fontconfig mesa pixman xserver git update 4 July 2014

xserver          commit a61ca6f006d70343c88fe45206fae0669d1e8971
pixman           commit 6d2cf40166d81bfc63108504c8022dc4fec37ff5
fontconfig       commit 5b22776999b6052afe0e829b1a0c0935bbe1e9f7
mesa             commit 9a37eb8adb6558a4abf47774b583cb582a0ae116

16 files changed, 128 insertions, 126 deletions

diff --git a/mesalib/src/mesa/drivers/common/meta.c b/mesalib/src/mesa/drivers/common/meta.c
index f313f5645..f1f57297c 100644
--- a/mesalib/src/mesa/drivers/common/meta.c
+++ b/mesalib/src/mesa/drivers/common/meta.c
@@ -2592,8 +2592,7 @@ _mesa_meta_setup_texture_coords(GLenum faceTarget,
             coord = coords3;
             break;
          default:
-            assert(0);
-            unreachable();
+            unreachable("not reached");
          }
 
          coord[3] = (float) (slice / 6);
diff --git a/mesalib/src/mesa/drivers/dri/common/dri_util.c b/mesalib/src/mesa/drivers/dri/common/dri_util.c
index 8506f5356..6c78928ee 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, bool *val)
+dri2ConfigQueryb(__DRIscreen *screen, const char *var, unsigned char *val)
 {
    if (!driCheckOption(&screen->optionCache, var, DRI_BOOL))
       return -1;
diff --git a/mesalib/src/mesa/drivers/dri/common/xmlconfig.c b/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
index 5091c115b..58d0e0656 100644
--- a/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
+++ b/mesalib/src/mesa/drivers/dri/common/xmlconfig.c
@@ -285,7 +285,7 @@ static float strToF (const XML_Char *string, const XML_Char **tail) {
 }
 
 /** \brief Parse a value of a given type. */
-static bool parseValue (driOptionValue *v, driOptionType type,
+static unsigned char parseValue (driOptionValue *v, driOptionType type,
 			     const XML_Char *string) {
     const XML_Char *tail = NULL;
   /* skip leading white-space */
@@ -328,7 +328,7 @@ static bool parseValue (driOptionValue *v, driOptionType type,
 }
 
 /** \brief Parse a list of ranges of type info->type. */
-static bool parseRanges (driOptionInfo *info, const XML_Char *string) {
+static unsigned char parseRanges (driOptionInfo *info, const XML_Char *string) {
     XML_Char *cp, *range;
     uint32_t nRanges, i;
     driOptionRange *ranges;
@@ -1003,13 +1003,13 @@ void driDestroyOptionCache (driOptionCache *cache) {
     free(cache->values);
 }
 
-bool driCheckOption (const driOptionCache *cache, const char *name,
+unsigned char driCheckOption (const driOptionCache *cache, const char *name,
 			  driOptionType type) {
     uint32_t i = findOption (cache, name);
     return cache->info[i].name != NULL && cache->info[i].type == type;
 }
 
-bool driQueryOptionb (const driOptionCache *cache, const char *name) {
+unsigned char 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);
diff --git a/mesalib/src/mesa/drivers/dri/common/xmlconfig.h b/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
index 386ddf19d..af0323485 100644
--- a/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
+++ b/mesalib/src/mesa/drivers/dri/common/xmlconfig.h
@@ -30,8 +30,6 @@
 #ifndef __XMLCONFIG_H
 #define __XMLCONFIG_H
 
-#include <stdbool.h>
-
 #define STRING_CONF_MAXLEN 25
 
 /** \brief Option data types */
@@ -41,7 +39,7 @@ typedef enum driOptionType {
 
 /** \brief Option value */
 typedef union driOptionValue {
-    bool _bool; /**< \brief Boolean */
+    unsigned char _bool; /**< \brief Boolean */
     int _int;      /**< \brief Integer or Enum */
     float _float;  /**< \brief Floating-point */
     char *_string;   /**< \brief String */
@@ -114,11 +112,11 @@ void driDestroyOptionInfo (driOptionCache *info);
 void driDestroyOptionCache (driOptionCache *cache);
 
 /** \brief Check if there exists a certain option */
-bool driCheckOption (const driOptionCache *cache, const char *name,
+unsigned char driCheckOption (const driOptionCache *cache, const char *name,
 			  driOptionType type);
 
