Bump version to rc4

Fixes memory leak on module unload.

CC: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
(cherry picked from commit 58272c1ad7)

.travis.yml

@@ -21,7 +21,7 @@ env:
     - LIBXCB_VERSION=libxcb-1.13
     - LIBXSHMFENCE_VERSION=libxshmfence-1.2
     - LIBVDPAU_VERSION=libvdpau-1.1
-    - LIBVA_VERSION=libva-1.6.2
+    - LIBVA_VERSION=libva-1.7.0
     - LIBWAYLAND_VERSION=wayland-1.11.1
     - WAYLAND_PROTOCOLS_VERSION=wayland-protocols-1.8
     - PKG_CONFIG_PATH=$HOME/prefix/lib/pkgconfig:$HOME/prefix/share/pkgconfig

VERSION

@@ -1 +1 @@
-18.1.0-devel
+18.1.0-rc4

bin/get-fixes-pick-list.sh

@@ -16,7 +16,7 @@ latest_branchpoint=`git merge-base origin/master HEAD`
 git log --reverse --pretty=%H $latest_branchpoint > already_landed
 
 # ... and the ones cherry-picked.
-git log --reverse --grep="cherry picked from commit" $latest_branchpoint..HEAD |\
+git log --reverse --pretty=medium --grep="cherry picked from commit" $latest_branchpoint..HEAD |\
 	grep "cherry picked from commit" |\
 	sed -e 's/^[[:space:]]*(cherry picked from commit[[:space:]]*//' -e 's/)//'  > already_picked
 
@@ -38,7 +38,7 @@ do
 
 	# Place every "fixes:" tag on its own line and join with the next word
 	# on its line or a later one.
-	fixes=`git show -s $sha | tr -d "\n" | sed -e 's/fixes:[[:space:]]*/\nfixes:/Ig' | grep "fixes:" | sed -e 's/\(fixes:[a-zA-Z0-9]*\).*$/\1/'`
+	fixes=`git show --pretty=medium -s $sha | tr -d "\n" | sed -e 's/fixes:[[:space:]]*/\nfixes:/Ig' | grep "fixes:" | sed -e 's/\(fixes:[a-zA-Z0-9]*\).*$/\1/'`
 
 	# For each one try to extract the tag
 	fixes_count=`echo "$fixes" | wc -l`

bin/get-pick-list.sh

@@ -12,7 +12,7 @@
 latest_branchpoint=`git merge-base origin/master HEAD`
 
 # Grep for commits with "cherry picked from commit" in the commit message.
-git log --reverse --grep="cherry picked from commit" $latest_branchpoint..HEAD |\
+git log --reverse --pretty=medium --grep="cherry picked from commit" $latest_branchpoint..HEAD |\
 	grep "cherry picked from commit" |\
 	sed -e 's/^[[:space:]]*(cherry picked from commit[[:space:]]*//' -e 's/)//' > already_picked
 

bin/install_megadrivers.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python
 # encoding=utf-8
-# Copyright © 2017 Intel Corporation
+# Copyright © 2017-2018 Intel Corporation
 
 # Permission is hereby granted, free of charge, to any person obtaining a copy
 # of this software and associated documentation files (the "Software"), to deal
@@ -35,7 +35,11 @@ def main():
     parser.add_argument('drivers', nargs='+')
     args = parser.parse_args()
 
-    to = os.path.join(os.environ.get('MESON_INSTALL_DESTDIR_PREFIX'), args.libdir)
+    if os.path.isabs(args.libdir):
+        to = os.path.join(os.environ.get('DESTDIR', '/'), args.libdir[1:])
+    else:
+        to = os.path.join(os.environ['MESON_INSTALL_DESTDIR_PREFIX'], args.libdir)
+
     master = os.path.join(to, os.path.basename(args.megadriver))
 
     if not os.path.exists(to):

configure.ac

@@ -86,7 +86,7 @@ DRI2PROTO_REQUIRED=2.8
 GLPROTO_REQUIRED=1.4.14
 LIBOMXIL_BELLAGIO_REQUIRED=0.0
 LIBOMXIL_TIZONIA_REQUIRED=0.10.0
-LIBVA_REQUIRED=0.38.0
+LIBVA_REQUIRED=0.39.0
 VDPAU_REQUIRED=1.1
 WAYLAND_REQUIRED=1.11
 WAYLAND_PROTOCOLS_REQUIRED=1.8

meson.build

@@ -584,7 +584,7 @@ endif
 with_gallium_va = _va == 'true'
 dep_va = null_dep
 if with_gallium_va
-  dep_va = dependency('libva', version : '>= 0.38.0')
+  dep_va = dependency('libva', version : '>= 0.39.0')
   dep_va_headers = declare_dependency(
     compile_args : run_command(prog_pkgconfig, ['libva', '--cflags']).stdout().split()
   )

src/amd/common/ac_llvm_build.c

@@ -1521,6 +1521,20 @@ LLVMValueRef ac_build_image_opcode(struct ac_llvm_context *ctx,
 	LLVMValueRef addr;
 	unsigned num_addr = 0;
 
+	if (a->opcode == ac_image_get_lod) {
+		switch (a->dim) {
+		case ac_image_1darray:
+			num_coords = 1;
+			break;
+		case ac_image_2darray:
+		case ac_image_cube:
+			num_coords = 2;
+			break;
+		default:
+			break;
+		}
+	}
+
 	if (a->offset)
 		args[num_addr++] = ac_to_integer(ctx, a->offset);
 	if (a->bias)

src/amd/common/ac_nir_to_llvm.c

@@ -1307,6 +1307,14 @@ static LLVMValueRef build_tex_intrinsic(struct ac_nir_context *ctx,
 		}
 	}
 
+	/* Fixup for GFX9 which allocates 1D textures as 2D. */
+	if (instr->op == nir_texop_lod && ctx->ac.chip_class >= GFX9) {
+		if ((args->dim == ac_image_2darray ||
+		     args->dim == ac_image_2d) && !args->coords[1]) {
+			args->coords[1] = ctx->ac.i32_0;
+		}
+	}
+
 	args->attributes = AC_FUNC_ATTR_READNONE;
 	return ac_build_image_opcode(&ctx->ac, args);
 }
@@ -3397,6 +3405,13 @@ static void visit_tex(struct ac_nir_context *ctx, nir_tex_instr *instr)
 	}
 
 	/* Texture coordinates fixups */
+	if (instr->coord_components > 1 &&
+	    instr->sampler_dim == GLSL_SAMPLER_DIM_1D &&
+	    instr->is_array &&
+	    instr->op != nir_texop_txf) {
+		args.coords[1] = apply_round_slice(&ctx->ac, args.coords[1]);
+	}
+
 	if (instr->coord_components > 2 &&
 	    (instr->sampler_dim == GLSL_SAMPLER_DIM_2D ||
 	     instr->sampler_dim == GLSL_SAMPLER_DIM_MS ||

src/amd/vulkan/radv_device.c

@@ -463,15 +463,6 @@ VkResult radv_CreateInstance(
 		client_version = VK_MAKE_VERSION(1, 0, 0);
 	}
 
-	if (VK_MAKE_VERSION(1, 0, 0) > client_version ||
-	    client_version > VK_MAKE_VERSION(1, 1, 0xfff)) {
-		return vk_errorf(VK_ERROR_INCOMPATIBLE_DRIVER,
-				 "Client requested version %d.%d.%d",
-				 VK_VERSION_MAJOR(client_version),
-				 VK_VERSION_MINOR(client_version),
-				 VK_VERSION_PATCH(client_version));
-	}
-
 	instance = vk_zalloc2(&default_alloc, pAllocator, sizeof(*instance), 8,
 			      VK_SYSTEM_ALLOCATION_SCOPE_INSTANCE);
 	if (!instance)

src/amd/vulkan/radv_entrypoints_gen.py

@@ -205,7 +205,7 @@ radv_entrypoint_is_enabled(int index, uint32_t core_version,
    % if not e.device_command:
       if (device) return false;
    % endif
-   % if e.name == 'vkCreateInstance' or e.name == 'vkEnumerateInstanceExtensionProperties' or e.name == 'vkEnumerateInstanceLayerProperties':
+   % if e.name == 'vkCreateInstance' or e.name == 'vkEnumerateInstanceExtensionProperties' or e.name == 'vkEnumerateInstanceLayerProperties' or e.name == 'vkEnumerateInstanceVersion':
       return !device;
    % elif e.core_version:
       return instance && ${e.core_version.c_vk_version()} <= core_version;

src/amd/vulkan/radv_image.c

@@ -133,12 +133,12 @@ radv_use_dcc_for_image(struct radv_device *device,
 	if (create_info->scanout)
 		return false;
 
-	/* FIXME: DCC for MSAA with 4x and 8x samples doesn't work yet. */
-	if (pCreateInfo->samples > 2)
-		return false;
-
-	/* TODO: Enable DCC for MSAA textures. */
-	if (!device->physical_device->dcc_msaa_allowed)
+	/* FIXME: DCC for MSAA with 4x and 8x samples doesn't work yet, while
+	 * 2x can be enabled with an option.
+	 */
+	if (pCreateInfo->samples > 2 ||
+	    (pCreateInfo->samples == 2 &&
+	     !device->physical_device->dcc_msaa_allowed))
 		return false;
 
 	/* Determine if the formats are DCC compatible. */
@@ -968,7 +968,7 @@ radv_image_create(VkDevice _device,
 	image->info.samples = pCreateInfo->samples;
 	image->info.array_size = pCreateInfo->arrayLayers;
 	image->info.levels = pCreateInfo->mipLevels;
-	image->info.num_channels = 4; /* TODO: set this correctly */
+	image->info.num_channels = vk_format_get_nr_components(pCreateInfo->format);
 
 	image->vk_format = pCreateInfo->format;
 	image->tiling = pCreateInfo->tiling;

src/amd/vulkan/radv_meta_blit2d.c

@@ -100,7 +100,8 @@ blit2d_bind_src(struct radv_cmd_buffer *cmd_buffer,
                 struct radv_meta_blit2d_buffer *src_buf,
                 struct blit2d_src_temps *tmp,
                 enum blit2d_src_type src_type, VkFormat depth_format,
-                VkImageAspectFlagBits aspects)
+                VkImageAspectFlagBits aspects,
+                uint32_t log2_samples)
 {
 	struct radv_device *device = cmd_buffer->device;
 
@@ -108,7 +109,7 @@ blit2d_bind_src(struct radv_cmd_buffer *cmd_buffer,
 		create_bview(cmd_buffer, src_buf, &tmp->bview, depth_format);
 
 		radv_meta_push_descriptor_set(cmd_buffer, VK_PIPELINE_BIND_POINT_GRAPHICS,
-					      device->meta_state.blit2d.p_layouts[src_type],
+					      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
 					      0, /* set */
 					      1, /* descriptorWriteCount */
 					      (VkWriteDescriptorSet[]) {
@@ -123,7 +124,7 @@ blit2d_bind_src(struct radv_cmd_buffer *cmd_buffer,
 					      });
 
 		radv_CmdPushConstants(radv_cmd_buffer_to_handle(cmd_buffer),
-				      device->meta_state.blit2d.p_layouts[src_type],
+				      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
 				      VK_SHADER_STAGE_FRAGMENT_BIT, 16, 4,
 				      &src_buf->pitch);
 	} else {
@@ -131,12 +132,12 @@ blit2d_bind_src(struct radv_cmd_buffer *cmd_buffer,
 
 		if (src_type == BLIT2D_SRC_TYPE_IMAGE_3D)
 			radv_CmdPushConstants(radv_cmd_buffer_to_handle(cmd_buffer),
-					      device->meta_state.blit2d.p_layouts[src_type],
+					      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
 					      VK_SHADER_STAGE_FRAGMENT_BIT, 16, 4,
 					      &src_img->layer);
 
 		radv_meta_push_descriptor_set(cmd_buffer, VK_PIPELINE_BIND_POINT_GRAPHICS,
-					      device->meta_state.blit2d.p_layouts[src_type],
+					      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
 					      0, /* set */
 					      1, /* descriptorWriteCount */
 					      (VkWriteDescriptorSet[]) {
@@ -190,10 +191,11 @@ blit2d_bind_dst(struct radv_cmd_buffer *cmd_buffer,
 
 static void
 bind_pipeline(struct radv_cmd_buffer *cmd_buffer,
-              enum blit2d_src_type src_type, unsigned fs_key)
+              enum blit2d_src_type src_type, unsigned fs_key,
+              uint32_t log2_samples)
 {
 	VkPipeline pipeline =
-		cmd_buffer->device->meta_state.blit2d.pipelines[src_type][fs_key];
+		cmd_buffer->device->meta_state.blit2d[log2_samples].pipelines[src_type][fs_key];
 
 	radv_CmdBindPipeline(radv_cmd_buffer_to_handle(cmd_buffer),
 			     VK_PIPELINE_BIND_POINT_GRAPHICS, pipeline);
@@ -201,10 +203,11 @@ bind_pipeline(struct radv_cmd_buffer *cmd_buffer,
 
 static void
 bind_depth_pipeline(struct radv_cmd_buffer *cmd_buffer,
-		    enum blit2d_src_type src_type)
+		    enum blit2d_src_type src_type,
+		    uint32_t log2_samples)
 {
 	VkPipeline pipeline =
-		cmd_buffer->device->meta_state.blit2d.depth_only_pipeline[src_type];
+		cmd_buffer->device->meta_state.blit2d[log2_samples].depth_only_pipeline[src_type];
 
 	radv_CmdBindPipeline(radv_cmd_buffer_to_handle(cmd_buffer),
 			     VK_PIPELINE_BIND_POINT_GRAPHICS, pipeline);
@@ -212,10 +215,11 @@ bind_depth_pipeline(struct radv_cmd_buffer *cmd_buffer,
 
 static void
 bind_stencil_pipeline(struct radv_cmd_buffer *cmd_buffer,
-		      enum blit2d_src_type src_type)
+		      enum blit2d_src_type src_type,
+		      uint32_t log2_samples)
 {
 	VkPipeline pipeline =
-		cmd_buffer->device->meta_state.blit2d.stencil_only_pipeline[src_type];
+		cmd_buffer->device->meta_state.blit2d[log2_samples].stencil_only_pipeline[src_type];
 
 	radv_CmdBindPipeline(radv_cmd_buffer_to_handle(cmd_buffer),
 			     VK_PIPELINE_BIND_POINT_GRAPHICS, pipeline);
@@ -227,7 +231,8 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 			    struct radv_meta_blit2d_buffer *src_buf,
 			    struct radv_meta_blit2d_surf *dst,
 			    unsigned num_rects,
-			    struct radv_meta_blit2d_rect *rects, enum blit2d_src_type src_type)
+			    struct radv_meta_blit2d_rect *rects, enum blit2d_src_type src_type,
+			    uint32_t log2_samples)
 {
 	struct radv_device *device = cmd_buffer->device;
 
