Merge remote-tracking branch 'origin/master' into vulkan

26 files changed, 266 insertions, 285 deletions

diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h
index ae37cbf7325..ab1afdbc475 100644
--- a/src/compiler/nir/nir.h
+++ b/src/compiler/nir/nir.h
@@ -88,9 +88,9 @@ typedef enum {
    nir_var_local,
    nir_var_uniform,
    nir_var_shader_storage,
-   nir_var_shared,
    nir_var_system_value,
    nir_var_param,
+   nir_var_shared,
 } nir_variable_mode;
 
 /**
diff --git a/src/compiler/nir/nir_lower_io.c b/src/compiler/nir/nir_lower_io.c
index 84e353775cf..9d502ee4e9c 100644
--- a/src/compiler/nir/nir_lower_io.c
+++ b/src/compiler/nir/nir_lower_io.c
@@ -361,10 +361,11 @@ nir_lower_io_block(nir_block *block, void *void_state)
          atomic->const_index[0] =
             intrin->variables[0]->var->data.driver_location;
 
-         nir_src_copy(&atomic->src[1], &intrin->src[0], atomic);
-
-         if (intrin->intrinsic == nir_intrinsic_var_atomic_comp_swap)
-            nir_src_copy(&atomic->src[2], &intrin->src[1], atomic);
+         for (unsigned i = 0;
+              i < nir_op_infos[intrin->intrinsic].num_inputs;
+              i++) {
+            nir_src_copy(&atomic->src[i+1], &intrin->src[i], atomic);
+         }
 
          if (intrin->dest.is_ssa) {
             nir_ssa_dest_init(&atomic->instr, &atomic->dest,
diff --git a/src/gallium/drivers/vc4/vc4_opt_dead_code.c b/src/gallium/drivers/vc4/vc4_opt_dead_code.c
index 9e79a2d1b2e..44d0779c611 100644
--- a/src/gallium/drivers/vc4/vc4_opt_dead_code.c
+++ b/src/gallium/drivers/vc4/vc4_opt_dead_code.c
@@ -86,12 +86,8 @@ qir_opt_dead_code(struct vc4_compile *c)
         /* Whether we're eliminating texture setup currently. */
         bool dce_tex = false;
 
-        struct list_head *node, *t;
-        for (node = c->instructions.prev, t = node->prev;
-             &c->instructions != node;
-             node = t, t = t->prev) {
-                struct qinst *inst = (struct qinst *)node;
-
+        list_for_each_entry_safe_rev(struct qinst, inst, &c->instructions,
+                                     link) {
                 if (inst->dst.file == QFILE_TEMP &&
                     !used[inst->dst.index] &&
                     !inst->sf &&
diff --git a/src/gallium/drivers/vc4/vc4_program.c b/src/gallium/drivers/vc4/vc4_program.c
index 81e8e9150d6..f5826d85174 100644
--- a/src/gallium/drivers/vc4/vc4_program.c
+++ b/src/gallium/drivers/vc4/vc4_program.c
@@ -1184,8 +1184,11 @@ emit_frag_end(struct vc4_compile *c)
                 color = qir_uniform_ui(c, 0);
         }
 
-        if (c->discard.file != QFILE_NULL)
-                qir_TLB_DISCARD_SETUP(c, c->discard);
+        uint32_t discard_cond = QPU_COND_ALWAYS;
+        if (c->discard.file != QFILE_NULL) {
+                qir_SF(c, c->discard);
+                discard_cond = QPU_COND_ZS;
+        }
 
         if (c->fs_key->stencil_enabled) {
                 qir_TLB_STENCIL_SETUP(c, qir_uniform(c, QUNIFORM_STENCIL, 0));
@@ -1209,14 +1212,18 @@ emit_frag_end(struct vc4_compile *c)
                 } else {
                         z = qir_FRAG_Z(c);
                 }
-                qir_TLB_Z_WRITE(c, z);
+                struct qinst *inst = qir_TLB_Z_WRITE(c, z);
+                inst->cond = discard_cond;
         }
 
         if (!c->msaa_per_sample_output) {
-                qir_TLB_COLOR_WRITE(c, color);
+                struct qinst *inst = qir_TLB_COLOR_WRITE(c, color);
+                inst->cond = discard_cond;
         } else {
-                for (int i = 0; i < VC4_MAX_SAMPLES; i++)
-                        qir_TLB_COLOR_WRITE_MS(c, c->sample_colors[i]);
+                for (int i = 0; i < VC4_MAX_SAMPLES; i++) {
+                        struct qinst *inst = qir_TLB_COLOR_WRITE_MS(c, c->sample_colors[i]);
+                        inst->cond = discard_cond;
+                }
         }
 }
 
diff --git a/src/gallium/drivers/vc4/vc4_qir.c b/src/gallium/drivers/vc4/vc4_qir.c
index 65f0067c61e..e73e3899410 100644
--- a/src/gallium/drivers/vc4/vc4_qir.c
+++ b/src/gallium/drivers/vc4/vc4_qir.c
@@ -69,7 +69,6 @@ static const struct qir_op_info qir_op_info[] = {
         [QOP_RSQ] = { "rsq", 1, 1, false, true },
         [QOP_EXP2] = { "exp2", 1, 2, false, true },
         [QOP_LOG2] = { "log2", 1, 2, false, true },
-        [QOP_TLB_DISCARD_SETUP] = { "discard", 0, 1, true },
         [QOP_TLB_STENCIL_SETUP] = { "tlb_stencil_setup", 0, 1, true },
         [QOP_TLB_Z_WRITE] = { "tlb_z", 0, 1, true },
         [QOP_TLB_COLOR_WRITE] = { "tlb_color", 0, 1, true },
@@ -488,11 +487,15 @@ qir_SF(struct vc4_compile *c, struct qreg src)
         if (!list_empty(&c->instructions))
                 last_inst = (struct qinst *)c->instructions.prev;
 
+        /* We don't have any way to guess which kind of MOV is implied. */
+        assert(!src.pack);
+
         if (src.file != QFILE_TEMP ||
             !c->defs[src.index] ||
             last_inst != c->defs[src.index] ||
             qir_is_multi_instruction(last_inst)) {
-                src = qir_MOV(c, src);
+                struct qreg null = { QFILE_NULL, 0 };
+                last_inst = qir_MOV_dest(c, null, src);
                 last_inst = (struct qinst *)c->instructions.prev;
         }
         last_inst->sf = true;
diff --git a/src/gallium/drivers/vc4/vc4_qir.h b/src/gallium/drivers/vc4/vc4_qir.h
index 4f39d72f552..3fbf5d749e7 100644
--- a/src/gallium/drivers/vc4/vc4_qir.h
+++ b/src/gallium/drivers/vc4/vc4_qir.h
@@ -101,7 +101,6 @@ enum qop {
         QOP_LOG2,
         QOP_VW_SETUP,
         QOP_VR_SETUP,
-        QOP_TLB_DISCARD_SETUP,
         QOP_TLB_STENCIL_SETUP,
         QOP_TLB_Z_WRITE,
         QOP_TLB_COLOR_WRITE,
@@ -551,17 +550,23 @@ qir_##name##_dest(struct vc4_compile *c, struct qreg dest,               \
 }
 
 #define QIR_NODST_1(name)                                               \
-static inline void                                                      \
+static inline struct qinst *                                            \
 qir_##name(struct vc4_compile *c, struct qreg a)                        \
 {                                                                       \
-        qir_emit(c, qir_inst(QOP_##name, c->undef, a, c->undef));       \
+        struct qinst *inst = qir_inst(QOP_##name, c->undef,             \
+                                      a, c->undef);                     \
+        qir_emit(c, inst);                                              \
+        return inst;                                                    \
 }
 
 #define QIR_NODST_2(name)                                               \
-static inline void                                                      \
+static inline struct qinst *                                            \
 qir_##name(struct vc4_compile *c, struct qreg a, struct qreg b)         \
 {                                                                       \
-        qir_emit(c, qir_inst(QOP_##name, c->undef, a, b));       \
+        struct qinst *inst = qir_inst(QOP_##name, c->undef,             \
+                                      a, b);                            \
+        qir_emit(c, inst);                                              \
+        return inst;                                                    \
 }
 