 /** \brief Query a boolean option value */
-bool driQueryOptionb (const driOptionCache *cache, const char *name);
+unsigned char driQueryOptionb (const driOptionCache *cache, const char *name);
 /** \brief Query an integer option value */
 int driQueryOptioni (const driOptionCache *cache, const char *name);
 /** \brief Query a floating-point option value */
diff --git a/mesalib/src/mesa/main/compiler.h b/mesalib/src/mesa/main/compiler.h
index 600691724..79d8740e5 100644
--- a/mesalib/src/mesa/main/compiler.h
+++ b/mesalib/src/mesa/main/compiler.h
@@ -253,15 +253,23 @@ static INLINE GLuint CPU_TO_LE32(GLuint x)
  * function" warnings.
  */
 #if __GNUC__ >= 4 && __GNUC_MINOR__ >= 5
-#define unreachable() __builtin_unreachable()
+#define unreachable(str)    \
+do {                        \
+   assert(!str);            \
+   __builtin_unreachable(); \
+} while (0)
 #elif (defined(__clang__) && defined(__has_builtin))
 # if __has_builtin(__builtin_unreachable)
-#  define unreachable() __builtin_unreachable()
+#  define unreachable(str)  \
+do {                        \
+   assert(!str);            \
+   __builtin_unreachable(); \
+} while (0)
 # endif
 #endif
 
 #ifndef unreachable
-#define unreachable()
+#define unreachable(str)
 #endif
 
 /*
diff --git a/mesalib/src/mesa/main/extensions.c b/mesalib/src/mesa/main/extensions.c
index 6f2536170..92e3f0d68 100644
--- a/mesalib/src/mesa/main/extensions.c
+++ b/mesalib/src/mesa/main/extensions.c
@@ -323,6 +323,7 @@ static const struct extension extension_table[] = {
    { "GL_AMD_shader_stencil_export",               o(ARB_shader_stencil_export),               GL,             2009 },
    { "GL_AMD_shader_trinary_minmax",               o(dummy_true),                              GL,             2012 },
    { "GL_AMD_vertex_shader_layer",                 o(AMD_vertex_shader_layer),                 GLC,            2012 },
+   { "GL_AMD_vertex_shader_viewport_index",        o(AMD_vertex_shader_viewport_index),        GLC,            2012 },
    { "GL_APPLE_object_purgeable",                  o(APPLE_object_purgeable),                  GL,             2006 },
    { "GL_APPLE_packed_pixels",                     o(dummy_true),                              GLL,            2002 },
    { "GL_APPLE_texture_max_level",                 o(dummy_true),                                   ES1 | ES2, 2009 },
diff --git a/mesalib/src/mesa/main/mtypes.h b/mesalib/src/mesa/main/mtypes.h
index eaf377610..a7126fd55 100644
--- a/mesalib/src/mesa/main/mtypes.h
+++ b/mesalib/src/mesa/main/mtypes.h
@@ -2560,7 +2560,7 @@ struct gl_uniform_block
    GLuint Binding;
 
    /**
-    * Minimum size of a buffer object to back this uniform buffer
+    * Minimum size (in bytes) of a buffer object to back this uniform buffer
     * (GL_UNIFORM_BLOCK_DATA_SIZE).
     */
    GLuint UniformBufferSize;
@@ -3616,6 +3616,7 @@ struct gl_extensions
    GLboolean AMD_performance_monitor;
    GLboolean AMD_seamless_cubemap_per_texture;
    GLboolean AMD_vertex_shader_layer;
+   GLboolean AMD_vertex_shader_viewport_index;
    GLboolean APPLE_object_purgeable;
    GLboolean ATI_texture_compression_3dc;
    GLboolean ATI_texture_mirror_once;
diff --git a/mesalib/src/mesa/main/shader_query.cpp b/mesalib/src/mesa/main/shader_query.cpp
index 36d1d9cc0..426774316 100644
--- a/mesalib/src/mesa/main/shader_query.cpp
+++ b/mesalib/src/mesa/main/shader_query.cpp
@@ -126,8 +126,8 @@ _mesa_GetActiveAttrib(GLhandleARB program, GLuint desired_index,
    exec_list *const ir = shProg->_LinkedShaders[MESA_SHADER_VERTEX]->ir;
    unsigned current_index = 0;
 
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       if (!is_active_attrib(var))
          continue;
@@ -236,8 +236,8 @@ _mesa_GetAttribLocation(GLhandleARB program, const GLcharARB * name)
       return -1;
 
    exec_list *ir = shProg->_LinkedShaders[MESA_SHADER_VERTEX]->ir;
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       /* The extra check against VERT_ATTRIB_GENERIC0 is because
        * glGetAttribLocation cannot be used on "conventional" attributes.
@@ -274,8 +274,8 @@ _mesa_count_active_attribs(struct gl_shader_program *shProg)
    exec_list *const ir = shProg->_LinkedShaders[MESA_SHADER_VERTEX]->ir;
    unsigned i = 0;
 