@@ -241,7 +246,7 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 			else if (aspect_mask == VK_IMAGE_ASPECT_DEPTH_BIT)
 				depth_format = vk_format_depth_only(dst->image->vk_format);
 			struct blit2d_src_temps src_temps;
-			blit2d_bind_src(cmd_buffer, src_img, src_buf, &src_temps, src_type, depth_format, aspect_mask);
+			blit2d_bind_src(cmd_buffer, src_img, src_buf, &src_temps, src_type, depth_format, aspect_mask, log2_samples);
 
 			struct blit2d_dst_temps dst_temps;
 			blit2d_bind_dst(cmd_buffer, dst, rects[r].dst_x + rects[r].width,
@@ -255,7 +260,7 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 			};
 
 			radv_CmdPushConstants(radv_cmd_buffer_to_handle(cmd_buffer),
-					device->meta_state.blit2d.p_layouts[src_type],
+					device->meta_state.blit2d[log2_samples].p_layouts[src_type],
 					VK_SHADER_STAGE_VERTEX_BIT, 0, 16,
 					vertex_push_constants);
 
@@ -266,7 +271,7 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 				radv_CmdBeginRenderPass(radv_cmd_buffer_to_handle(cmd_buffer),
 							&(VkRenderPassBeginInfo) {
 								.sType = VK_STRUCTURE_TYPE_RENDER_PASS_BEGIN_INFO,
-									.renderPass = device->meta_state.blit2d.render_passes[fs_key][dst_layout],
+									.renderPass = device->meta_state.blit2d_render_passes[fs_key][dst_layout],
 									.framebuffer = dst_temps.fb,
 									.renderArea = {
 									.offset = { rects[r].dst_x, rects[r].dst_y, },
@@ -277,13 +282,13 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 										}, VK_SUBPASS_CONTENTS_INLINE);
 
 
-				bind_pipeline(cmd_buffer, src_type, fs_key);
+				bind_pipeline(cmd_buffer, src_type, fs_key, log2_samples);
 			} else if (aspect_mask == VK_IMAGE_ASPECT_DEPTH_BIT) {
 				enum radv_blit_ds_layout ds_layout = radv_meta_blit_ds_to_type(dst->current_layout);
 				radv_CmdBeginRenderPass(radv_cmd_buffer_to_handle(cmd_buffer),
 							&(VkRenderPassBeginInfo) {
 								.sType = VK_STRUCTURE_TYPE_RENDER_PASS_BEGIN_INFO,
-									.renderPass = device->meta_state.blit2d.depth_only_rp[ds_layout],
+									.renderPass = device->meta_state.blit2d_depth_only_rp[ds_layout],
 									.framebuffer = dst_temps.fb,
 									.renderArea = {
 									.offset = { rects[r].dst_x, rects[r].dst_y, },
@@ -294,14 +299,14 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 										}, VK_SUBPASS_CONTENTS_INLINE);
 
 
-				bind_depth_pipeline(cmd_buffer, src_type);
+				bind_depth_pipeline(cmd_buffer, src_type, log2_samples);
 
 			} else if (aspect_mask == VK_IMAGE_ASPECT_STENCIL_BIT) {
 				enum radv_blit_ds_layout ds_layout = radv_meta_blit_ds_to_type(dst->current_layout);
 				radv_CmdBeginRenderPass(radv_cmd_buffer_to_handle(cmd_buffer),
 							&(VkRenderPassBeginInfo) {
 								.sType = VK_STRUCTURE_TYPE_RENDER_PASS_BEGIN_INFO,
-									.renderPass = device->meta_state.blit2d.stencil_only_rp[ds_layout],
+									.renderPass = device->meta_state.blit2d_stencil_only_rp[ds_layout],
 									.framebuffer = dst_temps.fb,
 									.renderArea = {
 									.offset = { rects[r].dst_x, rects[r].dst_y, },
@@ -312,7 +317,7 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 										}, VK_SUBPASS_CONTENTS_INLINE);
 
 
-				bind_stencil_pipeline(cmd_buffer, src_type);
+				bind_stencil_pipeline(cmd_buffer, src_type, log2_samples);
 			} else
 				unreachable("Processing blit2d with multiple aspects.");
 
@@ -332,7 +337,24 @@ radv_meta_blit2d_normal_dst(struct radv_cmd_buffer *cmd_buffer,
 
 
 
-			radv_CmdDraw(radv_cmd_buffer_to_handle(cmd_buffer), 3, 1, 0, 0);
+			if (log2_samples > 0) {
+				for (uint32_t sample = 0; sample < src_img->image->info.samples; sample++) {
+					uint32_t sample_mask = 1 << sample;
+					radv_CmdPushConstants(radv_cmd_buffer_to_handle(cmd_buffer),
+							      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
+							      VK_SHADER_STAGE_FRAGMENT_BIT, 20, 4,
+							      &sample);
+
+					radv_CmdPushConstants(radv_cmd_buffer_to_handle(cmd_buffer),
+							      device->meta_state.blit2d[log2_samples].p_layouts[src_type],
+							      VK_SHADER_STAGE_FRAGMENT_BIT, 24, 4,
+							      &sample_mask);
+
+					radv_CmdDraw(radv_cmd_buffer_to_handle(cmd_buffer), 3, 1, 0, 0);
+				}
+			}
+			else
+				radv_CmdDraw(radv_cmd_buffer_to_handle(cmd_buffer), 3, 1, 0, 0);
 			radv_CmdEndRenderPass(radv_cmd_buffer_to_handle(cmd_buffer));
 
 			/* At the point where we emit the draw call, all data from the
@@ -358,7 +380,8 @@ radv_meta_blit2d(struct radv_cmd_buffer *cmd_buffer,
 	enum blit2d_src_type src_type = src_buf ? BLIT2D_SRC_TYPE_BUFFER :
 		use_3d ? BLIT2D_SRC_TYPE_IMAGE_3D : BLIT2D_SRC_TYPE_IMAGE;
 	radv_meta_blit2d_normal_dst(cmd_buffer, src_img, src_buf, dst,
-				    num_rects, rects, src_type);
+				    num_rects, rects, src_type,
+				    src_img ? util_logbase2(src_img->image->info.samples) : 0);
 }
 
 static nir_shader *
@@ -421,13 +444,14 @@ build_nir_vertex_shader(void)
 
 typedef nir_ssa_def* (*texel_fetch_build_func)(struct nir_builder *,
                                                struct radv_device *,
-                                               nir_ssa_def *, bool);
+                                               nir_ssa_def *, bool, bool);
 
 static nir_ssa_def *
 build_nir_texel_fetch(struct nir_builder *b, struct radv_device *device,
-                      nir_ssa_def *tex_pos, bool is_3d)
+                      nir_ssa_def *tex_pos, bool is_3d, bool is_multisampled)
 {
-	enum glsl_sampler_dim dim = is_3d ? GLSL_SAMPLER_DIM_3D : GLSL_SAMPLER_DIM_2D;
+	enum glsl_sampler_dim dim =
+		is_3d ? GLSL_SAMPLER_DIM_3D : is_multisampled ? GLSL_SAMPLER_DIM_MS : GLSL_SAMPLER_DIM_2D;
 	const struct glsl_type *sampler_type =
 		glsl_sampler_type(dim, false, false, GLSL_TYPE_UINT);
 	nir_variable *sampler = nir_variable_create(b->shader, nir_var_uniform,
@@ -436,6 +460,7 @@ build_nir_texel_fetch(struct nir_builder *b, struct radv_device *device,
 	sampler->data.binding = 0;
 
 	nir_ssa_def *tex_pos_3d = NULL;
+	nir_intrinsic_instr *sample_idx = NULL;
 	if (is_3d) {
 		nir_intrinsic_instr *layer = nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_push_constant);
 		nir_intrinsic_set_base(layer, 16);
@@ -451,13 +476,26 @@ build_nir_texel_fetch(struct nir_builder *b, struct radv_device *device,
 		chans[2] = &layer->dest.ssa;
 		tex_pos_3d = nir_vec(b, chans, 3);
 	}
-	nir_tex_instr *tex = nir_tex_instr_create(b->shader, 2);
+	if (is_multisampled) {
+		sample_idx = nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_push_constant);
+		nir_intrinsic_set_base(sample_idx, 20);
+		nir_intrinsic_set_range(sample_idx, 4);
+		sample_idx->src[0] = nir_src_for_ssa(nir_imm_int(b, 0));
+		sample_idx->num_components = 1;
+		nir_ssa_dest_init(&sample_idx->instr, &sample_idx->dest, 1, 32, "sample_idx");
+		nir_builder_instr_insert(b, &sample_idx->instr);
+	}
+	nir_tex_instr *tex = nir_tex_instr_create(b->shader, is_multisampled ? 3 : 2);
 	tex->sampler_dim = dim;
-	tex->op = nir_texop_txf;
+	tex->op = is_multisampled ? nir_texop_txf_ms : nir_texop_txf;
 	tex->src[0].src_type = nir_tex_src_coord;
 	tex->src[0].src = nir_src_for_ssa(is_3d ? tex_pos_3d : tex_pos);
-	tex->src[1].src_type = nir_tex_src_lod;
-	tex->src[1].src = nir_src_for_ssa(nir_imm_int(b, 0));
+	tex->src[1].src_type = is_multisampled ? nir_tex_src_ms_index : nir_tex_src_lod;
+	tex->src[1].src = nir_src_for_ssa(is_multisampled ? &sample_idx->dest.ssa : nir_imm_int(b, 0));
+	if (is_multisampled) {
+		tex->src[2].src_type = nir_tex_src_lod;
+		tex->src[2].src = nir_src_for_ssa(nir_imm_int(b, 0));
+	}
 	tex->dest_type = nir_type_uint;
 	tex->is_array = false;
 	tex->coord_components = is_3d ? 3 : 2;
@@ -473,7 +511,7 @@ build_nir_texel_fetch(struct nir_builder *b, struct radv_device *device,
 
 static nir_ssa_def *
 build_nir_buffer_fetch(struct nir_builder *b, struct radv_device *device,
-		       nir_ssa_def *tex_pos, bool is_3d)
+		       nir_ssa_def *tex_pos, bool is_3d, bool is_multisampled)
 {
 	const struct glsl_type *sampler_type =
 		glsl_sampler_type(GLSL_SAMPLER_DIM_BUF, false, false, GLSL_TYPE_UINT);
@@ -519,9 +557,31 @@ static const VkPipelineVertexInputStateCreateInfo normal_vi_create_info = {
 	.vertexAttributeDescriptionCount = 0,
 };
 