 #define QIR_PACK(name)                                                   \
@@ -623,7 +628,6 @@ QIR_ALU0(TLB_COLOR_READ)
 QIR_NODST_1(TLB_COLOR_WRITE)
 QIR_NODST_1(TLB_COLOR_WRITE_MS)
 QIR_NODST_1(TLB_Z_WRITE)
-QIR_NODST_1(TLB_DISCARD_SETUP)
 QIR_NODST_1(TLB_STENCIL_SETUP)
 QIR_NODST_1(MS_MASK)
 
diff --git a/src/gallium/drivers/vc4/vc4_qir_schedule.c b/src/gallium/drivers/vc4/vc4_qir_schedule.c
index ee1e9aafbb9..186e81be750 100644
--- a/src/gallium/drivers/vc4/vc4_qir_schedule.c
+++ b/src/gallium/drivers/vc4/vc4_qir_schedule.c
@@ -236,11 +236,6 @@ calculate_deps(struct schedule_setup_state *state, struct schedule_node *n)
                 add_write_dep(dir, &state->last_tlb, n);
                 break;
 
-        case QOP_TLB_DISCARD_SETUP:
-                add_write_dep(dir, &state->last_sf, n);
-                add_write_dep(dir, &state->last_tlb, n);
-                break;
-
         default:
                 break;
         }
diff --git a/src/gallium/drivers/vc4/vc4_qpu_emit.c b/src/gallium/drivers/vc4/vc4_qpu_emit.c
index 450b97fc014..b507e370683 100644
--- a/src/gallium/drivers/vc4/vc4_qpu_emit.c
+++ b/src/gallium/drivers/vc4/vc4_qpu_emit.c
@@ -171,7 +171,6 @@ void
 vc4_generate_code(struct vc4_context *vc4, struct vc4_compile *c)
 {
         struct qpu_reg *temp_registers = vc4_register_allocate(vc4, c);
-        bool discard = false;
         uint32_t inputs_remaining = c->num_inputs;
         uint32_t vpm_read_fifo_count = 0;
         uint32_t vpm_read_offset = 0;
@@ -375,12 +374,6 @@ vc4_generate_code(struct vc4_context *vc4, struct vc4_compile *c)
                          */
                         break;
 
-                case QOP_TLB_DISCARD_SETUP:
-                        discard = true;
-                        queue(c, qpu_a_MOV(src[0], src[0]) | unpack);
-                        *last_inst(c) |= QPU_SF;
-                        break;
-
                 case QOP_TLB_STENCIL_SETUP:
                         assert(!unpack);
                         queue(c, qpu_a_MOV(qpu_ra(QPU_W_TLB_STENCIL_SETUP),
@@ -390,9 +383,8 @@ vc4_generate_code(struct vc4_context *vc4, struct vc4_compile *c)
                 case QOP_TLB_Z_WRITE:
                         queue(c, qpu_a_MOV(qpu_ra(QPU_W_TLB_Z),
                                            src[0]) | unpack);
-                        if (discard) {
-                                set_last_cond_add(c, QPU_COND_ZS);
-                        }
+                        set_last_cond_add(c, qinst->cond);
+                        handled_qinst_cond = true;
                         break;
 
                 case QOP_TLB_COLOR_READ:
@@ -406,16 +398,14 @@ vc4_generate_code(struct vc4_context *vc4, struct vc4_compile *c)
 
                 case QOP_TLB_COLOR_WRITE:
                         queue(c, qpu_a_MOV(qpu_tlbc(), src[0]) | unpack);
-                        if (discard) {
-                                set_last_cond_add(c, QPU_COND_ZS);
-                        }
+                        set_last_cond_add(c, qinst->cond);
+                        handled_qinst_cond = true;
                         break;
 
                 case QOP_TLB_COLOR_WRITE_MS:
                         queue(c, qpu_a_MOV(qpu_tlbc_ms(), src[0]));
-                        if (discard) {
-                                set_last_cond_add(c, QPU_COND_ZS);
-                        }
+                        set_last_cond_add(c, qinst->cond);
+                        handled_qinst_cond = true;
                         break;
 
                 case QOP_VARY_ADD_C:
diff --git a/src/intel/vulkan/anv_pipeline.c b/src/intel/vulkan/anv_pipeline.c
index abe93a50af8..56d79a02d79 100644
--- a/src/intel/vulkan/anv_pipeline.c
+++ b/src/intel/vulkan/anv_pipeline.c
@@ -366,10 +366,6 @@ anv_pipeline_compile(struct anv_pipeline *pipeline,
    if (pipeline->layout)
       anv_nir_apply_pipeline_layout(pipeline, nir, prog_data, map);
 
-   /* Finish the optimization and compilation process */
-   if (nir->stage == MESA_SHADER_COMPUTE)
-      brw_nir_lower_shared(nir);
-
    /* nir_lower_io will only handle the push constants; we need to set this
     * to the full number of possible uniforms.
     */
@@ -775,8 +771,6 @@ anv_pipeline_compile_cs(struct anv_pipeline *pipeline,
 
       anv_fill_binding_table(&prog_data.base, 1);
 
-      prog_data.base.total_shared = nir->num_shared;
-
       void *mem_ctx = ralloc_context(NULL);
 
       if (module->nir == NULL)
diff --git a/src/mesa/drivers/common/meta.c b/src/mesa/drivers/common/meta.c
index ab78f4565da..b673db44b0b 100644
--- a/src/mesa/drivers/common/meta.c
+++ b/src/mesa/drivers/common/meta.c
@@ -121,72 +121,51 @@ _mesa_meta_framebuffer_texture_image(struct gl_context *ctx,
                              level, layer, false, __func__);
 }
 
-GLuint
+struct gl_shader *
 _mesa_meta_compile_shader_with_debug(struct gl_context *ctx, GLenum target,
                                      const GLcharARB *source)
 {
-   GLuint shader;
-   GLint ok, size;
-   GLchar *info;
-
-   shader = _mesa_CreateShader(target);
-   _mesa_ShaderSource(shader, 1, &source, NULL);
-   _mesa_CompileShader(shader);
-
-   _mesa_GetShaderiv(shader, GL_COMPILE_STATUS, &ok);
-   if (ok)
-      return shader;
-
-   _mesa_GetShaderiv(shader, GL_INFO_LOG_LENGTH, &size);
-   if (size == 0) {
-      _mesa_DeleteShader(shader);
-      return 0;
-   }
+   const GLuint name = ~0;
+   struct gl_shader *sh;
+
+   sh = ctx->Driver.NewShader(ctx, name, target);
+   sh->Source = strdup(source);
+   sh->CompileStatus = false;
+   _mesa_compile_shader(ctx, sh);
+
+   if (!sh->CompileStatus) {
+      if (sh->InfoLog) {
+         _mesa_problem(ctx,
+                       "meta program compile failed:\n%s\nsource:\n%s\n",
+                       sh->InfoLog, source);
+      }
 
-   info = malloc(size);
-   if (!info) {
-      _mesa_DeleteShader(shader);
-      return 0;
+      _mesa_reference_shader(ctx, &sh, NULL);
    }
 
-   _mesa_GetShaderInfoLog(shader, size, NULL, info);
-   _mesa_problem(ctx,
-		 "meta program compile failed:\n%s\n"
-		 "source:\n%s\n",
-		 info, source);
-
-   free(info);
-   _mesa_DeleteShader(shader);
-
-   return 0;
+   return sh;
 }
 
-GLuint
-_mesa_meta_link_program_with_debug(struct gl_context *ctx, GLuint program)
+void
+_mesa_meta_link_program_with_debug(struct gl_context *ctx,
+                                   struct gl_shader_program *sh_prog)
 {
-   GLint ok, size;
-   GLchar *info;
-
-   _mesa_LinkProgram(program);
-
-   _mesa_GetProgramiv(program, GL_LINK_STATUS, &ok);
-   if (ok)
-      return program;
+   _mesa_link_program(ctx, sh_prog);
 