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       if (!is_active_attrib(var))
          continue;
@@ -298,8 +298,8 @@ _mesa_longest_attribute_name_length(struct gl_shader_program *shProg)
    exec_list *const ir = shProg->_LinkedShaders[MESA_SHADER_VERTEX]->ir;
    size_t longest = 0;
 
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       if (var == NULL
 	  || var->data.mode != ir_var_shader_in
@@ -400,8 +400,8 @@ _mesa_GetFragDataIndex(GLuint program, const GLchar *name)
       return -1;
 
    exec_list *ir = shProg->_LinkedShaders[MESA_SHADER_FRAGMENT]->ir;
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       /* The extra check against FRAG_RESULT_DATA0 is because
        * glGetFragDataLocation cannot be used on "conventional" attributes.
@@ -456,8 +456,8 @@ _mesa_GetFragDataLocation(GLuint program, const GLchar *name)
       return -1;
 
    exec_list *ir = shProg->_LinkedShaders[MESA_SHADER_FRAGMENT]->ir;
-   foreach_list(node, ir) {
-      const ir_variable *const var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, ir) {
+      const ir_variable *const var = node->as_variable();
 
       /* The extra check against FRAG_RESULT_DATA0 is because
        * glGetFragDataLocation cannot be used on "conventional" attributes.
diff --git a/mesalib/src/mesa/program/ir_to_mesa.cpp b/mesalib/src/mesa/program/ir_to_mesa.cpp
index 59cf1232a..1109051e9 100644
--- a/mesalib/src/mesa/program/ir_to_mesa.cpp
+++ b/mesalib/src/mesa/program/ir_to_mesa.cpp
@@ -663,11 +663,7 @@ ir_to_mesa_visitor::get_temp(const glsl_type *type)
 variable_storage *
 ir_to_mesa_visitor::find_variable_storage(const ir_variable *var)
 {
-   variable_storage *entry;
-
-   foreach_list(node, &this->variables) {
-      entry = (variable_storage *) node;
-
+   foreach_in_list(variable_storage, entry, &this->variables) {
       if (entry->var == var)
 	 return entry;
    }
@@ -801,9 +797,7 @@ ir_to_mesa_visitor::visit(ir_function *ir)
 
       assert(sig);
 
-      foreach_list(node, &sig->body) {
-	 ir_instruction *ir = (ir_instruction *) node;
-
+      foreach_in_list(ir_instruction, ir, &sig->body) {
 	 ir->accept(this);
       }
    }
@@ -1868,8 +1862,7 @@ ir_to_mesa_visitor::visit(ir_constant *ir)
       src_reg temp_base = get_temp(ir->type);
       dst_reg temp = dst_reg(temp_base);
 
-      foreach_list(node, &ir->components) {
-	 ir_constant *field_value = (ir_constant *) node;
+      foreach_in_list(ir_constant, field_value, &ir->components) {
 	 int size = type_size(field_value->type);
 
 	 assert(size > 0);
@@ -2338,9 +2331,7 @@ set_branchtargets(ir_to_mesa_visitor *v,
 	 mesa_instructions[loop_stack[loop_stack_pos]].BranchTarget = i;
 	 break;
       case OPCODE_CAL:
-	 foreach_list(n, &v->function_signatures) {
-	    function_entry *entry = (function_entry *) n;
-
+	 foreach_in_list(function_entry, entry, &v->function_signatures) {
 	    if (entry->sig_id == mesa_instructions[i].BranchTarget) {
 	       mesa_instructions[i].BranchTarget = entry->inst;
 	       break;
@@ -2495,8 +2486,8 @@ _mesa_generate_parameters_list_for_uniforms(struct gl_shader_program
 {
    add_uniform_to_shader add(shader_program, params, sh->Stage);
 
-   foreach_list(node, sh->ir) {
-      ir_variable *var = ((ir_instruction *) node)->as_variable();
+   foreach_in_list(ir_instruction, node, sh->ir) {
+      ir_variable *var = node->as_variable();
 
       if ((var == NULL) || (var->data.mode != ir_var_uniform)
 	  || var->is_in_uniform_block() || (strncmp(var->name, "gl_", 3) == 0))
@@ -2621,9 +2612,7 @@ ir_to_mesa_visitor::copy_propagate(void)
    int *acp_level = rzalloc_array(mem_ctx, int, this->next_temp * 4);
    int level = 0;
 
-   foreach_list(node, &this->instructions) {
-      ir_to_mesa_instruction *inst = (ir_to_mesa_instruction *) node;
-
+   foreach_in_list(ir_to_mesa_instruction, inst, &this->instructions) {
       assert(inst->dst.file != PROGRAM_TEMPORARY
 	     || inst->dst.index < this->next_temp);
 