+static void
+build_nir_store_sample_mask(struct nir_builder *b)
+{
+	nir_intrinsic_instr *sample_mask = nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_push_constant);
+	nir_intrinsic_set_base(sample_mask, 24);
+	nir_intrinsic_set_range(sample_mask, 4);
+	sample_mask->src[0] = nir_src_for_ssa(nir_imm_int(b, 0));
+	sample_mask->num_components = 1;
+	nir_ssa_dest_init(&sample_mask->instr, &sample_mask->dest, 1, 32, "sample_mask");
+	nir_builder_instr_insert(b, &sample_mask->instr);
+
+	const struct glsl_type *sample_mask_out_type = glsl_uint_type();
+
+	nir_variable *sample_mask_out =
+		nir_variable_create(b->shader, nir_var_shader_out,
+				    sample_mask_out_type, "sample_mask_out");
+	sample_mask_out->data.location = FRAG_RESULT_SAMPLE_MASK;
+
+	nir_store_var(b, sample_mask_out, &sample_mask->dest.ssa, 0x1);
+}
+
 static nir_shader *
 build_nir_copy_fragment_shader(struct radv_device *device,
-                               texel_fetch_build_func txf_func, const char* name, bool is_3d)
+                               texel_fetch_build_func txf_func, const char* name, bool is_3d,
+                               bool is_multisampled)
 {
 	const struct glsl_type *vec4 = glsl_vec4_type();
 	const struct glsl_type *vec2 = glsl_vector_type(GLSL_TYPE_FLOAT, 2);
@@ -538,11 +598,15 @@ build_nir_copy_fragment_shader(struct radv_device *device,
 						      vec4, "f_color");
 	color_out->data.location = FRAG_RESULT_DATA0;
 
+	if (is_multisampled) {
+		build_nir_store_sample_mask(&b);
+	}
+
 	nir_ssa_def *pos_int = nir_f2i32(&b, nir_load_var(&b, tex_pos_in));
 	unsigned swiz[4] = { 0, 1 };
 	nir_ssa_def *tex_pos = nir_swizzle(&b, pos_int, swiz, 2, false);
 
-	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d);
+	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d, is_multisampled);
 	nir_store_var(&b, color_out, color, 0xf);
 
 	return b.shader;
@@ -550,7 +614,8 @@ build_nir_copy_fragment_shader(struct radv_device *device,
 
 static nir_shader *
 build_nir_copy_fragment_shader_depth(struct radv_device *device,
-				     texel_fetch_build_func txf_func, const char* name, bool is_3d)
+				     texel_fetch_build_func txf_func, const char* name, bool is_3d,
+				     bool is_multisampled)
 {
 	const struct glsl_type *vec4 = glsl_vec4_type();
 	const struct glsl_type *vec2 = glsl_vector_type(GLSL_TYPE_FLOAT, 2);
@@ -567,11 +632,15 @@ build_nir_copy_fragment_shader_depth(struct radv_device *device,
 						      vec4, "f_color");
 	color_out->data.location = FRAG_RESULT_DEPTH;
 
+	if (is_multisampled) {
+		build_nir_store_sample_mask(&b);
+	}
+
 	nir_ssa_def *pos_int = nir_f2i32(&b, nir_load_var(&b, tex_pos_in));
 	unsigned swiz[4] = { 0, 1 };
 	nir_ssa_def *tex_pos = nir_swizzle(&b, pos_int, swiz, 2, false);
 
-	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d);
+	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d, is_multisampled);
 	nir_store_var(&b, color_out, color, 0x1);
 
 	return b.shader;
@@ -579,7 +648,8 @@ build_nir_copy_fragment_shader_depth(struct radv_device *device,
 
 static nir_shader *
 build_nir_copy_fragment_shader_stencil(struct radv_device *device,
-				       texel_fetch_build_func txf_func, const char* name, bool is_3d)
+				       texel_fetch_build_func txf_func, const char* name, bool is_3d,
+				       bool is_multisampled)
 {
 	const struct glsl_type *vec4 = glsl_vec4_type();
 	const struct glsl_type *vec2 = glsl_vector_type(GLSL_TYPE_FLOAT, 2);
@@ -596,11 +666,15 @@ build_nir_copy_fragment_shader_stencil(struct radv_device *device,
 						      vec4, "f_color");
 	color_out->data.location = FRAG_RESULT_STENCIL;
 
+	if (is_multisampled) {
+		build_nir_store_sample_mask(&b);
+	}
+
 	nir_ssa_def *pos_int = nir_f2i32(&b, nir_load_var(&b, tex_pos_in));
 	unsigned swiz[4] = { 0, 1 };
 	nir_ssa_def *tex_pos = nir_swizzle(&b, pos_int, swiz, 2, false);
 
-	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d);
+	nir_ssa_def *color = txf_func(&b, device, tex_pos, is_3d, is_multisampled);
 	nir_store_var(&b, color_out, color, 0x1);
 
 	return b.shader;
@@ -614,45 +688,48 @@ radv_device_finish_meta_blit2d_state(struct radv_device *device)
 	for(unsigned j = 0; j < NUM_META_FS_KEYS; ++j) {
 		for (unsigned k = 0; k < RADV_META_DST_LAYOUT_COUNT; ++k) {
 			radv_DestroyRenderPass(radv_device_to_handle(device),
-			                       state->blit2d.render_passes[j][k],
-			                       &state->alloc);
+					       state->blit2d_render_passes[j][k],
+					       &state->alloc);
 		}
 	}
 
 	for (enum radv_blit_ds_layout j = RADV_BLIT_DS_LAYOUT_TILE_ENABLE; j < RADV_BLIT_DS_LAYOUT_COUNT; j++) {
 		radv_DestroyRenderPass(radv_device_to_handle(device),
-				       state->blit2d.depth_only_rp[j], &state->alloc);
+				       state->blit2d_depth_only_rp[j], &state->alloc);
 		radv_DestroyRenderPass(radv_device_to_handle(device),
-				       state->blit2d.stencil_only_rp[j], &state->alloc);
+				       state->blit2d_stencil_only_rp[j], &state->alloc);
 	}
 
-	for (unsigned src = 0; src < BLIT2D_NUM_SRC_TYPES; src++) {
-		radv_DestroyPipelineLayout(radv_device_to_handle(device),
-					   state->blit2d.p_layouts[src],
-					   &state->alloc);
-		radv_DestroyDescriptorSetLayout(radv_device_to_handle(device),
-						state->blit2d.ds_layouts[src],
-						&state->alloc);
+	for (unsigned log2_samples = 0; log2_samples < 1 + MAX_SAMPLES_LOG2; ++log2_samples) {
+		for (unsigned src = 0; src < BLIT2D_NUM_SRC_TYPES; src++) {
+			radv_DestroyPipelineLayout(radv_device_to_handle(device),
+						   state->blit2d[log2_samples].p_layouts[src],
+						   &state->alloc);
+			radv_DestroyDescriptorSetLayout(radv_device_to_handle(device),
+							state->blit2d[log2_samples].ds_layouts[src],
+							&state->alloc);
+
+			for (unsigned j = 0; j < NUM_META_FS_KEYS; ++j) {
+				radv_DestroyPipeline(radv_device_to_handle(device),
+						     state->blit2d[log2_samples].pipelines[src][j],
+						     &state->alloc);
+			}
 
-		for (unsigned j = 0; j < NUM_META_FS_KEYS; ++j) {
 			radv_DestroyPipeline(radv_device_to_handle(device),
-					     state->blit2d.pipelines[src][j],
+					     state->blit2d[log2_samples].depth_only_pipeline[src],
+					     &state->alloc);
+			radv_DestroyPipeline(radv_device_to_handle(device),
+					     state->blit2d[log2_samples].stencil_only_pipeline[src],
 					     &state->alloc);
 		}
-
-		radv_DestroyPipeline(radv_device_to_handle(device),
-				     state->blit2d.depth_only_pipeline[src],
-				     &state->alloc);
-		radv_DestroyPipeline(radv_device_to_handle(device),
-				     state->blit2d.stencil_only_pipeline[src],
-				     &state->alloc);
 	}
 }
 
 static VkResult
 blit2d_init_color_pipeline(struct radv_device *device,
 			   enum blit2d_src_type src_type,
-			   VkFormat format)
+			   VkFormat format,
+			   uint32_t log2_samples)
 {
 	VkResult result;
 	unsigned fs_key = radv_format_meta_fs_key(format);
@@ -681,7 +758,7 @@ blit2d_init_color_pipeline(struct radv_device *device,
 	struct radv_shader_module fs = { .nir = NULL };
 
 
-	fs.nir = build_nir_copy_fragment_shader(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D);
+	fs.nir = build_nir_copy_fragment_shader(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D, log2_samples > 0);
 	vi_create_info = &normal_vi_create_info;
 
 	struct radv_shader_module vs = {
@@ -705,7 +782,7 @@ blit2d_init_color_pipeline(struct radv_device *device,
 	};
 
 	for (unsigned dst_layout = 0; dst_layout < RADV_META_DST_LAYOUT_COUNT; ++dst_layout) {
-		if (!device->meta_state.blit2d.render_passes[fs_key][dst_layout]) {
+		if (!device->meta_state.blit2d_render_passes[fs_key][dst_layout]) {
 			VkImageLayout layout = radv_meta_dst_layout_to_layout(dst_layout);
 
 			result = radv_CreateRenderPass(radv_device_to_handle(device),
@@ -737,7 +814,7 @@ blit2d_init_color_pipeline(struct radv_device *device,
 						.pPreserveAttachments = (uint32_t[]) { 0 },
 						},
 						.dependencyCount = 0,
-					}, &device->meta_state.alloc, &device->meta_state.blit2d.render_passes[fs_key][dst_layout]);
+					}, &device->meta_state.alloc, &device->meta_state.blit2d_render_passes[fs_key][dst_layout]);
 		}
 	}
 
@@ -765,7 +842,7 @@ blit2d_init_color_pipeline(struct radv_device *device,
 		},
 		.pMultisampleState = &(VkPipelineMultisampleStateCreateInfo) {
 			.sType = VK_STRUCTURE_TYPE_PIPELINE_MULTISAMPLE_STATE_CREATE_INFO,
-			.rasterizationSamples = 1,
+			.rasterizationSamples = 1 << log2_samples,
 			.sampleShadingEnable = false,
 			.pSampleMask = (VkSampleMask[]) { UINT32_MAX },
 		},
@@ -796,8 +873,8 @@ blit2d_init_color_pipeline(struct radv_device *device,
 			},
 		},
 		.flags = 0,
-		.layout = device->meta_state.blit2d.p_layouts[src_type],
-		.renderPass = device->meta_state.blit2d.render_passes[fs_key][0],
+		.layout = device->meta_state.blit2d[log2_samples].p_layouts[src_type],
+		.renderPass = device->meta_state.blit2d_render_passes[fs_key][0],
 		.subpass = 0,
 	};
 
@@ -809,7 +886,7 @@ blit2d_init_color_pipeline(struct radv_device *device,
 					       radv_pipeline_cache_to_handle(&device->meta_state.cache),
 					       &vk_pipeline_info, &radv_pipeline_info,
 					       &device->meta_state.alloc,
-					       &device->meta_state.blit2d.pipelines[src_type][fs_key]);
+					       &device->meta_state.blit2d[log2_samples].pipelines[src_type][fs_key]);
 
 
 	ralloc_free(vs.nir);
@@ -820,7 +897,8 @@ blit2d_init_color_pipeline(struct radv_device *device,
 
 static VkResult
 blit2d_init_depth_only_pipeline(struct radv_device *device,
-				enum blit2d_src_type src_type)
+				enum blit2d_src_type src_type,
+				uint32_t log2_samples)
 {
 	VkResult result;
 	const char *name;
@@ -847,7 +925,7 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 	const VkPipelineVertexInputStateCreateInfo *vi_create_info;
 	struct radv_shader_module fs = { .nir = NULL };
 
-	fs.nir = build_nir_copy_fragment_shader_depth(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D);
+	fs.nir = build_nir_copy_fragment_shader_depth(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D, log2_samples > 0);
 	vi_create_info = &normal_vi_create_info;
 
 	struct radv_shader_module vs = {
@@ -871,7 +949,7 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 	};
 
 	for (enum radv_blit_ds_layout ds_layout = RADV_BLIT_DS_LAYOUT_TILE_ENABLE; ds_layout < RADV_BLIT_DS_LAYOUT_COUNT; ds_layout++) {
-		if (!device->meta_state.blit2d.depth_only_rp[ds_layout]) {
+		if (!device->meta_state.blit2d_depth_only_rp[ds_layout]) {
 			VkImageLayout layout = radv_meta_blit_ds_to_layout(ds_layout);
 			result = radv_CreateRenderPass(radv_device_to_handle(device),
 						       &(VkRenderPassCreateInfo) {
@@ -899,7 +977,7 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 								       .pPreserveAttachments = (uint32_t[]) { 0 },
 							       },
 							       .dependencyCount = 0,
-							}, &device->meta_state.alloc, &device->meta_state.blit2d.depth_only_rp[ds_layout]);
+							}, &device->meta_state.alloc, &device->meta_state.blit2d_depth_only_rp[ds_layout]);
 		}
 	}
 
@@ -927,7 +1005,7 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 		},
 		.pMultisampleState = &(VkPipelineMultisampleStateCreateInfo) {
 			.sType = VK_STRUCTURE_TYPE_PIPELINE_MULTISAMPLE_STATE_CREATE_INFO,
-			.rasterizationSamples = 1,
+			.rasterizationSamples = 1 << log2_samples,
 			.sampleShadingEnable = false,
 			.pSampleMask = (VkSampleMask[]) { UINT32_MAX },
 		},
@@ -958,8 +1036,8 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 			},
 		},
 		.flags = 0,
-		.layout = device->meta_state.blit2d.p_layouts[src_type],
-		.renderPass = device->meta_state.blit2d.depth_only_rp[0],
+		.layout = device->meta_state.blit2d[log2_samples].p_layouts[src_type],
+		.renderPass = device->meta_state.blit2d_depth_only_rp[0],
 		.subpass = 0,
 	};
 
@@ -971,7 +1049,7 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 					       radv_pipeline_cache_to_handle(&device->meta_state.cache),
 					       &vk_pipeline_info, &radv_pipeline_info,
 					       &device->meta_state.alloc,
-					       &device->meta_state.blit2d.depth_only_pipeline[src_type]);
+					       &device->meta_state.blit2d[log2_samples].depth_only_pipeline[src_type]);
 
 
 	ralloc_free(vs.nir);
@@ -982,7 +1060,8 @@ blit2d_init_depth_only_pipeline(struct radv_device *device,
 
 static VkResult
 blit2d_init_stencil_only_pipeline(struct radv_device *device,
-				  enum blit2d_src_type src_type)
+				  enum blit2d_src_type src_type,
+				  uint32_t log2_samples)
 {
 	VkResult result;
 	const char *name;
@@ -1009,7 +1088,7 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 	const VkPipelineVertexInputStateCreateInfo *vi_create_info;
 	struct radv_shader_module fs = { .nir = NULL };
 
-	fs.nir = build_nir_copy_fragment_shader_stencil(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D);
+	fs.nir = build_nir_copy_fragment_shader_stencil(device, src_func, name, src_type == BLIT2D_SRC_TYPE_IMAGE_3D, log2_samples > 0);
 	vi_create_info = &normal_vi_create_info;
 
 	struct radv_shader_module vs = {
@@ -1033,7 +1112,7 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 	};
 
 	for (enum radv_blit_ds_layout ds_layout = RADV_BLIT_DS_LAYOUT_TILE_ENABLE; ds_layout < RADV_BLIT_DS_LAYOUT_COUNT; ds_layout++) {
-		if (!device->meta_state.blit2d.stencil_only_rp[ds_layout]) {
+		if (!device->meta_state.blit2d_stencil_only_rp[ds_layout]) {
 			VkImageLayout layout = radv_meta_blit_ds_to_layout(ds_layout);
 			result = radv_CreateRenderPass(radv_device_to_handle(device),
 						       &(VkRenderPassCreateInfo) {
@@ -1061,7 +1140,7 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 								       .pPreserveAttachments = (uint32_t[]) { 0 },
 							       },
 							       .dependencyCount = 0,
-						       }, &device->meta_state.alloc, &device->meta_state.blit2d.stencil_only_rp[ds_layout]);
+						       }, &device->meta_state.alloc, &device->meta_state.blit2d_stencil_only_rp[ds_layout]);
 		}
 	}
 
@@ -1089,7 +1168,7 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 		},
 		.pMultisampleState = &(VkPipelineMultisampleStateCreateInfo) {
 			.sType = VK_STRUCTURE_TYPE_PIPELINE_MULTISAMPLE_STATE_CREATE_INFO,
-			.rasterizationSamples = 1,
+			.rasterizationSamples = 1 << log2_samples,
 			.sampleShadingEnable = false,
 			.pSampleMask = (VkSampleMask[]) { UINT32_MAX },
 		},
@@ -1136,8 +1215,8 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 			},
 		},
 		.flags = 0,
-		.layout = device->meta_state.blit2d.p_layouts[src_type],
-		.renderPass = device->meta_state.blit2d.stencil_only_rp[0],
+		.layout = device->meta_state.blit2d[log2_samples].p_layouts[src_type],
+		.renderPass = device->meta_state.blit2d_stencil_only_rp[0],
 		.subpass = 0,
 	};
 
@@ -1149,7 +1228,7 @@ blit2d_init_stencil_only_pipeline(struct radv_device *device,
 					       radv_pipeline_cache_to_handle(&device->meta_state.cache),
 					       &vk_pipeline_info, &radv_pipeline_info,
 					       &device->meta_state.alloc,
-					       &device->meta_state.blit2d.stencil_only_pipeline[src_type]);
+					       &device->meta_state.blit2d[log2_samples].stencil_only_pipeline[src_type]);
 
 
 	ralloc_free(vs.nir);
@@ -1175,15 +1254,16 @@ static VkFormat pipeline_formats[] = {
 
 static VkResult
 meta_blit2d_create_pipe_layout(struct radv_device *device,
-			       int idx)
+			       int idx,
+			       uint32_t log2_samples)
 {
 	VkResult result;
 	VkDescriptorType desc_type = (idx == BLIT2D_SRC_TYPE_BUFFER) ? VK_DESCRIPTOR_TYPE_UNIFORM_TEXEL_BUFFER : VK_DESCRIPTOR_TYPE_SAMPLED_IMAGE;
 	const VkPushConstantRange push_constant_ranges[] = {
 		{VK_SHADER_STAGE_VERTEX_BIT, 0, 16},
-		{VK_SHADER_STAGE_FRAGMENT_BIT, 16, 4},
+		{VK_SHADER_STAGE_FRAGMENT_BIT, 16, 12},
 	};
-	int num_push_constant_range = (idx != BLIT2D_SRC_TYPE_IMAGE) ? 2 : 1;
+	int num_push_constant_range = (idx != BLIT2D_SRC_TYPE_IMAGE || log2_samples > 0) ? 2 : 1;
 
 	result = radv_CreateDescriptorSetLayout(radv_device_to_handle(device),
 						&(VkDescriptorSetLayoutCreateInfo) {
@@ -1199,7 +1279,7 @@ meta_blit2d_create_pipe_layout(struct radv_device *device,
 								.pImmutableSamplers = NULL
 							},
 							}
-						}, &device->meta_state.alloc, &device->meta_state.blit2d.ds_layouts[idx]);
+						}, &device->meta_state.alloc, &device->meta_state.blit2d[log2_samples].ds_layouts[idx]);
 	if (result != VK_SUCCESS)
 		goto fail;
 
@@ -1207,11 +1287,11 @@ meta_blit2d_create_pipe_layout(struct radv_device *device,
 					   &(VkPipelineLayoutCreateInfo) {
 						   .sType = VK_STRUCTURE_TYPE_PIPELINE_LAYOUT_CREATE_INFO,
 							   .setLayoutCount = 1,
-							   .pSetLayouts = &device->meta_state.blit2d.ds_layouts[idx],
+							   .pSetLayouts = &device->meta_state.blit2d[log2_samples].ds_layouts[idx],
 							   .pushConstantRangeCount = num_push_constant_range,
 							   .pPushConstantRanges = push_constant_ranges,
 							   },
-					   &device->meta_state.alloc, &device->meta_state.blit2d.p_layouts[idx]);
+					   &device->meta_state.alloc, &device->meta_state.blit2d[log2_samples].p_layouts[idx]);
 	if (result != VK_SUCCESS)
 		goto fail;
 	return VK_SUCCESS;
@@ -1225,27 +1305,33 @@ radv_device_init_meta_blit2d_state(struct radv_device *device)
 	VkResult result;
 	bool create_3d = device->physical_device->rad_info.chip_class >= GFX9;
 
-	for (unsigned src = 0; src < BLIT2D_NUM_SRC_TYPES; src++) {
-		if (src == BLIT2D_SRC_TYPE_IMAGE_3D && !create_3d)
-			continue;
+	for (unsigned log2_samples = 0; log2_samples < 1 + MAX_SAMPLES_LOG2; log2_samples++) {
+		for (unsigned src = 0; src < BLIT2D_NUM_SRC_TYPES; src++) {
+			if (src == BLIT2D_SRC_TYPE_IMAGE_3D && !create_3d)
+				continue;
 
-		result = meta_blit2d_create_pipe_layout(device, src);
-		if (result != VK_SUCCESS)
-			goto fail;
+			/* Don't need to handle copies between buffers and multisample images. */
+			if (src == BLIT2D_SRC_TYPE_BUFFER && log2_samples > 0)
+				continue;
 
-		for (unsigned j = 0; j < ARRAY_SIZE(pipeline_formats); ++j) {
-			result = blit2d_init_color_pipeline(device, src, pipeline_formats[j]);
+			result = meta_blit2d_create_pipe_layout(device, src, log2_samples);
+			if (result != VK_SUCCESS)
+				goto fail;
+
+			for (unsigned j = 0; j < ARRAY_SIZE(pipeline_formats); ++j) {
+				result = blit2d_init_color_pipeline(device, src, pipeline_formats[j], log2_samples);
+				if (result != VK_SUCCESS)
+					goto fail;
+			}
+
+			result = blit2d_init_depth_only_pipeline(device, src, log2_samples);
+			if (result != VK_SUCCESS)
+				goto fail;
+
+			result = blit2d_init_stencil_only_pipeline(device, src, log2_samples);
 			if (result != VK_SUCCESS)
 				goto fail;
 		}
-
-		result = blit2d_init_depth_only_pipeline(device, src);
-		if (result != VK_SUCCESS)
-			goto fail;
-
-		result = blit2d_init_stencil_only_pipeline(device, src);
-		if (result != VK_SUCCESS)
-			goto fail;
 	}
 
 	return VK_SUCCESS;

src/amd/vulkan/radv_pass.c

@@ -87,8 +87,8 @@ VkResult radv_CreateRenderPass(
 		subpass_attachment_count +=
 			desc->inputAttachmentCount +
 			desc->colorAttachmentCount +
-			/* Count colorAttachmentCount again for resolve_attachments */
-			desc->colorAttachmentCount;
+			(desc->pResolveAttachments ? desc->colorAttachmentCount : 0) +
+			(desc->pDepthStencilAttachment != NULL);
 	}
 
 	if (subpass_attachment_count) {

src/amd/vulkan/radv_private.h

@@ -465,18 +465,18 @@ struct radv_meta_state {
 	} blit;
 
 	struct {
-		VkRenderPass render_passes[NUM_META_FS_KEYS][RADV_META_DST_LAYOUT_COUNT];
+		VkPipelineLayout p_layouts[5];
+		VkDescriptorSetLayout ds_layouts[5];
+		VkPipeline pipelines[5][NUM_META_FS_KEYS];
 
-		VkPipelineLayout p_layouts[3];
-		VkDescriptorSetLayout ds_layouts[3];
-		VkPipeline pipelines[3][NUM_META_FS_KEYS];
+		VkPipeline depth_only_pipeline[5];
 
-		VkRenderPass depth_only_rp[RADV_BLIT_DS_LAYOUT_COUNT];
-		VkPipeline depth_only_pipeline[3];
+		VkPipeline stencil_only_pipeline[5];
+	} blit2d[1 + MAX_SAMPLES_LOG2];
 
-		VkRenderPass stencil_only_rp[RADV_BLIT_DS_LAYOUT_COUNT];
-		VkPipeline stencil_only_pipeline[3];
-	} blit2d;
+	VkRenderPass blit2d_render_passes[NUM_META_FS_KEYS][RADV_META_DST_LAYOUT_COUNT];
+	VkRenderPass blit2d_depth_only_rp[RADV_BLIT_DS_LAYOUT_COUNT];
+	VkRenderPass blit2d_stencil_only_rp[RADV_BLIT_DS_LAYOUT_COUNT];
 
 	struct {
 		VkPipelineLayout                          img_p_layout;

src/amd/vulkan/vk_format.h

@@ -488,4 +488,11 @@ vk_to_non_srgb_format(VkFormat format)
 	}
 }
 
+static inline unsigned
+vk_format_get_nr_components(VkFormat format)
+{
+	const struct vk_format_description *desc = vk_format_description(format);
+	return desc->nr_channels;
+}
+
 #endif /* VK_FORMAT_H */

src/amd/vulkan/winsys/amdgpu/radv_amdgpu_bo.c

@@ -501,6 +501,7 @@ radv_amdgpu_winsys_bo_from_fd(struct radeon_winsys *_ws,
 	bo->size = result.alloc_size;
 	bo->is_shared = true;
 	bo->ws = ws;
+	bo->ref_count = 1;
 	radv_amdgpu_add_buffer_to_global_list(bo);
 	return (struct radeon_winsys_bo *)bo;
 error_va_map:

src/compiler/glsl/ast.h

@@ -474,7 +474,13 @@ enum {
 
 struct ast_type_qualifier {
    DECLARE_RALLOC_CXX_OPERATORS(ast_type_qualifier);
-   DECLARE_BITSET_T(bitset_t, 128);
+   /* Note: this bitset needs to have at least as many bits as the 'q'
+    * struct has flags, below.  Previously, the size was 128 instead of 96.
+    * But an apparent bug in GCC 5.4.0 causes bad SSE code generation
+    * elsewhere, leading to a crash.  96 bits works around the issue.
+    * See https://bugs.freedesktop.org/show_bug.cgi?id=105497
+    */
+   DECLARE_BITSET_T(bitset_t, 96);
 
    union flags {
       struct {

src/compiler/glsl/glcpp/tests/glcpp_test.py

@@ -78,7 +78,8 @@ def _valgrind(glcpp, filename):
     extra_args = parse_test_file(filename, nl_format='\n')
 
     try:
-        _, tmpfile = tempfile.mkstemp()
+        fd, tmpfile = tempfile.mkstemp()
+        os.close(fd)
         with open(filename, 'rb') as f:
             proc = subprocess.Popen(
                 ['valgrind', '--error-exitcode=31', '--log-file', tmpfile, glcpp] + extra_args,
@@ -135,7 +136,8 @@ def _replace_test(args, replace):
         total += 1
         testfile = os.path.join(args.testdir, filename)
         try:
-            _, tmpfile = tempfile.mkstemp()
+            fd, tmpfile = tempfile.mkstemp()
+            os.close(fd)
             with io.open(testfile, 'rt') as f:
                 contents = f.read()
             with io.open(tmpfile, 'wt') as f:

src/compiler/nir/nir_lower_returns.c

@@ -37,6 +37,8 @@ struct lower_returns_state {
     * needs to be predicated on the return flag variable.
     */
    bool has_predicated_return;
+
+   bool removed_unreachable_code;
 };
 
 static bool lower_returns_in_cf_list(struct exec_list *cf_list,
@@ -162,8 +164,9 @@ lower_returns_in_block(nir_block *block, struct lower_returns_state *state)
           */
          return false;
       } else {
+         state->removed_unreachable_code = true;
          nir_cf_delete(&list);
-         return true;
+         return false;
       }
    }
 
@@ -262,9 +265,11 @@ nir_lower_returns_impl(nir_function_impl *impl)
    state.loop = NULL;
    state.return_flag = NULL;
    state.has_predicated_return = false;
+   state.removed_unreachable_code = false;
    nir_builder_init(&state.builder, impl);
 
    bool progress = lower_returns_in_cf_list(&impl->body, &state);
+   progress = progress || state.removed_unreachable_code;
 
    if (progress) {
       nir_metadata_preserve(impl, nir_metadata_none);

src/compiler/spirv/vtn_variables.c

@@ -1419,11 +1419,11 @@ apply_var_decoration(struct vtn_builder *b, nir_variable *nir_var,
       case SpvBuiltInTessLevelInner:
          nir_var->data.compact = true;
          break;
-      case SpvBuiltInSamplePosition:
-         nir_var->data.origin_upper_left = b->origin_upper_left;
-         /* fallthrough */
       case SpvBuiltInFragCoord:
          nir_var->data.pixel_center_integer = b->pixel_center_integer;
+         /* fallthrough */
+      case SpvBuiltInSamplePosition:
+         nir_var->data.origin_upper_left = b->origin_upper_left;
          break;
       default:
          break;

src/egl/drivers/dri2/platform_x11.c

@@ -864,19 +864,22 @@ dri2_x11_swap_buffers_msc(_EGLDriver *drv, _EGLDisplay *disp, _EGLSurface *draw,
    if (draw->Type == EGL_PIXMAP_BIT || draw->Type == EGL_PBUFFER_BIT)
       return 0;
 
-   if (draw->SwapBehavior == EGL_BUFFER_PRESERVED || !dri2_dpy->swap_available)
-      return dri2_copy_region(drv, disp, draw, dri2_surf->region) ? 0 : -1;
+   if (draw->SwapBehavior == EGL_BUFFER_PRESERVED || !dri2_dpy->swap_available) {
+      swap_count = dri2_copy_region(drv, disp, draw, dri2_surf->region) ? 0 : -1;
+   } else {
+      dri2_flush_drawable_for_swapbuffers(disp, draw);
 
-   dri2_flush_drawable_for_swapbuffers(disp, draw);
+      cookie = xcb_dri2_swap_buffers_unchecked(dri2_dpy->conn,
+                                               dri2_surf->drawable, msc_hi,
+                                               msc_lo, divisor_hi, divisor_lo,
+                                               remainder_hi, remainder_lo);
 
-   cookie = xcb_dri2_swap_buffers_unchecked(dri2_dpy->conn, dri2_surf->drawable,
-                  msc_hi, msc_lo, divisor_hi, divisor_lo, remainder_hi, remainder_lo);
+      reply = xcb_dri2_swap_buffers_reply(dri2_dpy->conn, cookie, NULL);
 
-   reply = xcb_dri2_swap_buffers_reply(dri2_dpy->conn, cookie, NULL);
-
-   if (reply) {
-      swap_count = (((int64_t)reply->swap_hi) << 32) | reply->swap_lo;
-      free(reply);
+      if (reply) {
+         swap_count = (((int64_t)reply->swap_hi) << 32) | reply->swap_lo;
+         free(reply);
+      }
    }
 
    /* Since we aren't watching for the server's invalidate events like we're

src/gallium/auxiliary/pipe-loader/pipe_loader_drm.c

@@ -207,6 +207,7 @@ pipe_loader_drm_probe_fd(struct pipe_loader_device **dev, int fd)
    if (ddev->lib)
       util_dl_close(ddev->lib);
 #endif
+   FREE(ddev->base.driver_name);
    FREE(ddev);
    return false;
 }

src/gallium/auxiliary/util/u_hash_table.c

@@ -270,6 +270,23 @@ util_hash_table_foreach(struct util_hash_table *ht,
 }
 
 
+static enum pipe_error
+util_hash_inc(void *k, void *v, void *d)
+{
+   ++*(size_t *)d;
+   return PIPE_OK;
+}
+
+
+size_t
+util_hash_table_count(struct util_hash_table *ht)
+{
+	size_t count = 0;
+	util_hash_table_foreach(ht, util_hash_inc, &count);
+	return count;
+}
+
+
 void
 util_hash_table_destroy(struct util_hash_table *ht)
 {

src/gallium/auxiliary/util/u_hash_table.h

@@ -85,6 +85,11 @@ util_hash_table_foreach(struct util_hash_table *ht,
                         (void *key, void *value, void *data),
                         void *data);
 
+
+size_t
+util_hash_table_count(struct util_hash_table *ht);
+
+
 void
 util_hash_table_destroy(struct util_hash_table *ht);
 

src/gallium/auxiliary/util/u_transfer_helper.c

@@ -138,8 +138,7 @@ u_transfer_helper_resource_destroy(struct pipe_screen *pscreen,
    if (helper->vtbl->get_stencil) {
       struct pipe_resource *stencil = helper->vtbl->get_stencil(prsc);
 
-      if (stencil)
-         helper->vtbl->resource_destroy(pscreen, stencil);
+      pipe_resource_reference(&stencil, NULL);
    }
 
    helper->vtbl->resource_destroy(pscreen, prsc);

src/gallium/drivers/r600/evergreen_compute.c

@@ -461,11 +461,10 @@ static void evergreen_delete_compute_state(struct pipe_context *ctx, void *state
 	} else {
 #ifdef HAVE_OPENCL
 		radeon_shader_binary_clean(&shader->binary);
+		pipe_resource_reference(&shader->code_bo, NULL);
+		pipe_resource_reference(&shader->kernel_param, NULL);
 #endif
 		r600_destroy_shader(&shader->bc);
-
-		/* TODO destroy shader->code_bo, shader->const_bo
-		 * we'll need something like r600_buffer_free */
 	}
 	FREE(shader);
 }

src/gallium/drivers/r600/evergreen_state.c

@@ -2202,7 +2202,7 @@ static void evergreen_emit_constant_buffers(struct r600_context *rctx,
 		radeon_emit(cs, PKT3(PKT3_SET_RESOURCE, 8, 0) | pkt_flags);
 		radeon_emit(cs, (buffer_id_base + buffer_index) * 8);
 		radeon_emit(cs, va); /* RESOURCEi_WORD0 */
-		radeon_emit(cs, rbuffer->b.b.width0 - cb->buffer_offset - 1); /* RESOURCEi_WORD1 */
+		radeon_emit(cs, cb->buffer_size -1); /* RESOURCEi_WORD1 */
 		radeon_emit(cs, /* RESOURCEi_WORD2 */
 			    S_030008_ENDIAN_SWAP(gs_ring_buffer ? ENDIAN_NONE : r600_endian_swap(32)) |
 			    S_030008_STRIDE(gs_ring_buffer ? 4 : 16) |

src/gallium/drivers/r600/r600_pipe.c

@@ -65,7 +65,7 @@ static const struct debug_named_value r600_debug_options[] = {
 static void r600_destroy_context(struct pipe_context *context)
 {
 	struct r600_context *rctx = (struct r600_context *)context;
-	unsigned sh;
+	unsigned sh, i;
 
 	r600_isa_destroy(rctx->isa);
 
@@ -104,6 +104,10 @@ static void r600_destroy_context(struct pipe_context *context)
 	}
 	util_unreference_framebuffer_state(&rctx->framebuffer.state);
 
+	for (sh = 0; sh < PIPE_SHADER_TYPES; ++sh)
+		for (i = 0; i < PIPE_MAX_CONSTANT_BUFFERS; ++i)
+			rctx->b.b.set_constant_buffer(context, sh, i, NULL);
+
 	if (rctx->blitter) {
 		util_blitter_destroy(rctx->blitter);
 	}

src/gallium/drivers/r600/r600_state.c

@@ -1729,7 +1729,7 @@ static void r600_emit_constant_buffers(struct r600_context *rctx,
 		radeon_emit(cs, PKT3(PKT3_SET_RESOURCE, 7, 0));
 		radeon_emit(cs, (buffer_id_base + buffer_index) * 7);
 		radeon_emit(cs, offset); /* RESOURCEi_WORD0 */
-		radeon_emit(cs, rbuffer->b.b.width0 - offset - 1); /* RESOURCEi_WORD1 */
+		radeon_emit(cs, cb->buffer_size - 1); /* RESOURCEi_WORD1 */
 		radeon_emit(cs, /* RESOURCEi_WORD2 */
 			    S_038008_ENDIAN_SWAP(gs_ring_buffer ? ENDIAN_NONE : r600_endian_swap(32)) |
 			    S_038008_STRIDE(gs_ring_buffer ? 4 : 16));

src/gallium/drivers/radeonsi/si_get.c

@@ -477,12 +477,28 @@ static int si_get_shader_param(struct pipe_screen* pscreen,
 
 	case PIPE_SHADER_CAP_INDIRECT_INPUT_ADDR:
 		/* TODO: Indirect indexing of GS inputs is unimplemented. */
-		return shader != PIPE_SHADER_GEOMETRY &&
-		       (sscreen->llvm_has_working_vgpr_indexing ||
-			/* TCS and TES load inputs directly from LDS or
-			 * offchip memory, so indirect indexing is trivial. */
-			shader == PIPE_SHADER_TESS_CTRL ||
-			shader == PIPE_SHADER_TESS_EVAL);
+		if (shader == PIPE_SHADER_GEOMETRY)
+			return 0;
+
+		if (shader == PIPE_SHADER_VERTEX &&
+		    !sscreen->llvm_has_working_vgpr_indexing)
+			return 0;
+
+		/* Doing indirect indexing on GFX9 with LLVM 6.0 hangs.
+		 * This means we don't support INTERP instructions with
+		 * indirect indexing on inputs.
+		 */
+		if (shader == PIPE_SHADER_FRAGMENT &&
+		    !sscreen->llvm_has_working_vgpr_indexing &&
+		    HAVE_LLVM < 0x0700)
+			return 0;
+
+		/* TCS and TES load inputs directly from LDS or offchip
+		 * memory, so indirect indexing is always supported.
+		 * PS has to support indirect indexing, because we can't
+		 * lower that to TEMPs for INTERP instructions.
+		 */
+		return 1;
 
 	case PIPE_SHADER_CAP_INDIRECT_OUTPUT_ADDR:
 		return sscreen->llvm_has_working_vgpr_indexing ||

src/gallium/state_trackers/omx/vid_enc_common.c

@@ -353,7 +353,7 @@ OMX_ERRORTYPE enc_LoadImage_common(vid_enc_PrivateType * priv, OMX_VIDEO_PORTDEF
       blit.src.resource = inp->resource;
       blit.src.format = inp->resource->format;
 
-      blit.src.box.x = 0;
+      blit.src.box.x = -1;
       blit.src.box.y = def->nFrameHeight;
       blit.src.box.width = def->nFrameWidth;
       blit.src.box.height = def->nFrameHeight / 2 ;
@@ -367,11 +367,11 @@ OMX_ERRORTYPE enc_LoadImage_common(vid_enc_PrivateType * priv, OMX_VIDEO_PORTDEF
       blit.dst.box.depth = 1;
       blit.filter = PIPE_TEX_FILTER_NEAREST;
 
-      blit.mask = PIPE_MASK_G;
+      blit.mask = PIPE_MASK_R;
       priv->s_pipe->blit(priv->s_pipe, &blit);
 
-      blit.src.box.x = 1;
-      blit.mask = PIPE_MASK_R;
+      blit.src.box.x = 0;
+      blit.mask = PIPE_MASK_G;
       priv->s_pipe->blit(priv->s_pipe, &blit);
       priv->s_pipe->flush(priv->s_pipe, NULL, 0);
 

src/gallium/targets/graw-xlib/meson.build

@@ -21,10 +21,9 @@
 libgraw_xlib = shared_library(
   'graw_xlib',
   ['graw_xlib.c'],
-  c_args : ['-DGALLIUM_TRACE', '-DGALLIUM_RBUG'],
   include_directories : [inc_common, inc_gallium_drivers, inc_gallium_winsys],
   link_with : [
-    libgraw_util, libtrace, librbug, libmesa_util, libgallium, libws_xlib
+    libgraw_util, libmesa_util, libgallium, libws_xlib
   ],
   dependencies : [dep_thread, driver_swrast],
   version : '1.0',

src/gallium/tests/unit/Makefile.am

@@ -2,7 +2,7 @@
 
 include $(top_srcdir)/src/gallium/Automake.inc
 
-EXTRA_DIST = SConscript
+EXTRA_DIST = SConscript meson.build
 
 AM_CFLAGS = \
 	$(GALLIUM_CFLAGS)
@@ -32,5 +32,3 @@ u_format_test_SOURCES = u_format_test.c
 u_format_compatible_test_SOURCES = u_format_compatible_test.c
 
 translate_test_SOURCES = translate_test.c
-
-EXTRA_DIST = meson.build

src/gallium/winsys/amdgpu/drm/amdgpu_winsys.c

@@ -220,8 +220,13 @@ static bool amdgpu_winsys_unref(struct radeon_winsys *rws)
    simple_mtx_lock(&dev_tab_mutex);
 
    destroy = pipe_reference(&ws->reference, NULL);
-   if (destroy && dev_tab)
+   if (destroy && dev_tab) {
       util_hash_table_remove(dev_tab, ws->dev);
+      if (util_hash_table_count(dev_tab) == 0) {
+         util_hash_table_destroy(dev_tab);
+         dev_tab = NULL;
+      }
+   }
 
    simple_mtx_unlock(&dev_tab_mutex);
    return destroy;

src/gallium/winsys/radeon/drm/radeon_drm_winsys.c

@@ -719,8 +719,13 @@ static bool radeon_winsys_unref(struct radeon_winsys *ws)
     mtx_lock(&fd_tab_mutex);
 
     destroy = pipe_reference(&rws->reference, NULL);
-    if (destroy && fd_tab)
+    if (destroy && fd_tab) {
         util_hash_table_remove(fd_tab, intptr_to_pointer(rws->fd));
+        if (util_hash_table_count(fd_tab) == 0) {
+           util_hash_table_destroy(fd_tab);
+           fd_tab = NULL;
+        }
+    }
 
     mtx_unlock(&fd_tab_mutex);
     return destroy;

src/intel/Makefile.vulkan.am

@@ -64,10 +64,12 @@ EXTRA_DIST += \
 	vulkan/TODO
 
 vulkan/dev_icd.json : vulkan/anv_extensions.py vulkan/anv_icd.py
+	$(MKDIR_GEN)
 	$(AM_V_GEN)$(PYTHON2) $(srcdir)/vulkan/anv_icd.py \
 		--lib-path="${abs_top_builddir}/${LIB_DIR}" --out $@
 
 vulkan/intel_icd.@host_cpu@.json : vulkan/anv_extensions.py vulkan/anv_icd.py
+	$(MKDIR_GEN)
 	$(AM_V_GEN)$(PYTHON2) $(srcdir)/vulkan/anv_icd.py \
 		--lib-path="${libdir}" --out $@
 

src/intel/compiler/brw_fs.cpp

@@ -842,6 +842,8 @@ fs_inst::size_read(int arg) const
    case SHADER_OPCODE_TYPED_ATOMIC:
    case SHADER_OPCODE_TYPED_SURFACE_READ:
    case SHADER_OPCODE_TYPED_SURFACE_WRITE:
+   case FS_OPCODE_INTERPOLATE_AT_SAMPLE:
+   case FS_OPCODE_INTERPOLATE_AT_SHARED_OFFSET:
    case FS_OPCODE_INTERPOLATE_AT_PER_SLOT_OFFSET:
    case SHADER_OPCODE_BYTE_SCATTERED_WRITE:
    case SHADER_OPCODE_BYTE_SCATTERED_READ:

src/intel/compiler/brw_ir_vec4.h

@@ -334,6 +334,31 @@ public:
                                   opcode != BRW_OPCODE_IF &&
                                   opcode != BRW_OPCODE_WHILE));
    }
+
+   bool reads_g0_implicitly() const
+   {
+      switch (opcode) {
+      case SHADER_OPCODE_TEX:
+      case SHADER_OPCODE_TXL:
+      case SHADER_OPCODE_TXD:
+      case SHADER_OPCODE_TXF:
+      case SHADER_OPCODE_TXF_CMS_W:
+      case SHADER_OPCODE_TXF_CMS:
+      case SHADER_OPCODE_TXF_MCS:
+      case SHADER_OPCODE_TXS:
+      case SHADER_OPCODE_TG4:
+      case SHADER_OPCODE_TG4_OFFSET:
+      case SHADER_OPCODE_SAMPLEINFO:
+      case VS_OPCODE_PULL_CONSTANT_LOAD:
+      case GS_OPCODE_SET_PRIMITIVE_ID:
+      case GS_OPCODE_GET_INSTANCE_ID:
+      case SHADER_OPCODE_GEN4_SCRATCH_READ:
+      case SHADER_OPCODE_GEN4_SCRATCH_WRITE:
+         return true;
+      default:
+         return false;
+      }
+   }
 };
 
 /**

src/intel/compiler/brw_reg.h

@@ -705,7 +705,7 @@ static inline struct brw_reg
 brw_imm_w(int16_t w)
 {
    struct brw_reg imm = brw_imm_reg(BRW_REGISTER_TYPE_W);
-   imm.d = w | (w << 16);
+   imm.ud = (uint16_t)w | (uint32_t)(uint16_t)w << 16;
    return imm;
 }
 

src/intel/compiler/brw_schedule_instructions.cpp

@@ -1267,6 +1267,9 @@ vec4_instruction_scheduler::calculate_deps()
          }
       }
 
+      if (inst->reads_g0_implicitly())
+         add_dep(last_fixed_grf_write, n);
+
       if (!inst->is_send_from_grf()) {
          for (int i = 0; i < inst->mlen; i++) {
             /* It looks like the MRF regs are released in the send

src/intel/compiler/brw_shader.cpp

@@ -580,9 +580,11 @@ brw_negate_immediate(enum brw_reg_type type, struct brw_reg *reg)
       reg->d = -reg->d;
       return true;
    case BRW_REGISTER_TYPE_W:
-   case BRW_REGISTER_TYPE_UW:
-      reg->d = -(int16_t)reg->ud;
+   case BRW_REGISTER_TYPE_UW: {
+      uint16_t value = -(int16_t)reg->ud;
+      reg->ud = value | (uint32_t)value << 16;
       return true;
+   }
    case BRW_REGISTER_TYPE_F:
       reg->f = -reg->f;
       return true;
@@ -618,9 +620,11 @@ brw_abs_immediate(enum brw_reg_type type, struct brw_reg *reg)
    case BRW_REGISTER_TYPE_D:
       reg->d = abs(reg->d);
       return true;
-   case BRW_REGISTER_TYPE_W:
-      reg->d = abs((int16_t)reg->ud);
+   case BRW_REGISTER_TYPE_W: {
+      uint16_t value = abs((int16_t)reg->ud);
+      reg->ud = value | (uint32_t)value << 16;
       return true;
+   }
    case BRW_REGISTER_TYPE_F:
       reg->f = fabsf(reg->f);
       return true;

src/intel/dev/gen_device_info.c

@@ -1044,11 +1044,7 @@ gen_device_info_update_from_topology(struct gen_device_info *devinfo,
    for (int b = 0; b < eu_mask_len; b++)
       n_eus += __builtin_popcount(devinfo->eu_masks[b]);
 
-   /* We expect the total number of EUs to be uniformly distributed throughout
-    * the subslices.
-    */
-   assert(n_subslices && (n_eus % n_subslices) == 0);
-   devinfo->num_eu_per_subslice = n_eus / n_subslices;
+   devinfo->num_eu_per_subslice = DIV_ROUND_UP(n_eus, n_subslices);
 }
 
 bool

src/intel/vulkan/anv_allocator.c

@@ -508,12 +508,12 @@ anv_block_pool_grow(struct anv_block_pool *pool, struct anv_block_state *state)
       assert(center_bo_offset >= back_used);
 
       /* Make sure we don't shrink the back end of the pool */
-      if (center_bo_offset < pool->back_state.end)
-         center_bo_offset = pool->back_state.end;
+      if (center_bo_offset < back_required)
+         center_bo_offset = back_required;
 
       /* Make sure that we don't shrink the front end of the pool */
-      if (size - center_bo_offset < pool->state.end)
-         center_bo_offset = size - pool->state.end;
+      if (size - center_bo_offset < front_required)
+         center_bo_offset = size - front_required;
    }
 
    assert(center_bo_offset % PAGE_SIZE == 0);

src/intel/vulkan/anv_device.c

@@ -1179,6 +1179,7 @@ PFN_vkVoidFunction anv_GetInstanceProcAddr(
 
    LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceExtensionProperties);
    LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceLayerProperties);
+   LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceVersion);
    LOOKUP_ANV_ENTRYPOINT(CreateInstance);
 
 #undef LOOKUP_ANV_ENTRYPOINT

src/intel/vulkan/genX_cmd_buffer.c

@@ -1420,10 +1420,18 @@ genX(BeginCommandBuffer)(
  * context restore, so the mentioned hang doesn't happen. However,
  * software must program push constant commands for all stages prior to
  * rendering anything. So we flag them dirty in BeginCommandBuffer.
+ *
+ * Finally, we also make sure to stall at pixel scoreboard to make sure the
+ * constants have been loaded into the EUs prior to disable the push constants
+ * so that it doesn't hang a previous 3DPRIMITIVE.
  */
 static void
 emit_isp_disable(struct anv_cmd_buffer *cmd_buffer)
 {
+   anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
+         pc.StallAtPixelScoreboard = true;
+         pc.CommandStreamerStallEnable = true;
+   }
    anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
          pc.IndirectStatePointersDisable = true;
          pc.CommandStreamerStallEnable = true;

src/loader/loader_dri3_helper.c

@@ -420,13 +420,6 @@ dri3_handle_present_event(struct loader_dri3_drawable *draw,
 
          if (buf && buf->pixmap == ie->pixmap)
             buf->busy = 0;
-
-         if (buf && draw->cur_blit_source != b && !buf->busy &&
-             (buf->reallocate ||
-             (draw->num_back <= b && b < LOADER_DRI3_MAX_BACK))) {
-            dri3_free_render_buffer(draw, buf);
-            draw->buffers[b] = NULL;
-         }
       }
       break;
    }
@@ -559,7 +552,6 @@ dri3_find_back(struct loader_dri3_drawable *draw)
    /* Check whether we need to reuse the current back buffer as new back.
     * In that case, wait until it's not busy anymore.
     */
-   dri3_update_num_back(draw);
    num_to_consider = draw->num_back;
    if (!loader_dri3_have_image_blit(draw) && draw->cur_blit_source != -1) {
       num_to_consider = 1;
@@ -1813,6 +1805,7 @@ loader_dri3_get_buffers(__DRIdrawable *driDrawable,
 {
    struct loader_dri3_drawable *draw = loaderPrivate;
    struct loader_dri3_buffer   *front, *back;
+   int buf_id;
 
    buffers->image_mask = 0;
    buffers->front = NULL;
@@ -1824,6 +1817,16 @@ loader_dri3_get_buffers(__DRIdrawable *driDrawable,
    if (!dri3_update_drawable(driDrawable, draw))
       return false;
 
+   dri3_update_num_back(draw);
+
+   /* Free no longer needed back buffers */
+   for (buf_id = draw->num_back; buf_id < LOADER_DRI3_MAX_BACK; buf_id++) {
+      if (draw->cur_blit_source != buf_id && draw->buffers[buf_id]) {
+         dri3_free_render_buffer(draw, draw->buffers[buf_id]);
+         draw->buffers[buf_id] = NULL;
+      }
+   }
+
    /* pixmaps always have front buffers.
     * Exchange swaps also mandate fake front buffers.
     */

src/mesa/drivers/dri/i965/brw_context.c

@@ -1108,7 +1108,6 @@ intelDestroyContext(__DRIcontext * driContextPriv)
    struct brw_context *brw =
       (struct brw_context *) driContextPriv->driverPrivate;
    struct gl_context *ctx = &brw->ctx;
-   const struct gen_device_info *devinfo = &brw->screen->devinfo;
 
    _mesa_meta_free(&brw->ctx);
 
@@ -1120,8 +1119,7 @@ intelDestroyContext(__DRIcontext * driContextPriv)
       brw_destroy_shader_time(brw);
    }
 
-   if (devinfo->gen >= 6)
-      blorp_finish(&brw->blorp);
+   blorp_finish(&brw->blorp);
 
    brw_destroy_state(brw);
    brw_draw_destroy(brw);

src/mesa/drivers/dri/i965/brw_pipe_control.c

@@ -349,10 +349,18 @@ gen7_emit_vs_workaround_flush(struct brw_context *brw)
  * context restore, so the mentioned hang doesn't happen. However,
  * software must program push constant commands for all stages prior to
  * rendering anything, so we flag them as dirty.
+ *
+ * Finally, we also make sure to stall at pixel scoreboard to make sure the
+ * constants have been loaded into the EUs prior to disable the push constants
+ * so that it doesn't hang a previous 3DPRIMITIVE.
  */
 void
 gen10_emit_isp_disable(struct brw_context *brw)
 {
+   brw_emit_pipe_control(brw,
+                         PIPE_CONTROL_STALL_AT_SCOREBOARD |
+                         PIPE_CONTROL_CS_STALL,
+                         NULL, 0, 0);
    brw_emit_pipe_control(brw,
                          PIPE_CONTROL_ISP_DIS |
                          PIPE_CONTROL_CS_STALL,

src/mesa/drivers/dri/i965/intel_tex_image.c

@@ -927,7 +927,7 @@ intelCompressedTexSubImage(struct gl_context *ctx, GLuint dims,
                         !_mesa_is_srgb_format(gl_format);
    struct brw_context *brw = (struct brw_context*) ctx;
    const struct gen_device_info *devinfo = &brw->screen->devinfo;
-   if (devinfo->gen == 9 && is_linear_astc)
+   if (devinfo->gen == 9 && !gen_device_info_is_9lp(devinfo) && is_linear_astc)
       flush_astc_denorms(ctx, dims, texImage,
                          xoffset, yoffset, zoffset,
                          width, height, depth);

src/mesa/main/fbobject.c

@@ -1488,45 +1488,66 @@ _mesa_FramebufferParameteri(GLenum target, GLenum pname, GLint param)
 }
 
 static bool
-_pname_valid_for_default_framebuffer(struct gl_context *ctx,
-                                     GLenum pname)
+validate_get_framebuffer_parameteriv_pname(struct gl_context *ctx,
+                                           struct gl_framebuffer *fb,
+                                           GLuint pname, const char *func)
 {
-   if (!_mesa_is_desktop_gl(ctx))
-      return false;
+   bool cannot_be_winsys_fbo = true;
 
    switch (pname) {
+   case GL_FRAMEBUFFER_DEFAULT_LAYERS:
+      /*
+       * According to the OpenGL ES 3.1 specification section 9.2.3, the
+       * GL_FRAMEBUFFER_LAYERS parameter name is not supported.
+       */
+      if (_mesa_is_gles31(ctx) && !ctx->Extensions.OES_geometry_shader) {
+         _mesa_error(ctx, GL_INVALID_ENUM, "%s(pname=0x%x)", func, pname);
+         return false;
+      }
+      break;
+   case GL_FRAMEBUFFER_DEFAULT_WIDTH:
+   case GL_FRAMEBUFFER_DEFAULT_HEIGHT:
+   case GL_FRAMEBUFFER_DEFAULT_SAMPLES:
+   case GL_FRAMEBUFFER_DEFAULT_FIXED_SAMPLE_LOCATIONS:
+      break;
    case GL_DOUBLEBUFFER:
    case GL_IMPLEMENTATION_COLOR_READ_FORMAT:
    case GL_IMPLEMENTATION_COLOR_READ_TYPE:
    case GL_SAMPLES:
    case GL_SAMPLE_BUFFERS:
    case GL_STEREO:
-      return true;
+      /* From OpenGL 4.5 spec, section 9.2.3 "Framebuffer Object Queries:
+       *
+       *    "An INVALID_OPERATION error is generated by GetFramebufferParameteriv
+       *     if the default framebuffer is bound to target and pname is not one
+       *     of the accepted values from table 23.73, other than
+       *     SAMPLE_POSITION."
+       *
+       * For OpenGL ES, using default framebuffer raises INVALID_OPERATION
+       * for any pname.
+       */
+      cannot_be_winsys_fbo = !_mesa_is_desktop_gl(ctx);
+      break;
    default:
+      _mesa_error(ctx, GL_INVALID_ENUM, "%s(pname=0x%x)", func, pname);
       return false;
    }
+
+   if (cannot_be_winsys_fbo && _mesa_is_winsys_fbo(fb)) {
+      _mesa_error(ctx, GL_INVALID_OPERATION,
+                  "%s(invalid pname=0x%x for default framebuffer)", func, pname);
+      return false;
+   }
+
+   return true;
 }
 
 static void
 get_framebuffer_parameteriv(struct gl_context *ctx, struct gl_framebuffer *fb,
                             GLenum pname, GLint *params, const char *func)
 {
-   /* From OpenGL 4.5 spec, section 9.2.3 "Framebuffer Object Queries:
-    *
-    *    "An INVALID_OPERATION error is generated by GetFramebufferParameteriv
-    *     if the default framebuffer is bound to target and pname is not one
-    *     of the accepted values from table 23.73, other than
-    *     SAMPLE_POSITION."
-    *
-    * For OpenGL ES, using default framebuffer still raises INVALID_OPERATION
-    * for any pname.
-    */
-   if (_mesa_is_winsys_fbo(fb) &&
-       !_pname_valid_for_default_framebuffer(ctx, pname)) {
-      _mesa_error(ctx, GL_INVALID_OPERATION,
-                  "%s(invalid pname=0x%x for default framebuffer)", func, pname);
+   if (!validate_get_framebuffer_parameteriv_pname(ctx, fb, pname, func))
       return;
-   }
 
    switch (pname) {
    case GL_FRAMEBUFFER_DEFAULT_WIDTH:
@@ -1536,14 +1557,6 @@ get_framebuffer_parameteriv(struct gl_context *ctx, struct gl_framebuffer *fb,
       *params = fb->DefaultGeometry.Height;
       break;
    case GL_FRAMEBUFFER_DEFAULT_LAYERS:
-      /*
-       * According to the OpenGL ES 3.1 specification section 9.2.3, the
-       * GL_FRAMEBUFFER_LAYERS parameter name is not supported.
-       */
-      if (_mesa_is_gles31(ctx) && !ctx->Extensions.OES_geometry_shader) {
-         _mesa_error(ctx, GL_INVALID_ENUM, "%s(pname=0x%x)", func, pname);
-         break;
-      }
       *params = fb->DefaultGeometry.Layers;
       break;
    case GL_FRAMEBUFFER_DEFAULT_SAMPLES:
@@ -1570,9 +1583,6 @@ get_framebuffer_parameteriv(struct gl_context *ctx, struct gl_framebuffer *fb,
    case GL_STEREO:
       *params = fb->Visual.stereoMode;
       break;
-   default:
-      _mesa_error(ctx, GL_INVALID_ENUM,
-                  "%s(pname=0x%x)", func, pname);
    }
 }
 

src/mesa/main/fog.c

@@ -62,6 +62,7 @@ _mesa_Fogiv(GLenum pname, const GLint *params )
       case GL_FOG_END:
       case GL_FOG_INDEX:
       case GL_FOG_COORDINATE_SOURCE_EXT:
+      case GL_FOG_DISTANCE_MODE_NV:
 	 p[0] = (GLfloat) *params;
 	 break;
       case GL_FOG_COLOR:

src/mesa/main/get.c

@@ -105,6 +105,8 @@ enum value_type {
    TYPE_ENUM,
    TYPE_ENUM_2,
    TYPE_BOOLEAN,
+   TYPE_UBYTE,
+   TYPE_SHORT,
    TYPE_BIT_0,
    TYPE_BIT_1,
    TYPE_BIT_2,
@@ -188,6 +190,8 @@ union value {
    GLint value_int_4[4];
    GLint64 value_int64;
    GLenum value_enum;
+   GLubyte value_ubyte;
+   GLshort value_short;
 
    /* Sigh, see GL_COMPRESSED_TEXTURE_FORMATS_ARB handling */
    struct {
@@ -235,10 +239,13 @@ union value {
 #define CONTEXT_MATRIX(field) CONTEXT_FIELD(field, TYPE_MATRIX)
 #define CONTEXT_MATRIX_T(field) CONTEXT_FIELD(field, TYPE_MATRIX_T)
 
+/* Vertex array fields */
 #define ARRAY_INT(field) ARRAY_FIELD(field, TYPE_INT)
 #define ARRAY_ENUM(field) ARRAY_FIELD(field, TYPE_ENUM)
 #define ARRAY_ENUM16(field) ARRAY_FIELD(field, TYPE_ENUM16)
 #define ARRAY_BOOL(field) ARRAY_FIELD(field, TYPE_BOOLEAN)
+#define ARRAY_UBYTE(field) ARRAY_FIELD(field, TYPE_UBYTE)
+#define ARRAY_SHORT(field) ARRAY_FIELD(field, TYPE_SHORT)
 
 #define EXT(f)					\
    offsetof(struct gl_extensions, f)
@@ -1517,6 +1524,10 @@ get_value_size(enum value_type type, const union value *v)
       return sizeof(GLenum) * 2;
    case TYPE_BOOLEAN:
       return sizeof(GLboolean);
+   case TYPE_UBYTE:
+      return sizeof(GLubyte);
+   case TYPE_SHORT:
+      return sizeof(GLshort);
    case TYPE_BIT_0:
    case TYPE_BIT_1:
    case TYPE_BIT_2:
@@ -1628,6 +1639,14 @@ _mesa_GetBooleanv(GLenum pname, GLboolean *params)
       params[0] = ((GLboolean*) p)[0];
       break;
 
+   case TYPE_UBYTE:
+      params[0] = INT_TO_BOOLEAN(((GLubyte *) p)[0]);
+      break;
+
+   case TYPE_SHORT:
+      params[0] = INT_TO_BOOLEAN(((GLshort *) p)[0]);
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) p;
       for (i = 0; i < 16; i++)
@@ -1735,6 +1754,14 @@ _mesa_GetFloatv(GLenum pname, GLfloat *params)
       params[0] = BOOLEAN_TO_FLOAT(*(GLboolean*) p);
       break;
 
+   case TYPE_UBYTE:
+      params[0] = (GLfloat) ((GLubyte *) p)[0];
+      break;
+
+   case TYPE_SHORT:
+      params[0] = (GLfloat) ((GLshort *) p)[0];
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) p;
       for (i = 0; i < 16; i++)
@@ -1842,6 +1869,14 @@ _mesa_GetIntegerv(GLenum pname, GLint *params)
       params[0] = BOOLEAN_TO_INT(*(GLboolean*) p);
       break;
 
+   case TYPE_UBYTE:
+      params[0] = ((GLubyte *) p)[0];
+      break;
+
+   case TYPE_SHORT:
+      params[0] = ((GLshort *) p)[0];
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) p;
       for (i = 0; i < 16; i++)
@@ -2062,6 +2097,14 @@ _mesa_GetDoublev(GLenum pname, GLdouble *params)
       params[0] = *(GLboolean*) p;
       break;
 
+   case TYPE_UBYTE:
+      params[0] = ((GLubyte *) p)[0];
+      break;
+
+   case TYPE_SHORT:
+      params[0] = ((GLshort *) p)[0];
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) p;
       for (i = 0; i < 16; i++)
@@ -2141,6 +2184,8 @@ _mesa_GetUnsignedBytevEXT(GLenum pname, GLubyte *data)
    case TYPE_ENUM:
    case TYPE_ENUM_2:
    case TYPE_BOOLEAN:
+   case TYPE_UBYTE:
+   case TYPE_SHORT:
    case TYPE_FLOAT:
    case TYPE_FLOATN:
    case TYPE_FLOAT_2:
@@ -2790,6 +2835,14 @@ _mesa_GetFloati_v(GLenum pname, GLuint index, GLfloat *params)
       params[0] = BOOLEAN_TO_FLOAT(v.value_bool);
       break;
 
+   case TYPE_UBYTE:
+      params[0] = (GLfloat) v.value_ubyte;
+      break;
+
+   case TYPE_SHORT:
+      params[0] = (GLfloat) v.value_short;
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) &v;
       for (i = 0; i < 16; i++)
@@ -2873,6 +2926,14 @@ _mesa_GetDoublei_v(GLenum pname, GLuint index, GLdouble *params)
       params[0] = (GLdouble) BOOLEAN_TO_FLOAT(v.value_bool);
       break;
 
+   case TYPE_UBYTE:
+      params[0] = (GLdouble) v.value_ubyte;
+      break;
+
+   case TYPE_SHORT:
+      params[0] = (GLdouble) v.value_short;
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) &v;
       for (i = 0; i < 16; i++)
@@ -2925,6 +2986,8 @@ _mesa_GetUnsignedBytei_vEXT(GLenum target, GLuint index, GLubyte *data)
    case TYPE_ENUM:
    case TYPE_ENUM_2:
    case TYPE_BOOLEAN:
+   case TYPE_UBYTE:
+   case TYPE_SHORT:
    case TYPE_FLOAT:
    case TYPE_FLOATN:
    case TYPE_FLOAT_2:
@@ -3018,6 +3081,14 @@ _mesa_GetFixedv(GLenum pname, GLfixed *params)
       params[0] = BOOLEAN_TO_FIXED(((GLboolean*) p)[0]);
       break;
 
+   case TYPE_UBYTE:
+      params[0] = INT_TO_FIXED(((GLubyte *) p)[0]);
+      break;
+
+   case TYPE_SHORT:
+      params[0] = INT_TO_FIXED(((GLshort *) p)[0]);
+      break;
+
    case TYPE_MATRIX:
       m = *(GLmatrix **) p;
       for (i = 0; i < 16; i++)

src/mesa/main/get_hash_params.py

@@ -212,20 +212,20 @@ descriptor=[
   [ "TEXTURE_MATRIX", "LOC_CUSTOM, TYPE_MATRIX, 0, extra_valid_texture_unit" ],
   [ "TEXTURE_STACK_DEPTH", "LOC_CUSTOM, TYPE_INT, 0, extra_valid_texture_unit" ],
   [ "VERTEX_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_POS].Enabled), NO_EXTRA" ],
-  [ "VERTEX_ARRAY_SIZE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_POS].Size), NO_EXTRA" ],
+  [ "VERTEX_ARRAY_SIZE", "ARRAY_UBYTE(VertexAttrib[VERT_ATTRIB_POS].Size), NO_EXTRA" ],
   [ "VERTEX_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_POS].Type), NO_EXTRA" ],
-  [ "VERTEX_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_POS].Stride), NO_EXTRA" ],
+  [ "VERTEX_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_POS].Stride), NO_EXTRA" ],
   [ "NORMAL_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_NORMAL].Enabled), NO_EXTRA" ],
   [ "NORMAL_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_NORMAL].Type), NO_EXTRA" ],
-  [ "NORMAL_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_NORMAL].Stride), NO_EXTRA" ],
+  [ "NORMAL_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_NORMAL].Stride), NO_EXTRA" ],
   [ "COLOR_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_COLOR0].Enabled), NO_EXTRA" ],
   [ "COLOR_ARRAY_SIZE", "LOC_CUSTOM, TYPE_INT, 0, NO_EXTRA" ],
   [ "COLOR_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_COLOR0].Type), NO_EXTRA" ],
-  [ "COLOR_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_COLOR0].Stride), NO_EXTRA" ],
+  [ "COLOR_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_COLOR0].Stride), NO_EXTRA" ],
   [ "TEXTURE_COORD_ARRAY", "LOC_CUSTOM, TYPE_BOOLEAN, offsetof(struct gl_array_attributes, Enabled), NO_EXTRA" ],
-  [ "TEXTURE_COORD_ARRAY_SIZE", "LOC_CUSTOM, TYPE_INT, offsetof(struct gl_array_attributes, Size), NO_EXTRA" ],
+  [ "TEXTURE_COORD_ARRAY_SIZE", "LOC_CUSTOM, TYPE_UBYTE, offsetof(struct gl_array_attributes, Size), NO_EXTRA" ],
   [ "TEXTURE_COORD_ARRAY_TYPE", "LOC_CUSTOM, TYPE_ENUM16, offsetof(struct gl_array_attributes, Type), NO_EXTRA" ],
-  [ "TEXTURE_COORD_ARRAY_STRIDE", "LOC_CUSTOM, TYPE_INT, offsetof(struct gl_array_attributes, Stride), NO_EXTRA" ],
+  [ "TEXTURE_COORD_ARRAY_STRIDE", "LOC_CUSTOM, TYPE_SHORT, offsetof(struct gl_array_attributes, Stride), NO_EXTRA" ],
 
 # GL_ARB_multitexture
   [ "MAX_TEXTURE_UNITS", "CONTEXT_INT(Const.MaxTextureUnits), NO_EXTRA" ],
@@ -255,7 +255,7 @@ descriptor=[
 # OES_point_size_array
   [ "POINT_SIZE_ARRAY_OES", "ARRAY_FIELD(VertexAttrib[VERT_ATTRIB_POINT_SIZE].Enabled, TYPE_BOOLEAN), NO_EXTRA" ],
   [ "POINT_SIZE_ARRAY_TYPE_OES", "ARRAY_FIELD(VertexAttrib[VERT_ATTRIB_POINT_SIZE].Type, TYPE_ENUM16), NO_EXTRA" ],
-  [ "POINT_SIZE_ARRAY_STRIDE_OES", "ARRAY_FIELD(VertexAttrib[VERT_ATTRIB_POINT_SIZE].Stride, TYPE_INT), NO_EXTRA" ],
+  [ "POINT_SIZE_ARRAY_STRIDE_OES", "ARRAY_FIELD(VertexAttrib[VERT_ATTRIB_POINT_SIZE].Stride, TYPE_SHORT), NO_EXTRA" ],
   [ "POINT_SIZE_ARRAY_BUFFER_BINDING_OES", "LOC_CUSTOM, TYPE_INT, 0, NO_EXTRA" ],
 ]},
 
@@ -777,11 +777,11 @@ descriptor=[
   [ "COLOR_ARRAY_COUNT_EXT", "CONST(0), NO_EXTRA" ],
   [ "INDEX_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_COLOR_INDEX].Enabled), NO_EXTRA" ],
   [ "INDEX_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_COLOR_INDEX].Type), NO_EXTRA" ],
-  [ "INDEX_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_COLOR_INDEX].Stride), NO_EXTRA" ],
+  [ "INDEX_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_COLOR_INDEX].Stride), NO_EXTRA" ],
   [ "INDEX_ARRAY_COUNT_EXT", "CONST(0), NO_EXTRA" ],
   [ "TEXTURE_COORD_ARRAY_COUNT_EXT", "CONST(0), NO_EXTRA" ],
   [ "EDGE_FLAG_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_EDGEFLAG].Enabled), NO_EXTRA" ],
-  [ "EDGE_FLAG_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_EDGEFLAG].Stride), NO_EXTRA" ],
+  [ "EDGE_FLAG_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_EDGEFLAG].Stride), NO_EXTRA" ],
   [ "EDGE_FLAG_ARRAY_COUNT_EXT", "CONST(0), NO_EXTRA" ],
 
 # GL_ARB_texture_compression
@@ -811,14 +811,14 @@ descriptor=[
   [ "CURRENT_SECONDARY_COLOR", "CONTEXT_FIELD(Current.Attrib[VERT_ATTRIB_COLOR1][0], TYPE_FLOATN_4), extra_flush_current" ],
   [ "SECONDARY_COLOR_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_COLOR1].Enabled), NO_EXTRA" ],
   [ "SECONDARY_COLOR_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_COLOR1].Type), NO_EXTRA" ],
-  [ "SECONDARY_COLOR_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_COLOR1].Stride), NO_EXTRA" ],
+  [ "SECONDARY_COLOR_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_COLOR1].Stride), NO_EXTRA" ],
   [ "SECONDARY_COLOR_ARRAY_SIZE", "LOC_CUSTOM, TYPE_INT, 0, NO_EXTRA" ],
 
 # GL_EXT_fog_coord
   [ "CURRENT_FOG_COORDINATE", "CONTEXT_FLOAT(Current.Attrib[VERT_ATTRIB_FOG][0]), extra_flush_current" ],
   [ "FOG_COORDINATE_ARRAY", "ARRAY_BOOL(VertexAttrib[VERT_ATTRIB_FOG].Enabled), NO_EXTRA" ],
   [ "FOG_COORDINATE_ARRAY_TYPE", "ARRAY_ENUM16(VertexAttrib[VERT_ATTRIB_FOG].Type), NO_EXTRA" ],
-  [ "FOG_COORDINATE_ARRAY_STRIDE", "ARRAY_INT(VertexAttrib[VERT_ATTRIB_FOG].Stride), NO_EXTRA" ],
+  [ "FOG_COORDINATE_ARRAY_STRIDE", "ARRAY_SHORT(VertexAttrib[VERT_ATTRIB_FOG].Stride), NO_EXTRA" ],
   [ "FOG_COORDINATE_SOURCE", "CONTEXT_ENUM16(Fog.FogCoordinateSource), NO_EXTRA" ],
 
 # GL_NV_fog_distance

src/mesa/meson.build

@@ -732,6 +732,6 @@ endif
 if with_glx == 'xlib'
   subdir('drivers/x11')
 endif
-if with_tests
+if with_tests and dri_drivers != []
   subdir('main/tests')
 endif

src/meson.build

@@ -82,7 +82,9 @@ if with_gallium
   subdir('gallium')
   # This has to be here since it requires libgallium, and subdir cannot
   # contain ..
-  subdir('mesa/state_tracker/tests')
+  if with_tests
+    subdir('mesa/state_tracker/tests')
+  endif
 endif
 
 # This must be after at least mesa, glx, and gallium, since libgl will be

src/util/u_queue.c

@@ -311,6 +311,7 @@ util_queue_init(struct util_queue *queue,
       goto fail;
 
    (void) mtx_init(&queue->lock, mtx_plain);
+   (void) mtx_init(&queue->finish_lock, mtx_plain);
 
    queue->num_queued = 0;
    cnd_init(&queue->has_queued_cond);
@@ -398,6 +399,7 @@ util_queue_destroy(struct util_queue *queue)
 
    cnd_destroy(&queue->has_space_cond);
    cnd_destroy(&queue->has_queued_cond);
+   mtx_destroy(&queue->finish_lock);
    mtx_destroy(&queue->lock);
    free(queue->jobs);
    free(queue->threads);
@@ -529,6 +531,12 @@ util_queue_finish(struct util_queue *queue)
 
    util_barrier_init(&barrier, queue->num_threads);
 
+   /* If 2 threads were adding jobs for 2 different barries at the same time,
+    * a deadlock would happen, because 1 barrier requires that all threads
+    * wait for it exclusively.
+    */
+   mtx_lock(&queue->finish_lock);
+
    for (unsigned i = 0; i < queue->num_threads; ++i) {
       util_queue_fence_init(&fences[i]);
       util_queue_add_job(queue, &barrier, &fences[i], util_queue_finish_execute, NULL);
@@ -538,6 +546,7 @@ util_queue_finish(struct util_queue *queue)
       util_queue_fence_wait(&fences[i]);
       util_queue_fence_destroy(&fences[i]);
    }
+   mtx_unlock(&queue->finish_lock);
 
    util_barrier_destroy(&barrier);
 

src/util/u_queue.h

@@ -200,6 +200,7 @@ struct util_queue_job {
 /* Put this into your context. */
 struct util_queue {
    const char *name;
+   mtx_t finish_lock; /* only for util_queue_finish */
    mtx_t lock;
    cnd_t has_queued_cond;
    cnd_t has_space_cond;

src/vulkan/wsi/wsi_common.c

@@ -442,6 +442,7 @@ fail:
 VkResult
 wsi_create_prime_image(const struct wsi_swapchain *chain,
                        const VkSwapchainCreateInfoKHR *pCreateInfo,
+                       bool use_modifier,
                        struct wsi_image *image)
 {
    const struct wsi_device *wsi = chain->wsi;
@@ -626,7 +627,7 @@ wsi_create_prime_image(const struct wsi_swapchain *chain,
    if (result != VK_SUCCESS)
       goto fail;
 
-   image->drm_modifier = DRM_FORMAT_MOD_LINEAR;
+   image->drm_modifier = use_modifier ? DRM_FORMAT_MOD_LINEAR : DRM_FORMAT_MOD_INVALID;
    image->num_planes = 1;
    image->sizes[0] = linear_size;
    image->row_pitches[0] = linear_stride;

src/vulkan/wsi/wsi_common_private.h

@@ -89,6 +89,7 @@ wsi_create_native_image(const struct wsi_swapchain *chain,
 VkResult
 wsi_create_prime_image(const struct wsi_swapchain *chain,
                        const VkSwapchainCreateInfoKHR *pCreateInfo,
+                       bool use_modifier,
                        struct wsi_image *image);
 
 void

src/vulkan/wsi/wsi_common_x11.c

@@ -1043,7 +1043,8 @@ x11_image_init(VkDevice device_h, struct x11_swapchain *chain,
    uint32_t bpp = 32;
 
    if (chain->base.use_prime_blit) {
-      result = wsi_create_prime_image(&chain->base, pCreateInfo, &image->base);
+      bool use_modifier = num_tranches > 0;
+      result = wsi_create_prime_image(&chain->base, pCreateInfo, use_modifier, &image->base);
    } else {
       result = wsi_create_native_image(&chain->base, pCreateInfo,
                                        num_tranches, num_modifiers, modifiers,