-   _mesa_GetProgramiv(program, GL_INFO_LOG_LENGTH, &size);
-   if (size == 0)
-      return 0;
-
-   info = malloc(size);
-   if (!info)
-      return 0;
-
-   _mesa_GetProgramInfoLog(program, size, NULL, info);
-   _mesa_problem(ctx, "meta program link failed:\n%s", info);
+   if (!sh_prog->LinkStatus) {
+      _mesa_problem(ctx, "meta program link failed:\n%s", sh_prog->InfoLog);
+   }
+}
 
-   free(info);
+void
+_mesa_meta_use_program(struct gl_context *ctx,
+                       struct gl_shader_program *sh_prog)
+{
+   /* Attach shader state to the binding point */
+   _mesa_reference_pipeline_object(ctx, &ctx->_Shader, &ctx->Shader);
 
-   return 0;
+   /* Update the program */
+   _mesa_use_program(ctx, sh_prog);
 }
 
 void
@@ -194,24 +173,25 @@ _mesa_meta_compile_and_link_program(struct gl_context *ctx,
                                     const char *vs_source,
                                     const char *fs_source,
                                     const char *name,
-                                    GLuint *program)
+                                    struct gl_shader_program **out_sh_prog)
 {
-   GLuint vs = _mesa_meta_compile_shader_with_debug(ctx, GL_VERTEX_SHADER,
-                                                    vs_source);
-   GLuint fs = _mesa_meta_compile_shader_with_debug(ctx, GL_FRAGMENT_SHADER,
-                                                    fs_source);
-
-   *program = _mesa_CreateProgram();
-   _mesa_ObjectLabel(GL_PROGRAM, *program, -1, name);
-   _mesa_AttachShader(*program, fs);
-   _mesa_DeleteShader(fs);
-   _mesa_AttachShader(*program, vs);
-   _mesa_DeleteShader(vs);
-   _mesa_BindAttribLocation(*program, 0, "position");
-   _mesa_BindAttribLocation(*program, 1, "texcoords");
-   _mesa_meta_link_program_with_debug(ctx, *program);
-
-   _mesa_UseProgram(*program);
+   struct gl_shader_program *sh_prog;
+   const GLuint id = ~0;
+
+   sh_prog = _mesa_new_shader_program(id);
+   sh_prog->Label = strdup(name);
+   sh_prog->NumShaders = 2;
+   sh_prog->Shaders = malloc(2 * sizeof(struct gl_shader *));
+   sh_prog->Shaders[0] =
+      _mesa_meta_compile_shader_with_debug(ctx, GL_VERTEX_SHADER, vs_source);
+   sh_prog->Shaders[1] =
+      _mesa_meta_compile_shader_with_debug(ctx, GL_FRAGMENT_SHADER, fs_source);
+
+   _mesa_meta_link_program_with_debug(ctx, sh_prog);
+
+   _mesa_meta_use_program(ctx, sh_prog);
+
+   *out_sh_prog = sh_prog;
 }
 
 /**
@@ -230,19 +210,15 @@ _mesa_meta_setup_blit_shader(struct gl_context *ctx,
 {
    char *vs_source, *fs_source;
    struct blit_shader *shader = choose_blit_shader(target, table);
-   const char *vs_input, *vs_output, *fs_input, *vs_preprocess, *fs_preprocess;
+   const char *fs_input, *vs_preprocess, *fs_preprocess;
    void *mem_ctx;
 
    if (ctx->Const.GLSLVersion < 130) {
       vs_preprocess = "";
-      vs_input = "attribute";
-      vs_output = "varying";
       fs_preprocess = "#extension GL_EXT_texture_array : enable";
       fs_input = "varying";
    } else {
       vs_preprocess = "#version 130";
-      vs_input = "in";
-      vs_output = "out";
       fs_preprocess = "#version 130";
       fs_input = "in";
       shader->func = "texture";
@@ -250,8 +226,8 @@ _mesa_meta_setup_blit_shader(struct gl_context *ctx,
 
    assert(shader != NULL);
 
-   if (shader->shader_prog != 0) {
-      _mesa_UseProgram(shader->shader_prog);
+   if (shader->shader_prog != NULL) {
+      _mesa_meta_use_program(ctx, shader->shader_prog);
       return;
    }
 
@@ -259,15 +235,16 @@ _mesa_meta_setup_blit_shader(struct gl_context *ctx,
 
    vs_source = ralloc_asprintf(mem_ctx,
                 "%s\n"
-                "%s vec2 position;\n"
-                "%s vec4 textureCoords;\n"
-                "%s vec4 texCoords;\n"
+                "#extension GL_ARB_explicit_attrib_location: enable\n"
+                "layout(location = 0) in vec2 position;\n"
+                "layout(location = 1) in vec4 textureCoords;\n"
+                "out vec4 texCoords;\n"
                 "void main()\n"
                 "{\n"
                 "   texCoords = textureCoords;\n"
                 "   gl_Position = vec4(position, 0.0, 1.0);\n"
                 "}\n",
-                vs_preprocess, vs_input, vs_input, vs_output);
+                vs_preprocess);
 
    fs_source = ralloc_asprintf(mem_ctx,
                 "%s\n"
@@ -1533,7 +1510,6 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
       "{\n"
       "   gl_FragColor = color;\n"
       "}\n";
-   GLuint vs, fs;
    bool has_integer_textures;
 
    _mesa_meta_setup_vertex_objects(ctx, &clear->VAO, &clear->buf_obj, true,
@@ -1542,21 +1518,8 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
    if (clear->ShaderProg != 0)
       return;
 
-   vs = _mesa_CreateShader(GL_VERTEX_SHADER);
-   _mesa_ShaderSource(vs, 1, &vs_source, NULL);
-   _mesa_CompileShader(vs);
-
-   fs = _mesa_CreateShader(GL_FRAGMENT_SHADER);
-   _mesa_ShaderSource(fs, 1, &fs_source, NULL);
-   _mesa_CompileShader(fs);
-
-   clear->ShaderProg = _mesa_CreateProgram();
-   _mesa_AttachShader(clear->ShaderProg, fs);
-   _mesa_DeleteShader(fs);
-   _mesa_AttachShader(clear->ShaderProg, vs);
-   _mesa_DeleteShader(vs);
-   _mesa_ObjectLabel(GL_PROGRAM, clear->ShaderProg, -1, "meta clear");
-   _mesa_LinkProgram(clear->ShaderProg);
+   _mesa_meta_compile_and_link_program(ctx, vs_source, fs_source, "meta clear",
+                                       &clear->ShaderProg);
 
    has_integer_textures = _mesa_is_gles3(ctx) ||
       (_mesa_is_desktop_gl(ctx) && ctx->Const.GLSLVersion >= 130);
@@ -1590,26 +1553,15 @@ meta_glsl_clear_init(struct gl_context *ctx, struct clear_state *clear)
                          "   out_color = color;\n"
                          "}\n");
 
-      vs = _mesa_meta_compile_shader_with_debug(ctx, GL_VERTEX_SHADER,
-                                                vs_int_source);
-      fs = _mesa_meta_compile_shader_with_debug(ctx, GL_FRAGMENT_SHADER,
-                                                fs_int_source);
+      _mesa_meta_compile_and_link_program(ctx, vs_int_source, fs_int_source,
+                                          "integer clear",
+                                          &clear->IntegerShaderProg);
       ralloc_free(shader_source_mem_ctx);
 
-      clear->IntegerShaderProg = _mesa_CreateProgram();
-      _mesa_AttachShader(clear->IntegerShaderProg, fs);
-      _mesa_DeleteShader(fs);
-      _mesa_AttachShader(clear->IntegerShaderProg, vs);
-      _mesa_DeleteShader(vs);
-
       /* Note that user-defined out attributes get automatically assigned
        * locations starting from 0, so we don't need to explicitly
        * BindFragDataLocation to 0.
        */
-
-      _mesa_ObjectLabel(GL_PROGRAM, clear->IntegerShaderProg, -1,
-                        "integer clear");
-      _mesa_meta_link_program_with_debug(ctx, clear->IntegerShaderProg);
    }
 }
 