@@ -2826,7 +2815,7 @@ get_mesa_program(struct gl_context *ctx,
    prog->NumTemporaries = v.next_temp;
 
    int num_instructions = 0;
-   foreach_list(node, &v.instructions) {
+   foreach_in_list(ir_instruction, node, &v.instructions) {
       num_instructions++;
    }
 
@@ -2842,9 +2831,7 @@ get_mesa_program(struct gl_context *ctx,
     */
    mesa_inst = mesa_instructions;
    i = 0;
-   foreach_list(node, &v.instructions) {
-      const ir_to_mesa_instruction *inst = (ir_to_mesa_instruction *) node;
-
+   foreach_in_list(const ir_to_mesa_instruction, inst, &v.instructions) {
       mesa_inst->Opcode = inst->op;
       mesa_inst->CondUpdate = inst->cond_update;
       if (inst->saturate)
diff --git a/mesalib/src/mesa/state_tracker/st_cb_bufferobjects.c b/mesalib/src/mesa/state_tracker/st_cb_bufferobjects.c
index 02624617b..3b4d28d47 100644
--- a/mesalib/src/mesa/state_tracker/st_cb_bufferobjects.c
+++ b/mesalib/src/mesa/state_tracker/st_cb_bufferobjects.c
@@ -225,6 +225,9 @@ st_bufferobj_data(struct gl_context *ctx,
    case GL_UNIFORM_BUFFER:
       bind = PIPE_BIND_CONSTANT_BUFFER;
       break;
+   case GL_DRAW_INDIRECT_BUFFER:
+      bind = PIPE_BIND_COMMAND_ARGS_BUFFER;
+      break;
    default:
       bind = 0;
    }
diff --git a/mesalib/src/mesa/state_tracker/st_cb_clear.c b/mesalib/src/mesa/state_tracker/st_cb_clear.c
index 371f7fcda..4bfa8d75a 100644
--- a/mesalib/src/mesa/state_tracker/st_cb_clear.c
+++ b/mesalib/src/mesa/state_tracker/st_cb_clear.c
@@ -136,7 +136,7 @@ set_vertex_shader_layered(struct st_context *st)
    struct pipe_context *pipe = st->pipe;
 
    if (!pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_INSTANCEID) ||
-       !pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_VS_LAYER)) {
+       !pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_VS_LAYER_VIEWPORT)) {
       assert(!"Got layered clear, but the VS layer output is unsupported");
       set_vertex_shader(st);
       return;
diff --git a/mesalib/src/mesa/state_tracker/st_cb_queryobj.c b/mesalib/src/mesa/state_tracker/st_cb_queryobj.c
index 78a737094..489f537d8 100644
--- a/mesalib/src/mesa/state_tracker/st_cb_queryobj.c
+++ b/mesalib/src/mesa/state_tracker/st_cb_queryobj.c
@@ -132,13 +132,13 @@ st_BeginQuery(struct gl_context *ctx, struct gl_query_object *q)
        type == PIPE_QUERY_TIMESTAMP) {
       /* Determine time elapsed by emitting two timestamp queries. */
       if (!stq->pq_begin) {
-         stq->pq_begin = pipe->create_query(pipe, type);
+         stq->pq_begin = pipe->create_query(pipe, type, 0);
          stq->type = type;
       }
       pipe->end_query(pipe, stq->pq_begin);
    } else {
       if (!stq->pq) {
-         stq->pq = pipe->create_query(pipe, type);
+         stq->pq = pipe->create_query(pipe, type, q->Stream);
          stq->type = type;
       }
       if (stq->pq) {
@@ -164,7 +164,7 @@ st_EndQuery(struct gl_context *ctx, struct gl_query_object *q)
    if ((q->Target == GL_TIMESTAMP ||
         q->Target == GL_TIME_ELAPSED) &&
        !stq->pq) {
-      stq->pq = pipe->create_query(pipe, PIPE_QUERY_TIMESTAMP);
+      stq->pq = pipe->create_query(pipe, PIPE_QUERY_TIMESTAMP, 0);
       stq->type = PIPE_QUERY_TIMESTAMP;
    }
 
diff --git a/mesalib/src/mesa/state_tracker/st_draw.c b/mesalib/src/mesa/state_tracker/st_draw.c
index c8189faad..64d6ef525 100644
--- a/mesalib/src/mesa/state_tracker/st_draw.c
+++ b/mesalib/src/mesa/state_tracker/st_draw.c
@@ -244,6 +244,14 @@ st_draw_vbo(struct gl_context *ctx,
       }
    }
 
