
 } gl_state_index;
 
diff --git a/mesalib/src/mesa/program/program.c b/mesalib/src/mesa/program/program.c
index 988def1a3..aedce3e3c 100644
--- a/mesalib/src/mesa/program/program.c
+++ b/mesalib/src/mesa/program/program.c
@@ -553,6 +553,7 @@ _mesa_clone_program(struct gl_context *ctx, const struct gl_program *prog)
          gpc->Invocations = gp->Invocations;
          gpc->OutputType = gp->OutputType;
          gpc->UsesEndPrimitive = gp->UsesEndPrimitive;
+         gpc->UsesStreams = gp->UsesStreams;
       }
       break;
    default:
diff --git a/mesalib/src/mesa/swrast/s_texcombine.c b/mesalib/src/mesa/swrast/s_texcombine.c
index c1a152aa8..def5eb19e 100644
--- a/mesalib/src/mesa/swrast/s_texcombine.c
+++ b/mesalib/src/mesa/swrast/s_texcombine.c
@@ -429,16 +429,6 @@ texture_combine( struct gl_context *ctx, GLuint unit,
                               arg1[i][BCOMP]) * scaleRGB;
 	 }
          break;
-      case GL_BUMP_ENVMAP_ATI:
-         /* this produces a fixed rgba color, and the coord calc is done elsewhere */
-         for (i = 0; i < n; i++) {
-            /* rgba result is 0,0,0,1 */
-            rgba[i][RCOMP] = 0.0;
-            rgba[i][GCOMP] = 0.0;
-            rgba[i][BCOMP] = 0.0;
-            rgba[i][ACOMP] = 1.0;
-	 }
-         goto end; /* no alpha processing */
       default:
          _mesa_problem(ctx, "invalid combine mode");
       }
@@ -653,78 +643,13 @@ _swrast_texture_span( struct gl_context *ctx, SWspan *span )
       }
    }
 
-   /* First must sample all bump maps */
-   for (unit = 0; unit < ctx->Const.MaxTextureUnits; unit++) {
-      const struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
-
-      if (texUnit->_Current &&
-         texUnit->_CurrentCombine->ModeRGB == GL_BUMP_ENVMAP_ATI) {
-         const GLfloat (*texcoords)[4] = (const GLfloat (*)[4])
-            span->array->attribs[VARYING_SLOT_TEX0 + unit];
-         float4_array targetcoords =
-            span->array->attribs[VARYING_SLOT_TEX0 +
-               ctx->Texture.Unit[unit].BumpTarget - GL_TEXTURE0];
-
-         const struct gl_sampler_object *samp = _mesa_get_samplerobj(ctx, unit);
-         GLfloat *lambda = span->array->lambda[unit];
-         float4_array texels = get_texel_array(swrast, unit);
-         GLuint i;
-         GLfloat rotMatrix00 = ctx->Texture.Unit[unit].RotMatrix[0];
-         GLfloat rotMatrix01 = ctx->Texture.Unit[unit].RotMatrix[1];
-         GLfloat rotMatrix10 = ctx->Texture.Unit[unit].RotMatrix[2];
-         GLfloat rotMatrix11 = ctx->Texture.Unit[unit].RotMatrix[3];
-
-         /* adjust texture lod (lambda) */
-         if (span->arrayMask & SPAN_LAMBDA) {
-            if (texUnit->LodBias + samp->LodBias != 0.0F) {
-               /* apply LOD bias, but don't clamp yet */
-               const GLfloat bias = CLAMP(texUnit->LodBias + samp->LodBias,
-                                          -ctx->Const.MaxTextureLodBias,
-                                          ctx->Const.MaxTextureLodBias);
-               GLuint i;
-               for (i = 0; i < span->end; i++) {
-                  lambda[i] += bias;
-               }
-            }
-
-            if (samp->MinLod != -1000.0 ||
-                samp->MaxLod != 1000.0) {
-               /* apply LOD clamping to lambda */
-               const GLfloat min = samp->MinLod;
-               const GLfloat max = samp->MaxLod;
-               GLuint i;
-               for (i = 0; i < span->end; i++) {
-                  GLfloat l = lambda[i];
-                  lambda[i] = CLAMP(l, min, max);
-               }
-            }
-         }
-
-         /* Sample the texture (span->end = number of fragments) */
-         swrast->TextureSample[unit]( ctx, samp,
-                                      ctx->Texture.Unit[unit]._Current,
-                                      span->end, texcoords, lambda, texels );
-
-         /* manipulate the span values of the bump target
-            not sure this can work correctly even ignoring
-            the problem that channel is unsigned */
-         for (i = 0; i < span->end; i++) {
-            targetcoords[i][0] += (texels[i][0] * rotMatrix00 + texels[i][1] *
-                                  rotMatrix01) / targetcoords[i][3];
-            targetcoords[i][1] += (texels[i][0] * rotMatrix10 + texels[i][1] *
-                                  rotMatrix11) / targetcoords[i][3];
-         }
-      }
-   }
-
    /*
     * Must do all texture sampling before combining in order to
     * accomodate GL_ARB_texture_env_crossbar.
     */
    for (unit = 0; unit < ctx->Const.MaxTextureUnits; unit++) {
       const struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
-      if (texUnit->_Current &&
-          texUnit->_CurrentCombine->ModeRGB != GL_BUMP_ENVMAP_ATI) {
+      if (texUnit->_Current) {
          const GLfloat (*texcoords)[4] = (const GLfloat (*)[4])
             span->array->attribs[VARYING_SLOT_TEX0 + unit];
          const struct gl_texture_object *curObj = texUnit->_Current;
diff --git a/mesalib/src/mesa/swrast/s_texfetch.c b/mesalib/src/mesa/swrast/s_texfetch.c
index 90a514c16..e508368c8 100644
--- a/mesalib/src/mesa/swrast/s_texfetch.c
+++ b/mesalib/src/mesa/swrast/s_texfetch.c
@@ -210,7 +210,6 @@ texfetch_funcs[] =
    },
    FETCH_FUNCS(YCBCR),
    FETCH_FUNCS(YCBCR_REV),
-   FETCH_FUNCS(DUDV8),
 
    /* Array unorm formats */
    FETCH_FUNCS(A_UNORM8),
diff --git a/mesalib/src/mesa/swrast/s_texfetch_tmp.h b/mesalib/src/mesa/swrast/s_texfetch_tmp.h
index d48e39bfd..deda59246 100644
--- a/mesalib/src/mesa/swrast/s_texfetch_tmp.h
+++ b/mesalib/src/mesa/swrast/s_texfetch_tmp.h
@@ -888,22 +888,6 @@ FETCH(RGBA_UINT32)(const struct swrast_texture_image *texImage,
 }
 
 
-/**
- * This format by definition produces 0,0,0,1 as rgba values,
- * however we'll return the dudv values as rg and fix up elsewhere.
- */
-static void
-FETCH(DUDV8)(const struct swrast_texture_image *texImage,
-             GLint i, GLint j, GLint k, GLfloat *texel)
-{
-   const GLbyte *src = TEXEL_ADDR(GLbyte, texImage, i, j, k, 2);
-   texel[RCOMP] = BYTE_TO_FLOAT(src[0]);
-   texel[GCOMP] = BYTE_TO_FLOAT(src[1]);
-   texel[BCOMP] = 0;
-   texel[ACOMP] = 0;
-}
-
-
 static void
 FETCH(R_SNORM8)(const struct swrast_texture_image *texImage,
                 GLint i, GLint j, GLint k, GLfloat *texel)