@@ -1621,12 +1573,10 @@ meta_glsl_clear_cleanup(struct gl_context *ctx, struct clear_state *clear)
    _mesa_DeleteVertexArrays(1, &clear->VAO);
    clear->VAO = 0;
    _mesa_reference_buffer_object(ctx, &clear->buf_obj, NULL);
-   _mesa_DeleteProgram(clear->ShaderProg);
-   clear->ShaderProg = 0;
+   _mesa_reference_shader_program(ctx, &clear->ShaderProg, NULL);
 
    if (clear->IntegerShaderProg) {
-      _mesa_DeleteProgram(clear->IntegerShaderProg);
-      clear->IntegerShaderProg = 0;
+      _mesa_reference_shader_program(ctx, &clear->IntegerShaderProg, NULL);
    }
 }
 
@@ -1740,10 +1690,10 @@ meta_clear(struct gl_context *ctx, GLbitfield buffers, bool glsl)
 
    if (fb->_IntegerColor) {
       assert(glsl);
-      _mesa_UseProgram(clear->IntegerShaderProg);
+      _mesa_meta_use_program(ctx, clear->IntegerShaderProg);
       _mesa_Uniform4iv(0, 1, ctx->Color.ClearColor.i);
    } else if (glsl) {
-      _mesa_UseProgram(clear->ShaderProg);
+      _mesa_meta_use_program(ctx, clear->ShaderProg);
       _mesa_Uniform4fv(0, 1, ctx->Color.ClearColor.f);
    }
 
@@ -2704,25 +2654,17 @@ choose_blit_shader(GLenum target, struct blit_shader_table *table)
 }
 
 void
-_mesa_meta_blit_shader_table_cleanup(struct blit_shader_table *table)
+_mesa_meta_blit_shader_table_cleanup(struct gl_context *ctx,
+                                     struct blit_shader_table *table)
 {
-   _mesa_DeleteProgram(table->sampler_1d.shader_prog);
-   _mesa_DeleteProgram(table->sampler_2d.shader_prog);
-   _mesa_DeleteProgram(table->sampler_3d.shader_prog);
-   _mesa_DeleteProgram(table->sampler_rect.shader_prog);
-   _mesa_DeleteProgram(table->sampler_cubemap.shader_prog);
-   _mesa_DeleteProgram(table->sampler_1d_array.shader_prog);
-   _mesa_DeleteProgram(table->sampler_2d_array.shader_prog);
-   _mesa_DeleteProgram(table->sampler_cubemap_array.shader_prog);
-
-   table->sampler_1d.shader_prog = 0;
-   table->sampler_2d.shader_prog = 0;
-   table->sampler_3d.shader_prog = 0;
-   table->sampler_rect.shader_prog = 0;
-   table->sampler_cubemap.shader_prog = 0;
-   table->sampler_1d_array.shader_prog = 0;
-   table->sampler_2d_array.shader_prog = 0;
-   table->sampler_cubemap_array.shader_prog = 0;
+   _mesa_reference_shader_program(ctx, &table->sampler_1d.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_2d.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_3d.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_rect.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_cubemap.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_1d_array.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_2d_array.shader_prog, NULL);
+   _mesa_reference_shader_program(ctx, &table->sampler_cubemap_array.shader_prog, NULL);
 }
 
 /**
diff --git a/src/mesa/drivers/common/meta.h b/src/mesa/drivers/common/meta.h
index c2efa50a33d..0a7321c9d88 100644
--- a/src/mesa/drivers/common/meta.h
+++ b/src/mesa/drivers/common/meta.h
@@ -218,7 +218,7 @@ struct blit_shader {
    const char *type;
    const char *func;
    const char *texcoords;
-   GLuint shader_prog;
+   struct gl_shader_program *shader_prog;
 };
 
 /**
@@ -302,7 +302,7 @@ struct blit_state
    struct gl_buffer_object *buf_obj;
    struct blit_shader_table shaders_with_depth;
    struct blit_shader_table shaders_without_depth;
-   GLuint msaa_shaders[BLIT_MSAA_SHADER_COUNT];
+   struct gl_shader_program *msaa_shaders[BLIT_MSAA_SHADER_COUNT];
    struct temp_texture depthTex;
    bool no_ctsi_fallback;
 };
@@ -324,8 +324,8 @@ struct clear_state
 {
    GLuint VAO;
    struct gl_buffer_object *buf_obj;
-   GLuint ShaderProg;
-   GLuint IntegerShaderProg;
+   struct gl_shader_program *ShaderProg;
+   struct gl_shader_program *IntegerShaderProg;
 };
 
 
@@ -577,20 +577,25 @@ _mesa_meta_DrawTex(struct gl_context *ctx, GLfloat x, GLfloat y, GLfloat z,
 void
 _mesa_meta_drawbuffers_from_bitfield(GLbitfield bits);
 
-GLuint
+struct gl_shader *
 _mesa_meta_compile_shader_with_debug(struct gl_context *ctx, GLenum target,
                                      const GLcharARB *source);
 
 
-GLuint
-_mesa_meta_link_program_with_debug(struct gl_context *ctx, GLuint program);
+void
+_mesa_meta_link_program_with_debug(struct gl_context *ctx,
+                                   struct gl_shader_program *sh_prog);
 
 void
 _mesa_meta_compile_and_link_program(struct gl_context *ctx,
                                     const char *vs_source,
                                     const char *fs_source,
                                     const char *name,
-                                    GLuint *program);
+                                    struct gl_shader_program **sh_prog_ptr);
+
+extern void
+_mesa_meta_use_program(struct gl_context *ctx,
+                       struct gl_shader_program *sh_prog);
 
 GLboolean
 _mesa_meta_alloc_texture(struct temp_texture *tex,
@@ -655,7 +660,8 @@ void
 _mesa_meta_glsl_blit_cleanup(struct gl_context *ctx, struct blit_state *blit);
 
 void
-_mesa_meta_blit_shader_table_cleanup(struct blit_shader_table *table);
+_mesa_meta_blit_shader_table_cleanup(struct gl_context *ctx,
+                                     struct blit_shader_table *table);
 
 void
 _mesa_meta_glsl_generate_mipmap_cleanup(struct gl_context *ctx,
diff --git a/src/mesa/drivers/common/meta_blit.c b/src/mesa/drivers/common/meta_blit.c
index 5d80f7d8fe1..0066f7f9184 100644
--- a/src/mesa/drivers/common/meta_blit.c
+++ b/src/mesa/drivers/common/meta_blit.c
@@ -105,12 +105,12 @@ setup_glsl_msaa_blit_scaled_shader(struct gl_context *ctx,
    }
 
    if (blit->msaa_shaders[shader_index]) {
-      _mesa_UseProgram(blit->msaa_shaders[shader_index]);
+      _mesa_meta_use_program(ctx, blit->msaa_shaders[shader_index]);
       /* Update the uniform values. */
       loc_src_width =
-         _mesa_GetUniformLocation(blit->msaa_shaders[shader_index], "src_width");
+         _mesa_program_resource_location(blit->msaa_shaders[shader_index], GL_UNIFORM, "src_width");
       loc_src_height =
-         _mesa_GetUniformLocation(blit->msaa_shaders[shader_index], "src_height");
+         _mesa_program_resource_location(blit->msaa_shaders[shader_index], GL_UNIFORM, "src_height");
       _mesa_Uniform1f(loc_src_width, src_rb->Width);
       _mesa_Uniform1f(loc_src_height, src_rb->Height);
       return;
@@ -168,8 +168,9 @@ setup_glsl_msaa_blit_scaled_shader(struct gl_context *ctx,
 
    static const char vs_source[] =
                                "#version 130\n"
-                               "in vec2 position;\n"
-                               "in vec3 textureCoords;\n"
+                               "#extension GL_ARB_explicit_attrib_location: enable\n"
+                               "layout(location = 0) in vec2 position;\n"
+                               "layout(location = 1) in vec3 textureCoords;\n"
                                "out vec2 texCoords;\n"
                                "flat out int layer;\n"
                                "void main()\n"
@@ -236,9 +237,9 @@ setup_glsl_msaa_blit_scaled_shader(struct gl_context *ctx,
    _mesa_meta_compile_and_link_program(ctx, vs_source, fs_source, name,
                                        &blit->msaa_shaders[shader_index]);
    loc_src_width =
-      _mesa_GetUniformLocation(blit->msaa_shaders[shader_index], "src_width");
+      _mesa_program_resource_location(blit->msaa_shaders[shader_index], GL_UNIFORM, "src_width");
    loc_src_height =
-      _mesa_GetUniformLocation(blit->msaa_shaders[shader_index], "src_height");
+      _mesa_program_resource_location(blit->msaa_shaders[shader_index], GL_UNIFORM, "src_height");
    _mesa_Uniform1f(loc_src_width, src_rb->Width);
    _mesa_Uniform1f(loc_src_height, src_rb->Height);
 