+   if (indirect) {
+      info.indirect = st_buffer_object(indirect)->buffer;
+
+      /* Primitive restart is not handled by the VBO module in this case. */
+      info.primitive_restart = ctx->Array._PrimitiveRestart;
+      info.restart_index = ctx->Array.RestartIndex;
+   }
+
    /* do actual drawing */
    for (i = 0; i < nr_prims; i++) {
       info.mode = translate_prim(ctx, prims[i].mode);
@@ -256,6 +264,7 @@ st_draw_vbo(struct gl_context *ctx,
          info.min_index = info.start;
          info.max_index = info.start + info.count - 1;
       }
+      info.indirect_offset = prims[i].indirect_offset;
 
       if (ST_DEBUG & DEBUG_DRAW) {
          debug_printf("st/draw: mode %s  start %u  count %u  indexed %d\n",
@@ -265,7 +274,7 @@ st_draw_vbo(struct gl_context *ctx,
                       info.indexed);
       }
 
-      if (info.count_from_stream_output) {
+      if (info.count_from_stream_output || info.indirect) {
          cso_draw_vbo(st->cso_context, &info);
       }
       else if (info.primitive_restart) {
diff --git a/mesalib/src/mesa/state_tracker/st_extensions.c b/mesalib/src/mesa/state_tracker/st_extensions.c
index e93804689..4207cb64a 100644
--- a/mesalib/src/mesa/state_tracker/st_extensions.c
+++ b/mesalib/src/mesa/state_tracker/st_extensions.c
@@ -287,6 +287,11 @@ void st_init_limits(struct st_context *st)
       screen->get_param(screen, PIPE_CAP_MAX_STREAM_OUTPUT_SEPARATE_COMPONENTS);
    c->MaxTransformFeedbackInterleavedComponents =
       screen->get_param(screen, PIPE_CAP_MAX_STREAM_OUTPUT_INTERLEAVED_COMPONENTS);
+   c->MaxVertexStreams =
+      MAX2(1, screen->get_param(screen, PIPE_CAP_MAX_VERTEX_STREAMS));
+
+   /* The vertex stream must fit into pipe_stream_output_info::stream */
+   assert(c->MaxVertexStreams <= 4);
 
    c->StripTextureBorder = GL_TRUE;
 
@@ -427,6 +432,7 @@ void st_init_extensions(struct st_context *st)
       { o(ARB_texture_multisample),          PIPE_CAP_TEXTURE_MULTISAMPLE              },
       { o(ARB_texture_query_lod),            PIPE_CAP_TEXTURE_QUERY_LOD                },
       { o(ARB_sample_shading),               PIPE_CAP_SAMPLE_SHADING                   },
+      { o(ARB_draw_indirect),                PIPE_CAP_DRAW_INDIRECT                    }
    };
 
    /* Required: render target and sampler support */
@@ -618,7 +624,7 @@ void st_init_extensions(struct st_context *st)
    /* This extension needs full OpenGL 3.2, but we don't know if that's
     * supported at this point. Only check the GLSL version. */
    if (ctx->Const.GLSLVersion >= 150 &&
-       screen->get_param(screen, PIPE_CAP_TGSI_VS_LAYER)) {
+       screen->get_param(screen, PIPE_CAP_TGSI_VS_LAYER_VIEWPORT)) {
       ctx->Extensions.AMD_vertex_shader_layer = GL_TRUE;
    }
 
@@ -809,6 +815,9 @@ void st_init_extensions(struct st_context *st)
          ctx->Const.ViewportBounds.Min = -16384.0;
          ctx->Const.ViewportBounds.Max = 16384.0;
          ctx->Extensions.ARB_viewport_array = GL_TRUE;
+         ctx->Extensions.ARB_fragment_layer_viewport = GL_TRUE;
+         if (ctx->Extensions.AMD_vertex_shader_layer)
+            ctx->Extensions.AMD_vertex_shader_viewport_index = GL_TRUE;
       }
    }
    if (ctx->Const.MaxProgramTextureGatherComponents > 0)
diff --git a/mesalib/src/mesa/state_tracker/st_glsl_to_tgsi.cpp b/mesalib/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
index cac1e0fe2..9e1943139 100644
--- a/mesalib/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
+++ b/mesalib/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
@@ -919,9 +919,7 @@ glsl_to_tgsi_visitor::add_constant(gl_register_file file,
       /* Search immediate storage to see if we already have an identical
        * immediate that we can use instead of adding a duplicate entry.
        */
-      foreach_list(node, &this->immediates) {
-         entry = (immediate_storage *) node;
-         
+      foreach_in_list(immediate_storage, entry, &this->immediates) {
          if (entry->size == size &&
              entry->type == datatype &&
              !memcmp(entry->values, values, size * sizeof(gl_constant_value))) {
@@ -1061,11 +1059,7 @@ variable_storage *
 glsl_to_tgsi_visitor::find_variable_storage(ir_variable *var)
 {
    
-   variable_storage *entry;
-
-   foreach_list(node, &this->variables) {
-      entry = (variable_storage *) node;
-
+   foreach_in_list(variable_storage, entry, &this->variables) {
       if (entry->var == var)
          return entry;
    }
@@ -1202,9 +1196,7 @@ glsl_to_tgsi_visitor::visit(ir_function *ir)
 
       assert(sig);
 
-      foreach_list(node, &sig->body) {
-         ir_instruction *ir = (ir_instruction *) node;
-
+      foreach_in_list(ir_instruction, ir, &sig->body) {
          ir->accept(this);
       }
    }
@@ -1971,9 +1963,17 @@ glsl_to_tgsi_visitor::visit(ir_expression *ir)
       assert(ir->type->is_vector() || ir->type->is_scalar());
 
       if (const_offset_ir) {
-         index_reg = st_src_reg_for_int(const_offset / 16);
-      } else {
-         emit(ir, TGSI_OPCODE_USHR, st_dst_reg(index_reg), op[1], st_src_reg_for_int(4));
+         /* Constant index into constant buffer */
+         cbuf.reladdr = NULL;
+         cbuf.index = const_offset / 16;
+         cbuf.has_index2 = true;
+      }
+      else {
+         /* Relative/variable index into constant buffer */
+         emit(ir, TGSI_OPCODE_USHR, st_dst_reg(index_reg), op[1],
+              st_src_reg_for_int(4));
+         cbuf.reladdr = ralloc(mem_ctx, st_src_reg);
+         memcpy(cbuf.reladdr, &index_reg, sizeof(index_reg));
       }
 
       cbuf.swizzle = swizzle_for_size(ir->type->vector_elements);
@@ -1982,9 +1982,6 @@ glsl_to_tgsi_visitor::visit(ir_expression *ir)
                                     const_offset % 16 / 4,
                                     const_offset % 16 / 4);
 
-      cbuf.reladdr = ralloc(mem_ctx, st_src_reg);
-      memcpy(cbuf.reladdr, &index_reg, sizeof(index_reg));
-
       if (ir->type->base_type == GLSL_TYPE_BOOL) {
          emit(ir, TGSI_OPCODE_USNE, result_dst, cbuf, st_src_reg_for_int(0));
       } else {
@@ -2549,8 +2546,7 @@ glsl_to_tgsi_visitor::visit(ir_constant *ir)
       st_src_reg temp_base = get_temp(ir->type);
       st_dst_reg temp = st_dst_reg(temp_base);
 
-      foreach_list(node, &ir->components) {
-         ir_constant *field_value = (ir_constant *) node;
+      foreach_in_list(ir_constant, field_value, &ir->components) {
          int size = type_size(field_value->type);
 
          assert(size > 0);
@@ -2664,11 +2660,7 @@ glsl_to_tgsi_visitor::visit(ir_constant *ir)
 function_entry *
 glsl_to_tgsi_visitor::get_function_signature(ir_function_signature *sig)
 {
-   function_entry *entry;
-
-   foreach_list(node, &this->function_signatures) {
-      entry = (function_entry *) node;
-
+   foreach_in_list_use_after(function_entry, entry, &this->function_signatures) {
       if (entry->sig == sig)
          return entry;
    }
@@ -2679,8 +2671,7 @@ glsl_to_tgsi_visitor::get_function_signature(ir_function_signature *sig)
    entry->bgn_inst = NULL;
 
    /* Allocate storage for all the parameters. */
-   foreach_list(node, &sig->parameters) {
-      ir_variable *param = (ir_variable *) node;
+   foreach_in_list(ir_variable, param, &sig->parameters) {
       variable_storage *storage;
 
       storage = find_variable_storage(param);
@@ -3139,14 +3130,18 @@ void
 glsl_to_tgsi_visitor::visit(ir_emit_vertex *ir)
 {
    assert(this->prog->Target == GL_GEOMETRY_PROGRAM_NV);
-   emit(ir, TGSI_OPCODE_EMIT);
+
+   ir->stream->accept(this);
+   emit(ir, TGSI_OPCODE_EMIT, undef_dst, this->result);
 }
 
 void
 glsl_to_tgsi_visitor::visit(ir_end_primitive *ir)
 {
    assert(this->prog->Target == GL_GEOMETRY_PROGRAM_NV);
-   emit(ir, TGSI_OPCODE_ENDPRIM);
+
+   ir->stream->accept(this);
+   emit(ir, TGSI_OPCODE_ENDPRIM, undef_dst, this->result);
 }
 
 glsl_to_tgsi_visitor::glsl_to_tgsi_visitor()
@@ -3190,9 +3185,7 @@ count_resources(glsl_to_tgsi_visitor *v, gl_program *prog)
 {
    v->samplers_used = 0;
 