@@ -346,7 +347,7 @@ setup_glsl_msaa_blit_shader(struct gl_context *ctx,
    }
 
    if (blit->msaa_shaders[shader_index]) {
-      _mesa_UseProgram(blit->msaa_shaders[shader_index]);
+      _mesa_meta_use_program(ctx, blit->msaa_shaders[shader_index]);
       return;
    }
 
@@ -384,8 +385,9 @@ setup_glsl_msaa_blit_shader(struct gl_context *ctx,
 
       vs_source = ralloc_asprintf(mem_ctx,
                                   "#version 130\n"
-                                  "in vec2 position;\n"
-                                  "in %s textureCoords;\n"
+                                  "#extension GL_ARB_explicit_attrib_location: enable\n"
+                                  "layout(location = 0) in vec2 position;\n"
+                                  "layout(location = 1) in %s textureCoords;\n"
                                   "out %s texCoords;\n"
                                   "void main()\n"
                                   "{\n"
@@ -506,8 +508,9 @@ setup_glsl_msaa_blit_shader(struct gl_context *ctx,
 
       vs_source = ralloc_asprintf(mem_ctx,
                                   "#version 130\n"
-                                  "in vec2 position;\n"
-                                  "in %s textureCoords;\n"
+                                  "#extension GL_ARB_explicit_attrib_location: enable\n"
+                                  "layout(location = 0) in vec2 position;\n"
+                                  "layout(location = 1) in %s textureCoords;\n"
                                   "out %s texCoords;\n"
                                   "void main()\n"
                                   "{\n"
@@ -1034,8 +1037,8 @@ _mesa_meta_glsl_blit_cleanup(struct gl_context *ctx, struct blit_state *blit)
       _mesa_reference_buffer_object(ctx, &blit->buf_obj, NULL);
    }
 
-   _mesa_meta_blit_shader_table_cleanup(&blit->shaders_with_depth);
-   _mesa_meta_blit_shader_table_cleanup(&blit->shaders_without_depth);
+   _mesa_meta_blit_shader_table_cleanup(ctx, &blit->shaders_with_depth);
+   _mesa_meta_blit_shader_table_cleanup(ctx, &blit->shaders_without_depth);
 
    _mesa_DeleteTextures(1, &blit->depthTex.TexObj);
    blit->depthTex.TexObj = 0;
diff --git a/src/mesa/drivers/common/meta_generate_mipmap.c b/src/mesa/drivers/common/meta_generate_mipmap.c
index 892d8d34619..d4b75390ebf 100644
--- a/src/mesa/drivers/common/meta_generate_mipmap.c
+++ b/src/mesa/drivers/common/meta_generate_mipmap.c
@@ -134,7 +134,7 @@ _mesa_meta_glsl_generate_mipmap_cleanup(struct gl_context *ctx,
    _mesa_reference_sampler_object(ctx, &mipmap->samp_obj, NULL);
    _mesa_reference_framebuffer(&mipmap->fb, NULL);
 
-   _mesa_meta_blit_shader_table_cleanup(&mipmap->shaders);
+   _mesa_meta_blit_shader_table_cleanup(ctx, &mipmap->shaders);
 }
 
 static GLboolean
diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index a953745b114..b45ee5eb69d 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -900,7 +900,7 @@ struct brw_context
    struct brw_cache cache;
 
    /** IDs for meta stencil blit shader programs. */
-   unsigned meta_stencil_blit_programs[2];
+   struct gl_shader_program *meta_stencil_blit_programs[2];
 
    /* Whether a meta-operation is in progress. */
    bool meta_in_progress;
diff --git a/src/mesa/drivers/dri/i965/brw_eu_emit.c b/src/mesa/drivers/dri/i965/brw_eu_emit.c
index 6f11f597492..0b99356b27d 100644
--- a/src/mesa/drivers/dri/i965/brw_eu_emit.c
+++ b/src/mesa/drivers/dri/i965/brw_eu_emit.c
@@ -202,8 +202,20 @@ brw_set_dest(struct brw_codegen *p, brw_inst *inst, struct brw_reg dest)
    /* Generators should set a default exec_size of either 8 (SIMD4x2 or SIMD8)
     * or 16 (SIMD16), as that's normally correct.  However, when dealing with
     * small registers, we automatically reduce it to match the register size.
+    *
+    * In platforms that support fp64 we can emit instructions with a width of
+    * 4 that need two SIMD8 registers and an exec_size of 8 or 16. In these
+    * cases we need to make sure that these instructions have their exec sizes
+    * set properly when they are emitted and we can't rely on this code to fix
+    * it.
     */
-   if (dest.width < BRW_EXECUTE_8)
+   bool fix_exec_size;
+   if (devinfo->gen >= 6)
+      fix_exec_size = dest.width < BRW_EXECUTE_4;
+   else
+      fix_exec_size = dest.width < BRW_EXECUTE_8;
+
+   if (fix_exec_size)
       brw_inst_set_exec_size(devinfo, inst, dest.width);
 }
 
@@ -1229,8 +1241,9 @@ brw_F16TO32(struct brw_codegen *p, struct brw_reg dst, struct brw_reg src)
 void brw_NOP(struct brw_codegen *p)
 {
    brw_inst *insn = next_insn(p, BRW_OPCODE_NOP);
-   brw_set_dest(p, insn, retype(brw_vec4_grf(0,0), BRW_REGISTER_TYPE_UD));
-   brw_set_src0(p, insn, retype(brw_vec4_grf(0,0), BRW_REGISTER_TYPE_UD));
+   brw_inst_set_exec_size(p->devinfo, insn, BRW_EXECUTE_1);
+   brw_set_dest(p, insn, retype(brw_vec1_grf(0,0), BRW_REGISTER_TYPE_UD));
+   brw_set_src0(p, insn, retype(brw_vec1_grf(0,0), BRW_REGISTER_TYPE_UD));
    brw_set_src1(p, insn, brw_imm_ud(0x0));
 }
 
@@ -2562,6 +2575,9 @@ brw_send_indirect_message(struct brw_codegen *p,
       brw_set_src1(p, send, addr);
    }
 
+   if (dst.width < BRW_EXECUTE_8)
+      brw_inst_set_exec_size(devinfo, send, dst.width);
+
    brw_set_dest(p, send, dst);
    brw_set_src0(p, send, retype(payload, BRW_REGISTER_TYPE_UD));
    brw_inst_set_sfid(devinfo, send, sfid);
@@ -3330,11 +3346,14 @@ brw_find_live_channel(struct brw_codegen *p, struct brw_reg dst)
          /* Overwrite the destination without and with execution masking to
           * find out which of the channels is active.
           */
+         brw_push_insn_state(p);
+         brw_set_default_exec_size(p, BRW_EXECUTE_4);
          brw_MOV(p, brw_writemask(vec4(dst), WRITEMASK_X),
                  brw_imm_ud(1));
 