-   foreach_list(node, &v->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &v->instructions) {
       if (is_tex_instruction(inst->op)) {
          v->samplers_used |= 1 << inst->sampler;
 
@@ -3336,8 +3329,7 @@ glsl_to_tgsi_visitor::simplify_cmp(void)
    memset(tempWrites, 0, sizeof(unsigned) * MAX_TEMPS);
    memset(outputWrites, 0, sizeof(outputWrites));
 
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       unsigned prevWriteMask = 0;
 
       /* Give up if we encounter relative addressing or flow control. */
@@ -3382,8 +3374,7 @@ glsl_to_tgsi_visitor::simplify_cmp(void)
 void
 glsl_to_tgsi_visitor::rename_temp_register(int index, int new_index)
 {
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       unsigned j;
       
       for (j=0; j < num_inst_src_regs(inst->op); j++) {
@@ -3413,9 +3404,7 @@ glsl_to_tgsi_visitor::get_first_temp_read(int index)
    int loop_start = -1; /* index of the first active BGNLOOP (if any) */
    unsigned i = 0, j;
    
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-      
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       for (j=0; j < num_inst_src_regs(inst->op); j++) {
          if (inst->src[j].file == PROGRAM_TEMPORARY && 
              inst->src[j].index == index) {
@@ -3451,9 +3440,7 @@ glsl_to_tgsi_visitor::get_first_temp_write(int index)
    int loop_start = -1; /* index of the first active BGNLOOP (if any) */
    int i = 0;
    
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-      
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       if (inst->dst.file == PROGRAM_TEMPORARY && inst->dst.index == index) {
          return (depth == 0) ? i : loop_start;
       }
@@ -3480,9 +3467,7 @@ glsl_to_tgsi_visitor::get_last_temp_read(int index)
    int last = -1; /* index of last instruction that reads the temporary */
    unsigned i = 0, j;
    
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-      
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       for (j=0; j < num_inst_src_regs(inst->op); j++) {
          if (inst->src[j].file == PROGRAM_TEMPORARY && 
              inst->src[j].index == index) {
@@ -3516,9 +3501,7 @@ glsl_to_tgsi_visitor::get_last_temp_write(int index)
    int last = -1; /* index of last instruction that writes to the temporary */
    int i = 0;
    
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-      
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       if (inst->dst.file == PROGRAM_TEMPORARY && inst->dst.index == index)
          last = (depth == 0) ? i : -2;
       
@@ -3565,9 +3548,7 @@ glsl_to_tgsi_visitor::copy_propagate(void)
    int *acp_level = rzalloc_array(mem_ctx, int, this->next_temp * 4);
    int level = 0;
 
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       assert(inst->dst.file != PROGRAM_TEMPORARY
              || inst->dst.index < this->next_temp);
 
@@ -3761,9 +3742,7 @@ glsl_to_tgsi_visitor::eliminate_dead_code(void)
    int level = 0;
    int removed = 0;
 
-   foreach_list(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &this->instructions) {
       assert(inst->dst.file != PROGRAM_TEMPORARY
              || inst->dst.index < this->next_temp);
       
@@ -3889,9 +3868,7 @@ glsl_to_tgsi_visitor::eliminate_dead_code(void)
    /* Now actually remove the instructions that are completely dead and update
     * the writemask of other instructions with dead channels.
     */
-   foreach_list_safe(node, &this->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
-      
+   foreach_in_list_safe(glsl_to_tgsi_instruction, inst, &this->instructions) {
       if (!inst->dead_mask || !inst->dst.writemask)
          continue;
       else if ((inst->dst.writemask & ~inst->dead_mask) == 0) {
@@ -4077,8 +4054,7 @@ get_pixel_transfer_visitor(struct st_fragment_program *fp,
 
    /* Now copy the instructions from the original glsl_to_tgsi_visitor into the
     * new visitor. */
-   foreach_list(node, &original->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &original->instructions) {
       glsl_to_tgsi_instruction *newinst;
       st_src_reg src_regs[3];
 
@@ -4162,8 +4138,7 @@ get_bitmap_visitor(struct st_fragment_program *fp,
 
    /* Now copy the instructions from the original glsl_to_tgsi_visitor into the
     * new visitor. */
-   foreach_list(node, &original->instructions) {
-      glsl_to_tgsi_instruction *inst = (glsl_to_tgsi_instruction *) node;
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &original->instructions) {
       glsl_to_tgsi_instruction *newinst;
       st_src_reg src_regs[3];
 