          inst = brw_MOV(p, brw_writemask(vec4(dst), WRITEMASK_X),
                         brw_imm_ud(0));
+         brw_pop_insn_state(p);
          brw_inst_set_mask_control(devinfo, inst, BRW_MASK_ENABLE);
       }
    }
diff --git a/src/mesa/drivers/dri/i965/brw_ff_gs_emit.c b/src/mesa/drivers/dri/i965/brw_ff_gs_emit.c
index 8589dabe4c5..fea2b93471c 100644
--- a/src/mesa/drivers/dri/i965/brw_ff_gs_emit.c
+++ b/src/mesa/drivers/dri/i965/brw_ff_gs_emit.c
@@ -406,9 +406,13 @@ gen6_sol_program(struct brw_ff_gs_compile *c, struct brw_ff_gs_prog_key *key,
                                                 : 0x00020001)); /* (1, 0, 2) */
          brw_inst_set_pred_control(p->devinfo, inst, BRW_PREDICATE_NORMAL);
       }
+
+      assert(c->reg.destination_indices.width == BRW_EXECUTE_4);
+      brw_push_insn_state(p);
+      brw_set_default_exec_size(p, BRW_EXECUTE_4);
       brw_ADD(p, c->reg.destination_indices,
               c->reg.destination_indices, get_element_ud(c->reg.SVBI, 0));
-
+      brw_pop_insn_state(p);
       /* For each vertex, generate code to output each varying using the
        * appropriate binding table entry.
        */
@@ -438,8 +442,13 @@ gen6_sol_program(struct brw_ff_gs_compile *c, struct brw_ff_gs_prog_key *key,
             vertex_slot.swizzle = varying == VARYING_SLOT_PSIZ
                ? BRW_SWIZZLE_WWWW : key->transform_feedback_swizzles[binding];
             brw_set_default_access_mode(p, BRW_ALIGN_16);
+            brw_push_insn_state(p);
+            brw_set_default_exec_size(p, BRW_EXECUTE_4);
+
             brw_MOV(p, stride(c->reg.header, 4, 4, 1),
                     retype(vertex_slot, BRW_REGISTER_TYPE_UD));
+            brw_pop_insn_state(p);
+
             brw_set_default_access_mode(p, BRW_ALIGN_1);
             brw_svb_write(p,
                           final_write ? c->reg.temp : brw_null_reg(), /* dest */
diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp b/src/mesa/drivers/dri/i965/brw_fs.cpp
index b5f1a874368..874053c8489 100644
--- a/src/mesa/drivers/dri/i965/brw_fs.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs.cpp
@@ -5788,6 +5788,8 @@ brw_compile_cs(const struct brw_compiler *compiler, void *log_data,
    nir_shader *shader = nir_shader_clone(mem_ctx, src_shader);
    shader = brw_nir_apply_sampler_key(shader, compiler->devinfo, &key->tex,
                                       true);
+   brw_nir_lower_cs_shared(shader);
+   prog_data->base.total_shared += shader->num_shared;
    shader = brw_postprocess_nir(shader, compiler->devinfo, true);
 
    prog_data->local_size[0] = shader->info.cs.local_size[0];
diff --git a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
index 75c29c597f5..ae80832544b 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
@@ -1288,6 +1288,7 @@ fs_generator::generate_uniform_pull_constant_load_gen7(fs_inst *inst,
       brw_set_default_compression_control(p, BRW_COMPRESSION_NONE);
       brw_set_default_mask_control(p, BRW_MASK_DISABLE);
       brw_inst *send = brw_next_insn(p, BRW_OPCODE_SEND);
+      brw_inst_set_exec_size(devinfo, send, BRW_EXECUTE_4);
       brw_pop_insn_state(p);
 
       brw_set_dest(p, send, dst);
diff --git a/src/mesa/drivers/dri/i965/brw_meta_fast_clear.c b/src/mesa/drivers/dri/i965/brw_meta_fast_clear.c
index b81b1438ba3..1fb5dc83885 100644
--- a/src/mesa/drivers/dri/i965/brw_meta_fast_clear.c
+++ b/src/mesa/drivers/dri/i965/brw_meta_fast_clear.c
@@ -26,6 +26,7 @@
 #include "main/context.h"
 #include "main/objectlabel.h"
 #include "main/shaderapi.h"
+#include "main/shaderobj.h"
 #include "main/arrayobj.h"
 #include "main/bufferobj.h"
 #include "main/buffers.h"
@@ -61,8 +62,8 @@
 struct brw_fast_clear_state {
    struct gl_buffer_object *buf_obj;
    struct gl_vertex_array_object *array_obj;
+   struct gl_shader_program *shader_prog;
    GLuint vao;
-   GLuint shader_prog;
    GLint color_location;
 };
 
@@ -110,7 +111,8 @@ brw_bind_rep_write_shader(struct brw_context *brw, float *color)
    const char *vs_source =
       "#extension GL_AMD_vertex_shader_layer : enable\n"
       "#extension GL_ARB_draw_instanced : enable\n"
-      "attribute vec4 position;\n"
+      "#extension GL_ARB_explicit_attrib_location : enable\n"
+      "layout(location = 0) in vec4 position;\n"
       "uniform int layer;\n"
       "void main()\n"
       "{\n"
@@ -126,32 +128,23 @@ brw_bind_rep_write_shader(struct brw_context *brw, float *color)
       "   gl_FragColor = color;\n"
       "}\n";
 
-   GLuint vs, fs;
    struct brw_fast_clear_state *clear = brw->fast_clear_state;
    struct gl_context *ctx = &brw->ctx;
 
    if (clear->shader_prog) {
-      _mesa_UseProgram(clear->shader_prog);
+      _mesa_meta_use_program(ctx, clear->shader_prog);
       _mesa_Uniform4fv(clear->color_location, 1, color);
       return;
    }
 
-   vs = _mesa_meta_compile_shader_with_debug(ctx, GL_VERTEX_SHADER, vs_source);
-   fs = _mesa_meta_compile_shader_with_debug(ctx, GL_FRAGMENT_SHADER, fs_source);
-
-   clear->shader_prog = _mesa_CreateProgram();
-   _mesa_AttachShader(clear->shader_prog, fs);
-   _mesa_DeleteShader(fs);
-   _mesa_AttachShader(clear->shader_prog, vs);
-   _mesa_DeleteShader(vs);
-   _mesa_BindAttribLocation(clear->shader_prog, 0, "position");
-   _mesa_ObjectLabel(GL_PROGRAM, clear->shader_prog, -1, "meta repclear");
-   _mesa_LinkProgram(clear->shader_prog);
+   _mesa_meta_compile_and_link_program(ctx, vs_source, fs_source,
+                                       "meta repclear",
+                                       &clear->shader_prog);
 
    clear->color_location =
-      _mesa_GetUniformLocation(clear->shader_prog, "color");
+      _mesa_program_resource_location(clear->shader_prog, GL_UNIFORM, "color");
 
-   _mesa_UseProgram(clear->shader_prog);
+   _mesa_meta_use_program(ctx, clear->shader_prog);
    _mesa_Uniform4fv(clear->color_location, 1, color);
 }
 
@@ -168,7 +161,7 @@ brw_meta_fast_clear_free(struct brw_context *brw)
 
    _mesa_DeleteVertexArrays(1, &clear->vao);
    _mesa_reference_buffer_object(&brw->ctx, &clear->buf_obj, NULL);
-   _mesa_DeleteProgram(clear->shader_prog);
+   _mesa_reference_shader_program(&brw->ctx, &clear->shader_prog, NULL);
    free(clear);
 
    if (old_context)
diff --git a/src/mesa/drivers/dri/i965/brw_meta_stencil_blit.c b/src/mesa/drivers/dri/i965/brw_meta_stencil_blit.c
index 5b0c2e9bdd5..7e0424846a5 100644
--- a/src/mesa/drivers/dri/i965/brw_meta_stencil_blit.c
+++ b/src/mesa/drivers/dri/i965/brw_meta_stencil_blit.c
@@ -193,6 +193,9 @@ static const char *fs_tmpl =
    "   %s;\n"
    "}\n";
 
+#define get_uniform_loc(sh_prog, name) \
+   _mesa_program_resource_location(sh_prog, GL_UNIFORM, name)
+
 /**
  * Setup uniforms telling the coordinates of the destination rectangle in the
  * native w-tiled space. These are needed to ignore pixels that lie outside.
@@ -201,12 +204,13 @@ static const char *fs_tmpl =
  * 16x2 y-tiled).
  */
 static void
-setup_bounding_rect(GLuint prog, const struct blit_dims *dims)
+setup_bounding_rect(struct gl_shader_program *sh_prog,
+                    const struct blit_dims *dims)
 {
-   _mesa_Uniform1i(_mesa_GetUniformLocation(prog, "dst_x0"), dims->dst_x0);
-   _mesa_Uniform1i(_mesa_GetUniformLocation(prog, "dst_x1"), dims->dst_x1);
-   _mesa_Uniform1i(_mesa_GetUniformLocation(prog, "dst_y0"), dims->dst_y0);
-   _mesa_Uniform1i(_mesa_GetUniformLocation(prog, "dst_y1"), dims->dst_y1);
+   _mesa_Uniform1i(get_uniform_loc(sh_prog, "dst_x0"), dims->dst_x0);
+   _mesa_Uniform1i(get_uniform_loc(sh_prog, "dst_x1"), dims->dst_x1);
+   _mesa_Uniform1i(get_uniform_loc(sh_prog, "dst_y0"), dims->dst_y0);
+   _mesa_Uniform1i(get_uniform_loc(sh_prog, "dst_y1"), dims->dst_y1);
 }
 
 /**
@@ -215,14 +219,15 @@ setup_bounding_rect(GLuint prog, const struct blit_dims *dims)
  * between destination and source that may have differing offsets.
  */
 static void
-setup_drawing_rect(GLuint prog, const struct blit_dims *dims)
+setup_drawing_rect(struct gl_shader_program *sh_prog,
+                   const struct blit_dims *dims)
 {
-   _mesa_Uniform1f(_mesa_GetUniformLocation(prog, "draw_rect_w"),
+   _mesa_Uniform1f(get_uniform_loc(sh_prog, "draw_rect_w"),
                    dims->dst_x1 - dims->dst_x0);
-   _mesa_Uniform1f(_mesa_GetUniformLocation(prog, "draw_rect_h"),
+   _mesa_Uniform1f(get_uniform_loc(sh_prog, "draw_rect_h"),
                    dims->dst_y1 - dims->dst_y0);
-   _mesa_Uniform1f(_mesa_GetUniformLocation(prog, "dst_x_off"), dims->dst_x0);
-   _mesa_Uniform1f(_mesa_GetUniformLocation(prog, "dst_y_off"), dims->dst_y0);
+   _mesa_Uniform1f(get_uniform_loc(sh_prog, "dst_x_off"), dims->dst_x0);
+   _mesa_Uniform1f(get_uniform_loc(sh_prog, "dst_y_off"), dims->dst_y0);
 }
 
 /**
@@ -241,7 +246,7 @@ setup_drawing_rect(GLuint prog, const struct blit_dims *dims)
  *   src_x = src_x0 + (dst_x1 -dst_x - 0.5) * scale
  */
 static void
-setup_coord_coeff(GLuint prog, GLuint multiplier, GLuint offset,
+setup_coord_coeff(GLuint multiplier, GLuint offset,
                   int src_0, int src_1, int dst_0, int dst_1, bool mirror)
 {
    const float scale = ((float)(src_1 - src_0)) / (dst_1 - dst_0);
@@ -265,22 +270,21 @@ setup_coord_coeff(GLuint prog, GLuint multiplier, GLuint offset,
  * destination rectangle is adjusted for possible msaa and Y-tiling.
  */
 static void
-setup_coord_transform(GLuint prog, const struct blit_dims *dims)
+setup_coord_transform(struct gl_shader_program *sh_prog,
+                      const struct blit_dims *dims)
 {
-   setup_coord_coeff(prog,
-                     _mesa_GetUniformLocation(prog, "src_x_scale"),
-                     _mesa_GetUniformLocation(prog, "src_x_off"),
+   setup_coord_coeff(get_uniform_loc(sh_prog, "src_x_scale"),
+                     get_uniform_loc(sh_prog, "src_x_off"),
                      dims->src_x0, dims->src_x1, dims->dst_x0, dims->dst_x1,
                      dims->mirror_x);
 
-   setup_coord_coeff(prog,
-                     _mesa_GetUniformLocation(prog, "src_y_scale"),
-                     _mesa_GetUniformLocation(prog, "src_y_off"),
+   setup_coord_coeff(get_uniform_loc(sh_prog, "src_y_scale"),
+                     get_uniform_loc(sh_prog, "src_y_off"),
                      dims->src_y0, dims->src_y1, dims->dst_y0, dims->dst_y1,
                      dims->mirror_y);
 }
 
-static GLuint
+static struct gl_shader_program *
 setup_program(struct brw_context *brw, bool msaa_tex)
 {
    struct gl_context *ctx = &brw->ctx;
@@ -291,21 +295,22 @@ setup_program(struct brw_context *brw, bool msaa_tex)
    _mesa_meta_setup_vertex_objects(&brw->ctx, &blit->VAO, &blit->buf_obj, true,
                                    2, 2, 0);
 
-   GLuint *prog_id = &brw->meta_stencil_blit_programs[msaa_tex];
+   struct gl_shader_program **sh_prog_p =
+      &brw->meta_stencil_blit_programs[msaa_tex];
 
-   if (*prog_id) {
-      _mesa_UseProgram(*prog_id);
-      return *prog_id;
+   if (*sh_prog_p) {
+      _mesa_meta_use_program(ctx, *sh_prog_p);
+      return *sh_prog_p;
    }
 
    fs_source = ralloc_asprintf(NULL, fs_tmpl, sampler->sampler,
                                sampler->fetch);
    _mesa_meta_compile_and_link_program(ctx, vs_source, fs_source,
                                        "i965 stencil blit",
-                                       prog_id);
+                                       sh_prog_p);
    ralloc_free(fs_source);
 
-   return *prog_id;
+   return *sh_prog_p;
 }
 
 /**
@@ -425,7 +430,7 @@ brw_meta_stencil_blit(struct brw_context *brw,
    struct gl_context *ctx = &brw->ctx;
    struct blit_dims dims = *orig_dims;
    struct fb_tex_blit_state blit;
-   GLuint prog;
+   struct gl_shader_program *prog;
    struct gl_framebuffer *drawFb = NULL;
    struct gl_renderbuffer *rb = NULL;
    GLenum target;
@@ -467,7 +472,7 @@ brw_meta_stencil_blit(struct brw_context *brw,
    setup_drawing_rect(prog, &dims);
    setup_coord_transform(prog, orig_dims);
 
-   _mesa_Uniform1i(_mesa_GetUniformLocation(prog, "dst_num_samples"),
+   _mesa_Uniform1i(get_uniform_loc(prog, "dst_num_samples"),
                    dst_mt->num_samples);
 
    prepare_vertex_data(ctx, ctx->Meta->Blit.buf_obj);
diff --git a/src/mesa/drivers/dri/i965/brw_nir.c b/src/mesa/drivers/dri/i965/brw_nir.c
index ba9cb3f608c..941920adccd 100644
--- a/src/mesa/drivers/dri/i965/brw_nir.c
+++ b/src/mesa/drivers/dri/i965/brw_nir.c
@@ -378,7 +378,7 @@ brw_nir_lower_uniforms(nir_shader *nir, bool is_scalar)
 }
 
 void
-brw_nir_lower_shared(nir_shader *nir)
+brw_nir_lower_cs_shared(nir_shader *nir)
 {
    nir_assign_var_locations(&nir->shared, &nir->num_shared,
                             type_size_scalar_bytes);
@@ -578,9 +578,6 @@ brw_create_nir(struct brw_context *brw,
       OPT_V(nir_lower_atomics, shader_prog);
    }
 
-   if (nir->stage == MESA_SHADER_COMPUTE)
-      OPT_V(brw_nir_lower_shared);
-
    return nir;
 }
 
diff --git a/src/mesa/drivers/dri/i965/brw_nir.h b/src/mesa/drivers/dri/i965/brw_nir.h
index 0ef34735561..440b4ceb669 100644
--- a/src/mesa/drivers/dri/i965/brw_nir.h
+++ b/src/mesa/drivers/dri/i965/brw_nir.h
@@ -83,7 +83,6 @@ nir_shader *brw_create_nir(struct brw_context *brw,
 
 nir_shader *brw_preprocess_nir(nir_shader *nir, bool is_scalar);
 
-void brw_nir_lower_shared(nir_shader *nir);
 void brw_nir_lower_vs_inputs(nir_shader *nir,
                              const struct brw_device_info *devinfo,
                              bool is_scalar,
@@ -96,6 +95,7 @@ void brw_nir_lower_fs_inputs(nir_shader *nir);
 void brw_nir_lower_vue_outputs(nir_shader *nir, bool is_scalar);
 void brw_nir_lower_tcs_outputs(nir_shader *nir, const struct brw_vue_map *vue);
 void brw_nir_lower_fs_outputs(nir_shader *nir);
+void brw_nir_lower_cs_shared(nir_shader *nir);
 
 nir_shader *brw_postprocess_nir(nir_shader *nir,
                                 const struct brw_device_info *devinfo,
diff --git a/src/mesa/drivers/dri/i965/brw_vec4_generator.cpp b/src/mesa/drivers/dri/i965/brw_vec4_generator.cpp
index 695c4df258b..c9728bfb694 100644
--- a/src/mesa/drivers/dri/i965/brw_vec4_generator.cpp
+++ b/src/mesa/drivers/dri/i965/brw_vec4_generator.cpp
@@ -485,10 +485,13 @@ generate_gs_svb_write(struct brw_codegen *p,
    bool final_write = inst->sol_final_write;
 
    brw_push_insn_state(p);
+   brw_set_default_exec_size(p, BRW_EXECUTE_4);
    /* Copy Vertex data into M0.x */
    brw_MOV(p, stride(dst, 4, 4, 1),
            stride(retype(src0, BRW_REGISTER_TYPE_UD), 4, 4, 1));
+   brw_pop_insn_state(p);
 
+   brw_push_insn_state(p);
    /* Send SVB Write */
    brw_svb_write(p,
                  final_write ? src1 : brw_null_reg(), /* dest == src1 */
@@ -702,8 +705,10 @@ generate_gs_ff_sync(struct brw_codegen *p,
    brw_MOV(p, get_element_ud(header, 0), get_element_ud(dst, 0));
 
    /* src1 is not an immediate when we use transform feedback */
-   if (src1.file != BRW_IMMEDIATE_VALUE)
+   if (src1.file != BRW_IMMEDIATE_VALUE) {
+      brw_set_default_exec_size(p, BRW_EXECUTE_4);
       brw_MOV(p, brw_vec4_grf(src1.nr, 0), brw_vec4_grf(dst.nr, 1));
+   }
 
    brw_pop_insn_state(p);
 }
@@ -1473,6 +1478,7 @@ generate_code(struct brw_codegen *p,
       assert(inst->mlen <= BRW_MAX_MSG_LENGTH);
 
       unsigned pre_emit_nr_insn = p->nr_insn;
+      bool fix_exec_size = false;
 
       if (dst.width == BRW_WIDTH_4) {
          /* This happens in attribute fixups for "dual instanced" geometry
@@ -1497,6 +1503,8 @@ generate_code(struct brw_codegen *p,
             if (src[i].file == BRW_GENERAL_REGISTER_FILE)
                src[i] = stride(src[i], 4, 4, 1);
          }
+         brw_set_default_exec_size(p, BRW_EXECUTE_4);
+         fix_exec_size = true;
       }
 
       switch (inst->opcode) {
@@ -1988,6 +1996,9 @@ generate_code(struct brw_codegen *p,
          unreachable("Unsupported opcode");
       }
 
+      if (fix_exec_size)
+         brw_set_default_exec_size(p, BRW_EXECUTE_8);
+
       if (inst->opcode == VEC4_OPCODE_PACK_BYTES) {
          /* Handled dependency hints in the generator. */
 
diff --git a/src/mesa/main/enable.c b/src/mesa/main/enable.c
index 3985457f21a..b90a60ba03f 100644
--- a/src/mesa/main/enable.c
+++ b/src/mesa/main/enable.c
@@ -105,6 +105,8 @@ client_state(struct gl_context *ctx, GLenum cap, GLboolean state)
       case GL_POINT_SIZE_ARRAY_OES:
          var = &vao->VertexAttrib[VERT_ATTRIB_POINT_SIZE].Enabled;
          flag = VERT_BIT_POINT_SIZE;
+         FLUSH_VERTICES(ctx, _NEW_PROGRAM);
+         ctx->VertexProgram.PointSizeEnabled = state;
          break;
 
       /* GL_NV_primitive_restart */
diff --git a/src/mesa/main/shaderapi.c b/src/mesa/main/shaderapi.c
index cdf15b48a0d..32fad56f651 100644
--- a/src/mesa/main/shaderapi.c
+++ b/src/mesa/main/shaderapi.c
@@ -959,12 +959,9 @@ shader_source(struct gl_shader *sh, const GLchar *source)
 /**
  * Compile a shader.
  */
-static void
-compile_shader(struct gl_context *ctx, GLuint shaderObj)
+void
+_mesa_compile_shader(struct gl_context *ctx, struct gl_shader *sh)
 {
-   struct gl_shader *sh;
-
-   sh = _mesa_lookup_shader_err(ctx, shaderObj, "glCompileShader");
    if (!sh)
       return;
 
@@ -1023,12 +1020,9 @@ compile_shader(struct gl_context *ctx, GLuint shaderObj)
 /**
  * Link a program's shaders.
  */
-static void
-link_program(struct gl_context *ctx, GLuint program)
+void
+_mesa_link_program(struct gl_context *ctx, struct gl_shader_program *shProg)
 {
-   struct gl_shader_program *shProg;
-
-   shProg = _mesa_lookup_shader_program_err(ctx, program, "glLinkProgram");
    if (!shProg)
       return;
 
@@ -1270,7 +1264,8 @@ _mesa_CompileShader(GLuint shaderObj)
    GET_CURRENT_CONTEXT(ctx);
    if (MESA_VERBOSE & VERBOSE_API)
       _mesa_debug(ctx, "glCompileShader %u\n", shaderObj);
-   compile_shader(ctx, shaderObj);
+   _mesa_compile_shader(ctx, _mesa_lookup_shader_err(ctx, shaderObj,
+                                                     "glCompileShader"));
 }
 
 
@@ -1517,7 +1512,8 @@ _mesa_LinkProgram(GLuint programObj)
    GET_CURRENT_CONTEXT(ctx);
    if (MESA_VERBOSE & VERBOSE_API)
       _mesa_debug(ctx, "glLinkProgram %u\n", programObj);
-   link_program(ctx, programObj);
+   _mesa_link_program(ctx, _mesa_lookup_shader_program_err(ctx, programObj,
+                                                           "glLinkProgram"));
 }
 
 #if defined(HAVE_SHA1)
@@ -2154,25 +2150,24 @@ _mesa_CreateShaderProgramv(GLenum type, GLsizei count,
    }
 
    if (shader) {
-      _mesa_ShaderSource(shader, count, strings, NULL);
+      struct gl_shader *sh = _mesa_lookup_shader(ctx, shader);
 
-      compile_shader(ctx, shader);
+      _mesa_ShaderSource(shader, count, strings, NULL);
+      _mesa_compile_shader(ctx, sh);
 
       program = create_shader_program(ctx);
       if (program) {
 	 struct gl_shader_program *shProg;
-	 struct gl_shader *sh;
 	 GLint compiled = GL_FALSE;
 
 	 shProg = _mesa_lookup_shader_program(ctx, program);
-	 sh = _mesa_lookup_shader(ctx, shader);
 
 	 shProg->SeparateShader = GL_TRUE;
 
 	 get_shaderiv(ctx, shader, GL_COMPILE_STATUS, &compiled);
 	 if (compiled) {
 	    attach_shader(ctx, program, shader);
-	    link_program(ctx, program);
+	    _mesa_link_program(ctx, shProg);
 	    detach_shader(ctx, program, shader);
 
 #if 0
diff --git a/src/mesa/main/shaderapi.h b/src/mesa/main/shaderapi.h
index 8922c4d0640..d2d7f16ec7c 100644
--- a/src/mesa/main/shaderapi.h
+++ b/src/mesa/main/shaderapi.h
@@ -54,6 +54,12 @@ extern void
 _mesa_active_program(struct gl_context *ctx, struct gl_shader_program *shProg,
 		     const char *caller);
 
+extern void
+_mesa_compile_shader(struct gl_context *ctx, struct gl_shader *sh);
+
+extern void
+_mesa_link_program(struct gl_context *ctx, struct gl_shader_program *sh_prog);
+
 extern unsigned
 _mesa_count_active_attribs(struct gl_shader_program *shProg);