@@ -4396,7 +4371,7 @@ src_register(struct st_translate *t,
    case PROGRAM_CONSTANT:       /* ie, immediate */
       if (index2D) {
          struct ureg_src src;
-         src = ureg_src_register(TGSI_FILE_CONSTANT, 0);
+         src = ureg_src_register(TGSI_FILE_CONSTANT, index);
          src.Dimension = 1;
          src.DimensionIndex = index2D;
          return src;
@@ -4484,7 +4459,10 @@ translate_src(struct st_translate *t, const st_src_reg *src_reg)
 {
    struct ureg_src src = src_register(t, src_reg->file, src_reg->index, src_reg->index2D);
 
-   if (t->procType == TGSI_PROCESSOR_GEOMETRY && src_reg->has_index2) {
+   if (src_reg->has_index2) {
+      /* 2D indexes occur with geometry shader inputs (attrib, vertex)
+       * and UBO constant buffers (buffer, position).
+       */
       src = src_register(t, src_reg->file, src_reg->index, src_reg->index2D);
       if (src_reg->reladdr2)
          src = ureg_src_dimension_indirect(src, ureg_src(t->address[1]),
@@ -5115,7 +5093,14 @@ st_translate_program(
       unsigned num_ubos = program->shader->NumUniformBlocks;
 
       for (i = 0; i < num_ubos; i++) {
-         ureg_DECL_constant2D(t->ureg, 0, program->shader->UniformBlocks[i].UniformBufferSize / 4, i + 1);
+         unsigned size =
+            program->shader_program->UniformBlocks[i].UniformBufferSize;
+         unsigned num_const_vecs = (size + 15) / 16;
+         unsigned first, last;
+         assert(num_const_vecs > 0);
+         first = 0;
+         last = num_const_vecs > 0 ? num_const_vecs - 1 : 0;
+         ureg_DECL_constant2D(t->ureg, first, last, i + 1);
       }
    }
    
@@ -5128,8 +5113,7 @@ st_translate_program(
       goto out;
    }
    i = 0;
-   foreach_list(node, &program->immediates) {
-      immediate_storage *imm = (immediate_storage *) node;
+   foreach_in_list(immediate_storage, imm, &program->immediates) {
       assert(i < program->num_immediates);
       t->immediates[i++] = emit_immediate(t, imm->values, imm->type, imm->size);
    }
@@ -5144,9 +5128,9 @@ st_translate_program(
 
    /* Emit each instruction in turn:
     */
-   foreach_list(n, &program->instructions) {
+   foreach_in_list(glsl_to_tgsi_instruction, inst, &program->instructions) {
       set_insn_start(t, ureg_get_instruction_number(ureg));
-      compile_tgsi_instruction(t, (glsl_to_tgsi_instruction *) n, clamp_color);
+      compile_tgsi_instruction(t, inst, clamp_color);
    }
 
    /* Fix up all emitted labels:
@@ -5257,9 +5241,7 @@ get_mesa_program(struct gl_context *ctx,
    do {
       progress = GL_FALSE;
 
-      foreach_list(node, &v->function_signatures) {
-         function_entry *entry = (function_entry *) node;
-
+      foreach_in_list(function_entry, entry, &v->function_signatures) {
          if (!entry->bgn_inst) {
             v->current_function = entry;
 
@@ -5524,6 +5506,7 @@ st_translate_stream_output_info(glsl_to_tgsi_visitor *glsl_to_tgsi,
       so->output[i].num_components = info->Outputs[i].NumComponents;
       so->output[i].output_buffer = info->Outputs[i].OutputBuffer;
       so->output[i].dst_offset = info->Outputs[i].DstOffset;
+      so->output[i].stream = info->Outputs[i].StreamId;
    }
 
    for (i = 0; i < PIPE_MAX_SO_BUFFERS; i++) {
diff --git a/mesalib/src/mesa/state_tracker/st_program.c b/mesalib/src/mesa/state_tracker/st_program.c
index 1df411c3b..3570557fe 100644
--- a/mesalib/src/mesa/state_tracker/st_program.c
+++ b/mesalib/src/mesa/state_tracker/st_program.c
@@ -262,6 +262,10 @@ st_prepare_vertex_program(struct gl_context *ctx,
             stvp->output_semantic_name[slot] = TGSI_SEMANTIC_LAYER;
             stvp->output_semantic_index[slot] = 0;
             break;
+         case VARYING_SLOT_VIEWPORT:
+            stvp->output_semantic_name[slot] = TGSI_SEMANTIC_VIEWPORT_INDEX;
+            stvp->output_semantic_index[slot] = 0;
+            break;
 
          case VARYING_SLOT_TEX0:
          case VARYING_SLOT_TEX1: