docs: add release notes for 11.1.3

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

Android.common.mk

@@ -37,6 +37,7 @@ LOCAL_C_INCLUDES += \
 MESA_VERSION := $(shell cat $(MESA_TOP)/VERSION)
 # define ANDROID_VERSION (e.g., 4.0.x => 0x0400)
 LOCAL_CFLAGS += \
+	-Wno-date-time \
 	-DPACKAGE_VERSION=\"$(MESA_VERSION)\" \
 	-DPACKAGE_BUGREPORT=\"https://bugs.freedesktop.org/enter_bug.cgi?product=Mesa\" \
 	-DANDROID_VERSION=0x0$(MESA_ANDROID_MAJOR_VERSION)0$(MESA_ANDROID_MINOR_VERSION)
@@ -88,3 +89,6 @@ LOCAL_CPPFLAGS += \
 ifeq ($(strip $(LOCAL_MODULE_TAGS)),)
 LOCAL_MODULE_TAGS := optional
 endif
+
+# Quiet down the build system and remove any .h files from the sources
+LOCAL_SRC_FILES := $(patsubst %.h, , $(LOCAL_SRC_FILES))

Android.mk

@@ -42,6 +42,10 @@ $(call local-intermediates-dir)
 endef
 endif
 
+MESA_DRI_MODULE_REL_PATH := dri
+MESA_DRI_MODULE_PATH := $(TARGET_OUT_SHARED_LIBRARIES)/$(MESA_DRI_MODULE_REL_PATH)
+MESA_DRI_MODULE_UNSTRIPPED_PATH := $(TARGET_OUT_SHARED_LIBRARIES_UNSTRIPPED)/$(MESA_DRI_MODULE_REL_PATH)
+
 MESA_COMMON_MK := $(MESA_TOP)/Android.common.mk
 MESA_PYTHON2 := python
 

Makefile.am

@@ -24,18 +24,21 @@ SUBDIRS = src
 AM_DISTCHECK_CONFIGURE_FLAGS = \
 	--enable-dri3 \
 	--enable-gallium-tests \
+	--enable-gallium-osmesa \
 	--enable-gbm \
 	--enable-gles1 \
 	--enable-gles2 \
 	--enable-glx-tls \
+	--enable-nine \
+	--enable-opencl \
 	--enable-va \
 	--enable-vdpau \
 	--enable-xa \
 	--enable-xvmc \
 	--disable-llvm-shared-libs \
-	--with-egl-platforms=x11,wayland,drm \
+	--with-egl-platforms=x11,wayland,drm,surfaceless \
 	--with-dri-drivers=i915,i965,nouveau,radeon,r200,swrast \
-	--with-gallium-drivers=i915,ilo,nouveau,r300,r600,radeonsi,freedreno,svga,swrast
+	--with-gallium-drivers=i915,ilo,nouveau,r300,r600,radeonsi,freedreno,svga,swrast,vc4,virgl
 
 ACLOCAL_AMFLAGS = -I m4
 

VERSION

@@ -1 +1 @@
-11.1.2
+11.1.3

bin/.cherry-ignore

@@ -7,3 +7,6 @@
 # .num_slices isn't available in 11.1
 f2c891353609b48459f27f205407d42823dd7d03 Add missing platform information for KBL
 
+# As requested by Ilia at https://lists.freedesktop.org/archives/mesa-stable/2016-April/003939.html
+6eeb284e4f74a2fe5ae6cba90f97f219935e24df nv50/ir: normalize cube coordinates after derivatives have been computed
+

bin/get-extra-pick-list.sh

@@ -0,0 +1,35 @@
+#!/bin/sh
+
+# Script for generating a list of candidates which fix commits that have been
+# previously cherry-picked to a stable branch.
+#
+# Usage examples:
+#
+# $ bin/get-extra-pick-list.sh
+# $ bin/get-extra-pick-list.sh > picklist
+# $ bin/get-extra-pick-list.sh | tee picklist
+
+# Use the last branchpoint as our limit for the search
+# XXX: there should be a better way for this
+latest_branchpoint=`git branch | grep \* | cut -c 3-`-branchpoint
+
+# Grep for commits with "cherry picked from commit" in the commit message.
+git log --reverse --grep="cherry picked from commit" $latest_branchpoint..HEAD |\
+	grep "cherry picked from commit" |\
+	sed -e 's/^[[:space:]]*(cherry picked from commit[[:space:]]*//' -e 's/)//' |\
+	cut -c -8 |\
+while read sha
+do
+	# Check if the original commit is referenced in master
+	git log -n1 --pretty=oneline --grep=$sha $latest_branchpoint..origin/master |\
+		cut -c -8 |\
+	while read candidate
+	do
+		# Check if the potential fix, hasn't landed in branch yet.
+		found=`git log -n1 --pretty=oneline --reverse --grep=$candidate $latest_branchpoint..HEAD |wc -l`
+		if test $found = 0
+		then
+			echo Commit $candidate might need to be picked, as it references $sha
+		fi
+	done
+done

bin/get-pick-list.sh

@@ -14,7 +14,7 @@ git log --reverse --grep="cherry picked from commit" origin/master..HEAD |\
 	sed -e 's/^[[:space:]]*(cherry picked from commit[[:space:]]*//' -e 's/)//' > already_picked
 
 # Grep for commits that were marked as a candidate for the stable tree.
-git log --reverse --pretty=%H -i --grep='^\([[:space:]]*NOTE: .*[Cc]andidate\|CC:.*mesa-stable\)' HEAD..origin/master |\
+git log --reverse --pretty=%H -i --grep='^\([[:space:]]*NOTE: .*[Cc]andidate\|CC:.*11\.1.*mesa-stable\)' HEAD..origin/master |\
 while read sha
 do
 	# Check to see whether the patch is on the ignore list.

docs/relnotes/11.1.2.html

@@ -31,7 +31,8 @@ because compatibility contexts are not supported.
 
 <h2>SHA256 checksums</h2>
 <pre>
-TBD
+ba0e7462b2936b86e6684c26fbb55519f8d9ad31d13a1c1e1afbe41e73466eea  mesa-11.1.2.tar.gz
+8f72aead896b340ba0f7a4a474bfaf71681f5d675592aec1cb7ba698e319148b  mesa-11.1.2.tar.xz
 </pre>
 
 

docs/relnotes/11.1.3.html

@@ -0,0 +1,318 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
+<html lang="en">
+<head>
+  <meta http-equiv="content-type" content="text/html; charset=utf-8">
+  <title>Mesa Release Notes</title>
+  <link rel="stylesheet" type="text/css" href="../mesa.css">
+</head>
+<body>
+
+<div class="header">
+  <h1>The Mesa 3D Graphics Library</h1>
+</div>
+
+<iframe src="../contents.html"></iframe>
+<div class="content">
+
+<h1>Mesa 11.1.3 Release Notes / April 17, 2016</h1>
+
+<p>
+Mesa 11.1.3 is a bug fix release which fixes bugs found since the 11.1.2 release.
+</p>
+<p>
+Mesa 11.1.3 implements the OpenGL 4.1 API, but the version reported by
+glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) /
+glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being used.
+Some drivers don't support all the features required in OpenGL 4.1.  OpenGL
+4.1 is <strong>only</strong> available if requested at context creation
+because compatibility contexts are not supported.
+</p>
+
+
+<h2>SHA256 checksums</h2>
+<pre>
+TBD
+</pre>
+
+
+<h2>New features</h2>
+<p>None</p>
+
+<h2>Bug fixes</h2>
+
+<p>This list is likely incomplete.</p>
+
+<ul>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=27512">Bug 27512</a> - Illegal instruction _mesa_x86_64_transform_points4_general</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=91526">Bug 91526</a> - World of Warcraft (on Wine) has UI corruption with nouveau</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=92193">Bug 92193</a> - [SKL] ES2-CTS.gtf.GL2ExtensionTests.compressed_astc_texture.compressed_astc_texture fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=93358">Bug 93358</a> - [HSW] Unreal Elemental demo - assertion error in copy_image_with_blitter</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=93418">Bug 93418</a> - Geometry Shaders output wrong vertices on Sandy Bridge</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=93524">Bug 93524</a> - Clover doesn't build</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=93667">Bug 93667</a> - Crash in eglCreateImageKHR with huge texture size</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=93813">Bug 93813</a> - Incorrect viewport range when GL_CLIP_ORIGIN is GL_UPPER_LEFT</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94050">Bug 94050</a> - test_vec4_register_coalesce regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94073">Bug 94073</a> - Miscompilation of abs_vec3_vert_xvary_ref.vert in WebGL conformance</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94088">Bug 94088</a> - [llvmpipe] SIGFPE pthread_barrier_destroy.c:40</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94193">Bug 94193</a> - [llvmpipe] Line antialiasing looks different when GL_LINE_STIPPLE is enabled with pattern 0xffff</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94195">Bug 94195</a> - [llvmpipe] Does not build with LLVM 3.7.x on Windows</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94388">Bug 94388</a> - r600_blit.c:281: r600_decompress_depth_textures: Assertion `tex-&gt;is_depth &amp;&amp; !tex-&gt;is_flushing_texture' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94412">Bug 94412</a> - Trine 3 misrender</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94481">Bug 94481</a> - softpipe - access violation in img_filter_2d_nearest</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94595">Bug 94595</a> - [Mesa AMD&amp;swrast] Texture views attached as framebuffers return their viewed tecture's color encoding and render incorrectly</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=94954">Bug 94954</a> - test_vec4_copy_propagation fails in `make check`</li>
+
+</ul>
+
+
+<h2>Changes</h2>
+
+<p>Anuj Phogat (1):</p>
+<ul>
+  <li>i965: Fix assert conditions for src/dst x/y offsets</li>
+</ul>
+
+<p>Ben Widawsky (2):</p>
+<ul>
+  <li>i965: Make sure we blit a full compressed block</li>
+  <li>i965/skl: Add two missing device IDs</li>
+</ul>
+
+<p>Brian Paul (1):</p>
+<ul>
+  <li>mesa: fix incorrect viewport position when GL_CLIP_ORIGIN = GL_LOWER_LEFT</li>
+</ul>
+
+<p>Chris Forbes (1):</p>
+<ul>
+  <li>i965/blorp: Fix hiz ops on MSAA surfaces</li>
+</ul>
+
+<p>Christian König (1):</p>
+<ul>
+  <li>radeon/uvd: disable MPEG1</li>
+</ul>
+
+<p>Christian Schmidbauer (1):</p>
+<ul>
+  <li>st/nine: specify WINAPI only for i386 and amd64</li>
+</ul>
+
+<p>Daniel Czarnowski (3):</p>
+<ul>
+  <li>egl_dri2: NULL check for xcb_dri2_get_buffers_reply()</li>
+  <li>egl_dri2: set correct error code if swapbuffers fails</li>
+  <li>egl: support EGL_LARGEST_PBUFFER in eglCreatePbufferSurface(...)</li>
+</ul>
+
+<p>Dave Airlie (1):</p>
+<ul>
+  <li>mesa/fbobject: propogate Layered when reusing attachments.</li>
+</ul>
+
+<p>Derek Foreman (1):</p>
+<ul>
+  <li>egl/wayland: Try to use wl_surface.damage_buffer for SwapBuffersWithDamage</li>
+</ul>
+
+<p>Dongwon Kim (1):</p>
+<ul>
+  <li>egl: move Null check to eglGetSyncAttribKHR to prevent Segfault</li>
+</ul>
+
+<p>Emil Velikov (10):</p>
+<ul>
+  <li>docs: add sha256 checksums for 11.1.2</li>
+  <li>get-pick-list.sh: Require explicit "11.1" for nominating stable patches</li>
+  <li>cherry-ignore: do not pick nv50/ir commit</li>
+  <li>automake: add nine to make distcheck</li>
+  <li>install-gallium-links: port changes from install-lib-links</li>
+  <li>automake: add more missing options for make distcheck</li>
+  <li>mesa; add get-extra-pick-list.sh script into bin/</li>
+  <li>egl/x11: check the return value of xcb_dri2_get_buffers_reply()</li>
+  <li>nvc/ir: remove duplicate variable declaration</li>
+  <li>Update version to 11.1.3</li>
+</ul>
+
+<p>Francisco Jerez (4):</p>
+<ul>
+  <li>i965: Reupload push and pull constants when we get new shader image unit state.</li>
+  <li>i965/fs: Add missing analysis invalidation in opt_sampler_eot().</li>
+  <li>i965/fs: Add missing analysis invalidation in fixup_3src_null_dest().</li>
+  <li>i965/vec4: Consider removal of no-op MOVs as progress during register coalesce.</li>
+</ul>
+
+<p>Ilia Mirkin (21):</p>
+<ul>
+  <li>nvc0/ir: fix converting between predicate and gpr</li>
+  <li>nvc0: add some missing PUSH_SPACE's</li>
+  <li>nvc0: avoid negatives in PUSH_SPACE argument</li>
+  <li>glsl: make sure builtins are initialized before getting the shader</li>
+  <li>glsl: return cloned signature, not the builtin one</li>
+  <li>nv50/ir: fix quadop emission in the presence of predication</li>
+  <li>st/mesa: fix up result_src.type when doing i2u/u2i conversions</li>
+  <li>meta/copy_image: use precomputed dst_internal_format to avoid segfault</li>
+  <li>st/mesa: force depth mode to GL_RED for sized depth/stencil formats</li>
+  <li>glx: update to updated version of EXT_create_context_es2_profile</li>
+  <li>nv50,nvc0: bump minimum texture buffer offset alignment</li>
+  <li>nvc0: reset TFB bufctx when we no longer hold a reference to the buffers</li>
+  <li>glsl: avoid stack smashing when there are too many attributes</li>
+  <li>nvc0: fix blit triangle size to fully cover FB's &gt; 8192x8192</li>
+  <li>nv50: reset TFB bufctx when we no longer hold a reference to the buffers</li>
+  <li>nv50/ir: force-enable derivatives on TXD ops</li>
+  <li>st/mesa: only minify depth for 3d targets</li>
+  <li>nv50/ir: fix indirect texturing for non-array textures on nvc0</li>
+  <li>nvc0/ir: fix picking of coordinates from tex instruction for textureGrad</li>
+  <li>nvc0: disable primitive restart and index bias during blits</li>
+  <li>nv50/ir: we can't load local memory directly into an output</li>
+</ul>
+
+<p>Jason Ekstrand (1):</p>
+<ul>
+  <li>nir/lower_vec_to_movs: Better report channels handled by insert_mov</li>
+</ul>
+
+<p>Kenneth Graunke (3):</p>
+<ul>
+  <li>mesa: Make glGet queries initialize ctx-&gt;Debug when necessary.</li>
+  <li>mesa: Allow Get*() of several forgotten IsEnabled() pnames.</li>
+  <li>i965: Only magnify depth for 3D textures, not array textures.</li>
+</ul>
+
+<p>Koop Mast (1):</p>
+<ul>
+  <li>st/clover: Add libelf cflags to the build</li>
+</ul>
+
+<p>Marc-André Lureau (1):</p>
+<ul>
+  <li>virtio_gpu: Add virtio 1.0 PCI ID to driver map</li>
+</ul>
+
+<p>Marek Olšák (3):</p>
+<ul>
+  <li>radeonsi: fix Hyper-Z on Stoney</li>
+  <li>gallium/radeon: don't use temporary buffers for persistent mappings</li>
+  <li>radeonsi: fix Hyper-Z hangs on P2 configs</li>
+</ul>
+
+<p>Matt Turner (3):</p>
+<ul>
+  <li>i965/vec4: don't copy ATTR into 3src instructions with complex swizzles</li>
+  <li>i965/fs: Don't CSE negated multiplies with saturation.</li>
+  <li>i965/vec4: Update vec4 unit tests for commit 01dacc83ff.</li>
+</ul>
+
+<p>Nanley Chery (2):</p>
+<ul>
+  <li>mesa/image: Make _mesa_clip_readpixels() work with renderbuffers</li>
+  <li>mesa/readpix: Clip ReadPixels() area to the ReadBuffer's</li>
+</ul>
+
+<p>Nicolai Hähnle (2):</p>
+<ul>
+  <li>r600g: clear compressed_depthtex/colortex_mask when binding buffer texture</li>
+  <li>st/mesa: use the texture view's format for render-to-texture</li>
+</ul>
+
+<p>Nishanth Peethambaran (2):</p>
+<ul>
+  <li>st/omx: Remove trailing spaces</li>
+  <li>st/omx/dec: Correct the timestamping</li>
+</ul>
+
+<p>Oded Gabbay (8):</p>
+<ul>
+  <li>gallium/radeon: Correctly translate colorswaps for big endian</li>
+  <li>llvmpipe: use vpkswss when dst is signed</li>
+  <li>gallium/radeon: return correct values for BE in r600_translate_colorswap</li>
+  <li>gallium/radeon: remove separate BE path in r600_translate_colorswap</li>
+  <li>gallium/r600: Don't let h/w do endian swap for colorformat</li>
+  <li>gallium/radeon: disable evergreen_do_fast_color_clear for BE</li>
+  <li>r600g: Do colorformat endian swap for PIPE_USAGE_STAGING</li>
+  <li>radeonsi: Do colorformat endian swap for PIPE_USAGE_STAGING</li>
+</ul>
+
+<p>Olivier Pena (1):</p>
+<ul>
+  <li>scons: support for LLVM 3.7.</li>
+</ul>
+
+<p>Patrick Baggett (1):</p>
+<ul>
+  <li>mesa: Use SSE prefetch instructions rather than 3DNow instructions</li>
+</ul>
+
+<p>Rob Herring (10):</p>
+<ul>
+  <li>Android: remove dependence on .SECONDEXPANSION</li>
+  <li>Android: glsl: fix dependence on YACC_HEADER_SUFFIX from build system</li>
+  <li>Android: add -Wno-date-time flag for clang</li>
+  <li>Android: remove headers from LOCAL_SRC_FILES</li>
+  <li>Android: clean-up and fix DRI module path handling</li>
+  <li>freedreno: drop unnecessary -Wno-packed-bitfield-compat</li>
+  <li>gallium/radeon: Add space between string literal and identifier</li>
+  <li>r600: Make enum alu_op_flags unsigned</li>
+  <li>virtio_gpu: Add PCI ID to driver map</li>
+  <li>Android: fix x86 gallium builds</li>
+</ul>
+
+<p>Roland Scheidegger (2):</p>
+<ul>
+  <li>softpipe: fix anisotropic filtering crash</li>
+  <li>draw: fix line stippling</li>
+</ul>
+
+<p>Samuel Pitoiset (1):</p>
+<ul>
+  <li>nvc0: make sure to delete samplers used by compute shaders</li>
+</ul>
+
+<p>Steinar H. Gunderson (1):</p>
+<ul>
+  <li>mesa: Fix locking of GLsync objects.</li>
+</ul>
+
+<p>Tamil velan (1):</p>
+<ul>
+  <li>radeon/uvd: increase max height to 4096 for VI and newer</li>
+</ul>
+
+<p>Thomas Hellstrom (2):</p>
+<ul>
+  <li>winsys/svga: Fix an uninitialized return value</li>
+  <li>winsys/svga: Increase the fence timeout</li>
+</ul>
+
+<p>Vinson Lee (1):</p>
+<ul>
+  <li>llvmpipe: Do not use barriers if not using threads.</li>
+</ul>
+
+<p>xavier (1):</p>
+<ul>
+  <li>r600/sb: Do not distribute neg in expr_handler::fold_assoc() when folding multiplications.</li>
+</ul>
+
+
+</div>
+</body>
+</html>

include/D3D9/d3d9types.h

@@ -178,11 +178,17 @@ typedef struct _RGNDATA {
 #undef WINAPI
 #endif /* WINAPI*/
 
-#if defined(__x86_64__) || defined(_M_X64)
-#define WINAPI __attribute__((ms_abi))
-#else /* x86_64 */
-#define WINAPI __attribute__((__stdcall__))
-#endif /* x86_64 */
+#ifdef __GNUC__
+  #if (defined(__x86_64__) && !defined(__ILP32__)) || defined(_M_X64)
+    #define WINAPI __attribute__((ms_abi))
+  #elif defined(__i386) || defined(_M_IX86)
+    #define WINAPI __attribute__((__stdcall__))
+  #else /* neither amd64 nor i386 */
+    #define WINAPI
+  #endif
+#else /* __GNUC__ */
+  #define WINAPI
+#endif
 
 /* Implementation caps */
 #define D3DPRESENT_BACK_BUFFERS_MAX    3

include/pci_ids/i965_pci_ids.h

@@ -112,6 +112,7 @@ CHIPSET(0x162E, bdw_gt3, "Intel(R) Broadwell GT3")
 CHIPSET(0x1902, skl_gt1, "Intel(R) HD Graphics 510 (Skylake GT1)")
 CHIPSET(0x1906, skl_gt1, "Intel(R) HD Graphics 510 (Skylake GT1)")
 CHIPSET(0x190A, skl_gt1, "Intel(R) Skylake GT1")
+CHIPSET(0x190B, skl_gt1, "Intel(R) HD Graphics 510 (Skylake GT1)")
 CHIPSET(0x190E, skl_gt1, "Intel(R) Skylake GT1")
 CHIPSET(0x1912, skl_gt2, "Intel(R) HD Graphics 530 (Skylake GT2)")
 CHIPSET(0x1913, skl_gt2, "Intel(R) Skylake GT2f")
@@ -128,6 +129,7 @@ CHIPSET(0x1926, skl_gt3, "Intel(R) HD Graphics 535 (Skylake GT3)")
 CHIPSET(0x1927, skl_gt3, "Intel(R) Iris Graphics 550 (Skylake GT3e)")
 CHIPSET(0x192A, skl_gt4, "Intel(R) Skylake GT4")
 CHIPSET(0x192B, skl_gt3, "Intel(R) Iris Graphics (Skylake GT3fe)")
+CHIPSET(0x192D, skl_gt3, "Intel(R) Skylake GT3")
 CHIPSET(0x1932, skl_gt4, "Intel(R) Skylake GT4")
 CHIPSET(0x193A, skl_gt4, "Intel(R) Skylake GT4")
 CHIPSET(0x193B, skl_gt4, "Intel(R) Skylake GT4")

include/pci_ids/virtio_gpu_pci_ids.h

@@ -0,0 +1,2 @@
+CHIPSET(0x0010, VIRTGL, VIRTGL)
+CHIPSET(0x1050, VIRTGL, VIRTGL)

install-gallium-links.mk

@@ -3,9 +3,9 @@
 
 if BUILD_SHARED
 if HAVE_COMPAT_SYMLINKS
-all-local : .libs/install-gallium-links
+all-local : .install-gallium-links
 
-.libs/install-gallium-links : $(dri_LTLIBRARIES) $(egl_LTLIBRARIES) $(lib_LTLIBRARIES)
+.install-gallium-links : $(dri_LTLIBRARIES) $(egl_LTLIBRARIES) $(lib_LTLIBRARIES)
 	$(AM_V_GEN)$(MKDIR_P) $(top_builddir)/$(LIB_DIR);	\
 	link_dir=$(top_builddir)/$(LIB_DIR)/gallium;		\
 	if test x$(egl_LTLIBRARIES) != x; then			\
@@ -23,4 +23,15 @@ all-local : .libs/install-gallium-links
 		fi;						\
 	done && touch $@
 endif
+
+clean-local:
+	for f in $(notdir $(dri_LTLIBRARIES:%.la=.libs/%.$(LIB_EXT)*)) \
+		 $(notdir $(egl_LTLIBRARIES:%.la=.libs/%.$(LIB_EXT)*)) \
+		 $(notdir $(lib_LTLIBRARIES:%.la=.libs/%.$(LIB_EXT)*)); do \
+		echo $$f; \
+		$(RM) $(top_builddir)/$(LIB_DIR)/gallium/$$f;   \
+	done;
+	rmdir $(top_builddir)/$(LIB_DIR)/gallium || true
+	$(RM) .install-gallium-links
+
 endif

scons/llvm.py

@@ -106,7 +106,19 @@ def generate(env):
         ])
         env.Prepend(LIBPATH = [os.path.join(llvm_dir, 'lib')])
         # LIBS should match the output of `llvm-config --libs engine mcjit bitwriter x86asmprinter`
-        if llvm_version >= distutils.version.LooseVersion('3.6'):
+        if llvm_version >= distutils.version.LooseVersion('3.7'):
+            env.Prepend(LIBS = [
+                'LLVMBitWriter', 'LLVMX86Disassembler', 'LLVMX86AsmParser',
+                'LLVMX86CodeGen', 'LLVMSelectionDAG', 'LLVMAsmPrinter',
+                'LLVMCodeGen', 'LLVMScalarOpts', 'LLVMProfileData',
+                'LLVMInstCombine', 'LLVMInstrumentation', 'LLVMTransformUtils', 'LLVMipa',
+                'LLVMAnalysis', 'LLVMX86Desc', 'LLVMMCDisassembler',
+                'LLVMX86Info', 'LLVMX86AsmPrinter', 'LLVMX86Utils',
+                'LLVMMCJIT', 'LLVMTarget', 'LLVMExecutionEngine',
+                'LLVMRuntimeDyld', 'LLVMObject', 'LLVMMCParser',
+                'LLVMBitReader', 'LLVMMC', 'LLVMCore', 'LLVMSupport'
+            ])
+        elif llvm_version >= distutils.version.LooseVersion('3.6'):
             env.Prepend(LIBS = [
                 'LLVMBitWriter', 'LLVMX86Disassembler', 'LLVMX86AsmParser',
                 'LLVMX86CodeGen', 'LLVMSelectionDAG', 'LLVMAsmPrinter',

src/egl/Android.mk

@@ -44,11 +44,10 @@ LOCAL_CFLAGS := \
 	-DHAVE_ANDROID_PLATFORM
 
 ifeq ($(MESA_LOLLIPOP_BUILD),true)
-LOCAL_CFLAGS_arm := -DDEFAULT_DRIVER_DIR=\"/system/lib/dri\"
-LOCAL_CFLAGS_x86 := -DDEFAULT_DRIVER_DIR=\"/system/lib/dri\"
-LOCAL_CFLAGS_x86_64 := -DDEFAULT_DRIVER_DIR=\"/system/lib64/dri\"
+LOCAL_CFLAGS_32 := -DDEFAULT_DRIVER_DIR=\"/system/lib/$(MESA_DRI_MODULE_REL_PATH)\"
+LOCAL_CFLAGS_64 := -DDEFAULT_DRIVER_DIR=\"/system/lib64/$(MESA_DRI_MODULE_REL_PATH)\"
 else
-LOCAL_CFLAGS += -DDEFAULT_DRIVER_DIR=\"/system/lib/dri\"
+LOCAL_CFLAGS += -DDEFAULT_DRIVER_DIR=\"/system/lib/$(MESA_DRI_MODULE_REL_PATH)\"
 endif
 
 LOCAL_C_INCLUDES := \

src/egl/drivers/dri2/platform_wayland.c

@@ -653,6 +653,37 @@ create_wl_buffer(struct dri2_egl_surface *dri2_surf)
                           &wl_buffer_listener, dri2_surf);
 }
 
+static EGLBoolean
+try_damage_buffer(struct dri2_egl_surface *dri2_surf,
+                  const EGLint *rects,
+                  EGLint n_rects)
+{
+/* The WL_SURFACE_DAMAGE_BUFFER_SINCE_VERSION macro and
+ * wl_proxy_get_version() were both introduced in wayland 1.10.
+ * Instead of bumping our wayland dependency we just make this
+ * function conditional on the required 1.10 features, falling
+ * back to old (correct but suboptimal) behaviour for older
+ * wayland.
+ */
+#ifdef WL_SURFACE_DAMAGE_BUFFER_SINCE_VERSION
+   int i;
+
+   if (wl_proxy_get_version((struct wl_proxy *) dri2_surf->wl_win->surface)
+       < WL_SURFACE_DAMAGE_BUFFER_SINCE_VERSION)
+      return EGL_FALSE;
+
+   for (i = 0; i < n_rects; i++) {
+      const int *rect = &rects[i * 4];
+
+      wl_surface_damage_buffer(dri2_surf->wl_win->surface,
+                               rect[0],
+                               dri2_surf->base.Height - rect[1] - rect[3],
+                               rect[2], rect[3]);
+   }
+   return EGL_TRUE;
+#endif
+   return EGL_FALSE;
+}
 /**
  * Called via eglSwapBuffers(), drv->API.SwapBuffers().
  */
@@ -703,10 +734,12 @@ dri2_wl_swap_buffers_with_damage(_EGLDriver *drv,
    dri2_surf->dx = 0;
    dri2_surf->dy = 0;
 
-   /* We deliberately ignore the damage region and post maximum damage, due to
+   /* If the compositor doesn't support damage_buffer, we deliberately
+    * ignore the damage region and post maximum damage, due to
     * https://bugs.freedesktop.org/78190 */
-   wl_surface_damage(dri2_surf->wl_win->surface,
-                     0, 0, INT32_MAX, INT32_MAX);
+   if (!n_rects || !try_damage_buffer(dri2_surf, rects, n_rects))
+      wl_surface_damage(dri2_surf->wl_win->surface,
+                        0, 0, INT32_MAX, INT32_MAX);
 
    if (dri2_dpy->is_different_gpu) {
       _EGLContext *ctx = _eglGetCurrentContext();

src/egl/drivers/dri2/platform_x11.c

@@ -472,6 +472,8 @@ dri2_x11_get_buffers(__DRIdrawable * driDrawable,
 					    dri2_surf->drawable,
 					    count, count, attachments);
    reply = xcb_dri2_get_buffers_reply (dri2_dpy->conn, cookie, NULL);
+   if (reply == NULL)
+      return NULL;
    buffers = xcb_dri2_get_buffers_buffers (reply);
    if (buffers == NULL)
       return NULL;
@@ -870,7 +872,12 @@ dri2_x11_swap_buffers(_EGLDriver *drv, _EGLDisplay *disp, _EGLSurface *draw)
    struct dri2_egl_surface *dri2_surf = dri2_egl_surface(draw);
 
    if (dri2_dpy->dri2) {
-      return dri2_x11_swap_buffers_msc(drv, disp, draw, 0, 0, 0) != -1;
+      if (dri2_x11_swap_buffers_msc(drv, disp, draw, 0, 0, 0) != -1) {
+          return EGL_TRUE;
+      }
+      /* Swap failed with a window drawable. */
+      _eglError(EGL_BAD_NATIVE_WINDOW, __FUNCTION__);
+      return EGL_FALSE;
    } else {
       assert(dri2_dpy->swrast);
 
@@ -999,6 +1006,9 @@ dri2_create_image_khr_pixmap(_EGLDisplay *disp, _EGLContext *ctx,
    geometry_cookie = xcb_get_geometry (dri2_dpy->conn, drawable);
    buffers_reply = xcb_dri2_get_buffers_reply (dri2_dpy->conn,
 					       buffers_cookie, NULL);
+   if (buffers_reply == NULL)
+     return NULL;
+
    buffers = xcb_dri2_get_buffers_buffers (buffers_reply);
    if (buffers == NULL) {
       return NULL;

src/egl/main/eglapi.c

@@ -1555,8 +1555,14 @@ eglGetSyncAttrib(EGLDisplay dpy, EGLSync sync, EGLint attribute, EGLAttrib *valu
 static EGLBoolean EGLAPIENTRY
 eglGetSyncAttribKHR(EGLDisplay dpy, EGLSync sync, EGLint attribute, EGLint *value)
 {
-   EGLAttrib attrib = *value;
-   EGLBoolean result = eglGetSyncAttrib(dpy, sync, attribute, &attrib);
+   EGLAttrib attrib;
+   EGLBoolean result;
+
+   if (!value)
+      RETURN_EGL_ERROR(NULL, EGL_BAD_PARAMETER, EGL_FALSE);
+
+   attrib = *value;
+   result = eglGetSyncAttrib(dpy, sync, attribute, &attrib);
 
    /* The EGL_KHR_fence_sync spec says this about eglGetSyncAttribKHR:
     *

src/egl/main/eglconfig.c

@@ -44,7 +44,6 @@
 #include "egllog.h"
 
 
-#define MIN2(A, B)  (((A) < (B)) ? (A) : (B))
 
 
 /**

src/egl/main/egldefines.h

@@ -40,9 +40,16 @@ extern "C" {
 
 #define _EGL_MAX_EXTENSIONS_LEN 1000
 
+/* Hardcoded, conservative default for EGL_LARGEST_PBUFFER,
+ * this is used to implement EGL_LARGEST_PBUFFER.
+ */
+#define _EGL_MAX_PBUFFER_WIDTH 4096
+#define _EGL_MAX_PBUFFER_HEIGHT 4096
+
 #define _EGL_VENDOR_STRING "Mesa Project"
 
 #define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0]))
+#define MIN2(A, B)  (((A) < (B)) ? (A) : (B))
 
 #ifdef __cplusplus
 }

src/egl/main/eglsurface.c

@@ -307,6 +307,12 @@ _eglInitSurface(_EGLSurface *surf, _EGLDisplay *dpy, EGLint type,
    if (err != EGL_SUCCESS)
       return _eglError(err, func);
 
+   /* if EGL_LARGEST_PBUFFER in use, clamp width and height */
+   if (surf->LargestPbuffer) {
+      surf->Width = MIN2(surf->Width, _EGL_MAX_PBUFFER_WIDTH);
+      surf->Height = MIN2(surf->Height, _EGL_MAX_PBUFFER_HEIGHT);
+   }
+
    return EGL_TRUE;
 }
 

src/egl/main/eglsync.c

@@ -144,9 +144,6 @@ EGLBoolean
 _eglGetSyncAttrib(_EGLDriver *drv, _EGLDisplay *dpy, _EGLSync *sync,
                   EGLint attribute, EGLAttrib *value)
 {
-   if (!value)
-      return _eglError(EGL_BAD_PARAMETER, "eglGetSyncAttribKHR");
-
    switch (attribute) {
    case EGL_SYNC_TYPE_KHR:
       *value = sync->Type;

src/gallium/auxiliary/draw/draw_pipe_stipple.c

@@ -108,11 +108,11 @@ emit_segment(struct draw_stage *stage, struct prim_header *header,
 }
 
 
-static inline unsigned
+static inline bool
 stipple_test(int counter, ushort pattern, int factor)
 {
    int b = (counter / factor) & 0xf;
-   return (1 << b) & pattern;
+   return !!((1 << b) & pattern);
 }
 
 
@@ -126,7 +126,7 @@ stipple_line(struct draw_stage *stage, struct prim_header *header)
    const float *pos0 = v0->data[pos];
    const float *pos1 = v1->data[pos];
    float start = 0;
-   int state = 0;
+   bool state = 0;
 
    float x0 = pos0[0];
    float x1 = pos1[0];
@@ -143,29 +143,29 @@ stipple_line(struct draw_stage *stage, struct prim_header *header)
       stipple->counter = 0;
 
 
-   /* XXX ToDo: intead of iterating pixel-by-pixel, use a look-up table.
+   /* XXX ToDo: instead of iterating pixel-by-pixel, use a look-up table.
     */
    for (i = 0; i < length; i++) {
-      int result = stipple_test( (int) stipple->counter+i,
-                                 (ushort) stipple->pattern, stipple->factor );
+      bool result = stipple_test((int)stipple->counter + i,
+                                 (ushort)stipple->pattern, stipple->factor);
       if (result != state) {
          /* changing from "off" to "on" or vice versa */
-	 if (state) {
-	    if (start != i) {
+         if (state) {
+            if (start != i) {
                /* finishing an "on" segment */
-	       emit_segment( stage, header, start / length, i / length );
+               emit_segment(stage, header, start / length, i / length);
             }
-	 }
-	 else {
+         }
+         else {
             /* starting an "on" segment */
-	    start = (float) i;
-	 }
-	 state = result;	   
+            start = (float)i;
+         }
+         state = result;
       }
    }
 
    if (state && start < length)
-      emit_segment( stage, header, start / length, 1.0 );
+      emit_segment(stage, header, start / length, 1.0);
 
    stipple->counter += length;
 }

src/gallium/auxiliary/gallivm/lp_bld_pack.c

@@ -461,50 +461,49 @@ lp_build_pack2(struct gallivm_state *gallivm,
    assert(src_type.length * 2 == dst_type.length);
 
    /* Check for special cases first */
-   if((util_cpu_caps.has_sse2 || util_cpu_caps.has_altivec) &&
-       src_type.width * src_type.length >= 128) {
+   if ((util_cpu_caps.has_sse2 || util_cpu_caps.has_altivec) &&
+        src_type.width * src_type.length >= 128) {
       const char *intrinsic = NULL;
       boolean swap_intrinsic_operands = FALSE;
 
       switch(src_type.width) {
       case 32:
          if (util_cpu_caps.has_sse2) {
-           if(dst_type.sign) {
+           if (dst_type.sign) {
               intrinsic = "llvm.x86.sse2.packssdw.128";
-           }
-           else {
+           } else {
               if (util_cpu_caps.has_sse4_1) {
                  intrinsic = "llvm.x86.sse41.packusdw";
               }
            }
          } else if (util_cpu_caps.has_altivec) {
             if (dst_type.sign) {
-              intrinsic = "llvm.ppc.altivec.vpkswus";
-           } else {
-              intrinsic = "llvm.ppc.altivec.vpkuwus";
-           }
+               intrinsic = "llvm.ppc.altivec.vpkswss";
+            } else {
+               intrinsic = "llvm.ppc.altivec.vpkuwus";
+            }
 #ifdef PIPE_ARCH_LITTLE_ENDIAN
-           swap_intrinsic_operands = TRUE;
+            swap_intrinsic_operands = TRUE;
 #endif
          }
          break;
       case 16:
          if (dst_type.sign) {
             if (util_cpu_caps.has_sse2) {
-              intrinsic = "llvm.x86.sse2.packsswb.128";
+               intrinsic = "llvm.x86.sse2.packsswb.128";
             } else if (util_cpu_caps.has_altivec) {
-              intrinsic = "llvm.ppc.altivec.vpkshss";
+               intrinsic = "llvm.ppc.altivec.vpkshss";
 #ifdef PIPE_ARCH_LITTLE_ENDIAN
-              swap_intrinsic_operands = TRUE;
+               swap_intrinsic_operands = TRUE;
 #endif
             }
          } else {
             if (util_cpu_caps.has_sse2) {
-              intrinsic = "llvm.x86.sse2.packuswb.128";
+               intrinsic = "llvm.x86.sse2.packuswb.128";
             } else if (util_cpu_caps.has_altivec) {
-	      intrinsic = "llvm.ppc.altivec.vpkshus";
+               intrinsic = "llvm.ppc.altivec.vpkshus";
 #ifdef PIPE_ARCH_LITTLE_ENDIAN
-              swap_intrinsic_operands = TRUE;
+               swap_intrinsic_operands = TRUE;
 #endif
             }
          }

src/gallium/drivers/freedreno/Android.mk

@@ -32,8 +32,8 @@ LOCAL_SRC_FILES := \
 	$(a4xx_SOURCES) \
 	$(ir3_SOURCES)
 
-LOCAL_CFLAGS := \
-	-Wno-packed-bitfield-compat
+#LOCAL_CFLAGS := \
+#	-Wno-packed-bitfield-compat
 
 LOCAL_C_INCLUDES := \
 	$(LOCAL_PATH)/ir3

src/gallium/drivers/llvmpipe/lp_rast.c

@@ -910,7 +910,9 @@ lp_rast_create( unsigned num_threads )
    create_rast_threads(rast);
 
    /* for synchronizing rasterization threads */
-   pipe_barrier_init( &rast->barrier, rast->num_threads );
+   if (rast->num_threads > 0) {
+      pipe_barrier_init( &rast->barrier, rast->num_threads );
+   }
 
    memset(lp_dummy_tile, 0, sizeof lp_dummy_tile);
 
@@ -967,7 +969,9 @@ void lp_rast_destroy( struct lp_rasterizer *rast )
    }
 
    /* for synchronizing rasterization threads */
-   pipe_barrier_destroy( &rast->barrier );
+   if (rast->num_threads > 0) {
+      pipe_barrier_destroy( &rast->barrier );
+   }
 
    lp_scene_queue_destroy(rast->full_scenes);
 

src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gk110.cpp

@@ -1232,7 +1232,7 @@ CodeEmitterGK110::emitQUADOP(const Instruction *i, uint8_t qOp, uint8_t laneMask
 
    defId(i->def(0), 2);
    srcId(i->src(0), 10);
-   srcId(i->srcExists(1) ? i->src(1) : i->src(0), 23);
+   srcId((i->srcExists(1) && i->predSrc != 1) ? i->src(1) : i->src(0), 23);
 
    if (i->op == OP_QUADOP && progType != Program::TYPE_FRAGMENT)
       code[1] |= 1 << 9; // dall
@@ -1825,10 +1825,16 @@ CodeEmitterGK110::emitInstruction(Instruction *insn)
    case OP_CEIL:
    case OP_FLOOR:
    case OP_TRUNC:
-   case OP_CVT:
    case OP_SAT:
       emitCVT(insn);
       break;
+   case OP_CVT:
+      if (insn->def(0).getFile() == FILE_PREDICATE ||
+          insn->src(0).getFile() == FILE_PREDICATE)
+         emitMOV(insn);
+      else
+         emitCVT(insn);
+      break;
    case OP_RSQ:
       emitSFnOp(insn, 5 + 2 * insn->subOp);
       break;

src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp

@@ -671,7 +671,12 @@ CodeEmitterGM107::emitMOV()
        (insn->sType != TYPE_F32 && !longIMMD(insn->src(0)))) {
       switch (insn->src(0).getFile()) {
       case FILE_GPR:
-         emitInsn(0x5c980000);
+         if (insn->def(0).getFile() == FILE_PREDICATE) {
+            emitInsn(0x5b6a0000);
+            emitGPR (0x08);
+         } else {
+            emitInsn(0x5c980000);
+         }
          emitGPR (0x14, insn->src(0));
          break;
       case FILE_MEMORY_CONST:
@@ -682,18 +687,32 @@ CodeEmitterGM107::emitMOV()
          emitInsn(0x38980000);
          emitIMMD(0x14, 19, insn->src(0));
          break;
+      case FILE_PREDICATE:
+         emitInsn(0x50880000);
+         emitPRED(0x0c, insn->src(0));
+         emitPRED(0x1d);
+         emitPRED(0x27);
+         break;
       default:
          assert(!"bad src file");
          break;
       }
-      emitField(0x27, 4, insn->lanes);
+      if (insn->def(0).getFile() != FILE_PREDICATE &&
+          insn->src(0).getFile() != FILE_PREDICATE)
+         emitField(0x27, 4, insn->lanes);
    } else {
       emitInsn (0x01000000);
       emitIMMD (0x14, 32, insn->src(0));
       emitField(0x0c, 4, insn->lanes);
    }
 
-   emitGPR(0x00, insn->def(0));
+   if (insn->def(0).getFile() == FILE_PREDICATE) {
+      emitPRED(0x27);
+      emitPRED(0x03, insn->def(0));
+      emitPRED(0x00);
+   } else {
+      emitGPR(0x00, insn->def(0));
+   }
 }
 
 void
@@ -1509,7 +1528,10 @@ CodeEmitterGM107::emitFSWZADD()
    emitRND  (0x27);
    emitField(0x26, 1, insn->lanes); /* abused for .ndv */
    emitField(0x1c, 8, insn->subOp);
-   emitGPR  (0x14, insn->src(1));
+   if (insn->predSrc != 1)
+      emitGPR  (0x14, insn->src(1));
+   else
+      emitGPR  (0x14);
    emitGPR  (0x08, insn->src(0));
    emitGPR  (0x00, insn->def(0));
 }
@@ -2631,11 +2653,7 @@ CodeEmitterGM107::emitInstruction(Instruction *i)
       emitRAM();
       break;
    case OP_MOV:
-      if (insn->def(0).getFile() == FILE_GPR &&
-          insn->src(0).getFile() != FILE_PREDICATE)
-         emitMOV();
-      else
-         assert(!"R2P/P2R");
+      emitMOV();
       break;
    case OP_RDSV:
       emitS2R();
@@ -2647,7 +2665,10 @@ CodeEmitterGM107::emitInstruction(Instruction *i)
    case OP_CEIL:
    case OP_TRUNC:
    case OP_CVT:
-      if (isFloatType(insn->dType)) {
+      if (insn->op == OP_CVT && (insn->def(0).getFile() == FILE_PREDICATE ||
+                                 insn->src(0).getFile() == FILE_PREDICATE)) {
+         emitMOV();
+      } else if (isFloatType(insn->dType)) {
          if (isFloatType(insn->sType))
             emitF2F();
          else

src/gallium/drivers/nouveau/codegen/nv50_ir_emit_nv50.cpp

@@ -527,7 +527,8 @@ CodeEmitterNV50::emitForm_ADD(const Instruction *i)
 
    setSrcFileBits(i, NV50_OP_ENC_LONG_ALT);
    setSrc(i, 0, 0);
-   setSrc(i, 1, 2);
+   if (i->predSrc != 1)
+      setSrc(i, 1, 2);
 
    if (i->getIndirect(0, 0)) {
       assert(!i->getIndirect(1, 0));
@@ -839,7 +840,7 @@ CodeEmitterNV50::emitQUADOP(const Instruction *i, uint8_t lane, uint8_t quOp)
 
    emitForm_ADD(i);
 
-   if (!i->srcExists(1))
+   if (!i->srcExists(1) || i->predSrc == 1)
       srcId(i->src(0), 32 + 14);
 }
 
@@ -1608,7 +1609,9 @@ CodeEmitterNV50::emitTEX(const TexInstruction *i)
    code[1] |= (i->tex.mask & 0xc) << 12;
 
    if (i->tex.liveOnly)
-      code[1] |= 4;
+      code[1] |= 1 << 2;
+   if (i->tex.derivAll)
+      code[1] |= 1 << 3;
 
    defId(i->def(0), 2);
 

src/gallium/drivers/nouveau/codegen/nv50_ir_emit_nvc0.cpp

@@ -1331,7 +1331,7 @@ CodeEmitterNVC0::emitQUADOP(const Instruction *i, uint8_t qOp, uint8_t laneMask)
 
    defId(i->def(0), 14);
    srcId(i->src(0), 20);
-   srcId(i->srcExists(1) ? i->src(1) : i->src(0), 26);
+   srcId((i->srcExists(1) && i->predSrc != 1) ? i->src(1) : i->src(0), 26);
 
    if (i->op == OP_QUADOP && progType != Program::TYPE_FRAGMENT)
       code[0] |= 1 << 9; // dall
@@ -2430,10 +2430,16 @@ CodeEmitterNVC0::emitInstruction(Instruction *insn)
    case OP_CEIL:
    case OP_FLOOR:
    case OP_TRUNC:
-   case OP_CVT:
    case OP_SAT:
       emitCVT(insn);
       break;
+   case OP_CVT:
+      if (insn->def(0).getFile() == FILE_PREDICATE ||
+          insn->src(0).getFile() == FILE_PREDICATE)
+         emitMOV(insn);
+      else
+         emitCVT(insn);
+      break;
    case OP_RSQ:
       emitSFnOp(insn, 5 + 2 * insn->subOp);
       break;

src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nv50.cpp

@@ -354,7 +354,8 @@ NV50LegalizeSSA::propagateWriteToOutput(Instruction *st)
       return;
 
    for (int s = 0; di->srcExists(s); ++s)
-      if (di->src(s).getFile() == FILE_IMMEDIATE)
+      if (di->src(s).getFile() == FILE_IMMEDIATE ||
+          di->src(s).getFile() == FILE_MEMORY_LOCAL)
          return;
 
    if (prog->getType() == Program::TYPE_GEOMETRY) {
@@ -916,6 +917,7 @@ NV50LoweringPreSSA::handleTXD(TexInstruction *i)
 
    handleTEX(i);
    i->op = OP_TEX; // no need to clone dPdx/dPdy later
+   i->tex.derivAll = true;
 
    for (c = 0; c < dim; ++c)
       crd[c] = bld.getScratch();

src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp

@@ -722,9 +722,13 @@ NVC0LoweringPass::handleTEX(TexInstruction *i)
       }
 
       Value *arrayIndex = i->tex.target.isArray() ? i->getSrc(lyr) : NULL;
-      for (int s = dim; s >= 1; --s)
-         i->setSrc(s, i->getSrc(s - 1));
-      i->setSrc(0, arrayIndex);
+      if (arrayIndex) {
+         for (int s = dim; s >= 1; --s)
+            i->setSrc(s, i->getSrc(s - 1));
+         i->setSrc(0, arrayIndex);
+      } else {
+         i->moveSources(0, 1);
+      }
 
       if (arrayIndex) {
          int sat = (i->op == OP_TXF) ? 1 : 0;
@@ -846,7 +850,17 @@ NVC0LoweringPass::handleManualTXD(TexInstruction *i)
    Value *zero = bld.loadImm(bld.getSSA(), 0);
    int l, c;
    const int dim = i->tex.target.getDim();
-   const int array = i->tex.target.isArray();
+
+   // This function is invoked after handleTEX lowering, so we have to expect
+   // the arguments in the order that the hw wants them. For Fermi, array and
+   // indirect are both in the leading arg, while for Kepler, array and
+   // indirect are separate (and both precede the coordinates). Maxwell is
+   // handled in a separate function.
+   unsigned array;
+   if (targ->getChipset() < NVISA_GK104_CHIPSET)
+      array = i->tex.target.isArray() || i->tex.rIndirectSrc >= 0;
+   else
+      array = i->tex.target.isArray() + (i->tex.rIndirectSrc >= 0);
 
    i->op = OP_TEX; // no need to clone dPdx/dPdy later
 

src/gallium/drivers/nouveau/nv50/nv50_screen.c

@@ -120,7 +120,7 @@ nv50_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_CONSTANT_BUFFER_OFFSET_ALIGNMENT:
       return 256;
    case PIPE_CAP_TEXTURE_BUFFER_OFFSET_ALIGNMENT:
-      return 1; /* 256 for binding as RT, but that's not possible in GL */
+      return 16; /* 256 for binding as RT, but that's not possible in GL */
    case PIPE_CAP_MIN_MAP_BUFFER_ALIGNMENT:
       return NOUVEAU_MIN_BUFFER_MAP_ALIGN;
    case PIPE_CAP_MAX_VIEWPORTS:

src/gallium/drivers/nouveau/nv50/nv50_shader_state.c

@@ -633,8 +633,6 @@ nv50_stream_output_validate(struct nv50_context *nv50)
    BEGIN_NV04(push, NV50_3D(STRMOUT_BUFFERS_CTRL), 1);
    PUSH_DATA (push, ctrl);
 
-   nouveau_bufctx_reset(nv50->bufctx_3d, NV50_BIND_SO);
-
    for (i = 0; i < nv50->num_so_targets; ++i) {
       struct nv50_so_target *targ = nv50_so_target(nv50->so_target[i]);
       struct nv04_resource *buf = nv04_resource(targ->pipe.buffer);

src/gallium/drivers/nouveau/nv50/nv50_state.c

@@ -1158,8 +1158,10 @@ nv50_set_stream_output_targets(struct pipe_context *pipe,
    }
    nv50->num_so_targets = num_targets;
 
-   if (nv50->so_targets_dirty)
+   if (nv50->so_targets_dirty) {
+      nouveau_bufctx_reset(nv50->bufctx_3d, NV50_BIND_SO);
       nv50->dirty |= NV50_NEW_STRMOUT;
+   }
 }
 
 static void

src/gallium/drivers/nouveau/nvc0/nvc0_screen.c

@@ -109,7 +109,7 @@ nvc0_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_CONSTANT_BUFFER_OFFSET_ALIGNMENT:
       return 256;
    case PIPE_CAP_TEXTURE_BUFFER_OFFSET_ALIGNMENT:
-      return 1; /* 256 for binding as RT, but that's not possible in GL */
+      return 16; /* 256 for binding as RT, but that's not possible in GL */
    case PIPE_CAP_MIN_MAP_BUFFER_ALIGNMENT:
       return NOUVEAU_MIN_BUFFER_MAP_ALIGN;
    case PIPE_CAP_MAX_VIEWPORTS:

src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c

@@ -297,7 +297,6 @@ nvc0_tfb_validate(struct nvc0_context *nvc0)
 
    if (!(nvc0->dirty & NVC0_NEW_TFB_TARGETS))
       return;
-   nouveau_bufctx_reset(nvc0->bufctx_3d, NVC0_BIND_TFB);
 
    for (b = 0; b < nvc0->num_tfbbufs; ++b) {
       struct nvc0_so_target *targ = nvc0_so_target(nvc0->tfbbuf[b]);

src/gallium/drivers/nouveau/nvc0/nvc0_state.c

@@ -413,7 +413,7 @@ nvc0_sampler_state_delete(struct pipe_context *pipe, void *hwcso)
 {
    unsigned s, i;
 
-   for (s = 0; s < 5; ++s)
+   for (s = 0; s < 6; ++s)
       for (i = 0; i < nvc0_context(pipe)->num_samplers[s]; ++i)
          if (nvc0_context(pipe)->samplers[s][i] == hwcso)
             nvc0_context(pipe)->samplers[s][i] = NULL;
@@ -1157,8 +1157,10 @@ nvc0_set_transform_feedback_targets(struct pipe_context *pipe,
    }
    nvc0->num_tfbbufs = num_targets;
 
-   if (nvc0->tfbbuf_dirty)
+   if (nvc0->tfbbuf_dirty) {
+      nouveau_bufctx_reset(nvc0->bufctx_3d, NVC0_BIND_TFB);
       nvc0->dirty |= NVC0_NEW_TFB_TARGETS;
+   }
 }
 
 static void

src/gallium/drivers/nouveau/nvc0/nvc0_surface.c

@@ -1195,8 +1195,8 @@ nvc0_blit_3d(struct nvc0_context *nvc0, const struct pipe_blit_info *info)
    x0 = (float)info->src.box.x - x_range * (float)info->dst.box.x;
    y0 = (float)info->src.box.y - y_range * (float)info->dst.box.y;
 
-   x1 = x0 + 16384.0f * x_range;
-   y1 = y0 + 16384.0f * y_range;
+   x1 = x0 + 32768.0f * x_range;
+   y1 = y0 + 32768.0f * y_range;
 
    x0 *= (float)(1 << nv50_miptree(src)->ms_x);
    x1 *= (float)(1 << nv50_miptree(src)->ms_x);
@@ -1294,6 +1294,17 @@ nvc0_blit_3d(struct nvc0_context *nvc0, const struct pipe_blit_info *info)
    }
    nvc0->state.num_vtxelts = 2;
 
+   if (nvc0->state.prim_restart) {
+      IMMED_NVC0(push, NVC0_3D(PRIM_RESTART_ENABLE), 0);
+      nvc0->state.prim_restart = 0;
+   }
+
+   if (nvc0->state.index_bias) {
+      IMMED_NVC0(push, NVC0_3D(VB_ELEMENT_BASE), 0);
+      IMMED_NVC0(push, NVC0_3D(VERTEX_ID_BASE), 0);
+      nvc0->state.index_bias = 0;
+   }
+
    for (i = 0; i < info->dst.box.depth; ++i, z += dz) {
       if (info->dst.box.z + i) {
          BEGIN_NVC0(push, NVC0_3D(LAYER), 1);
@@ -1306,14 +1317,14 @@ nvc0_blit_3d(struct nvc0_context *nvc0, const struct pipe_blit_info *info)
       *(vbuf++) = fui(y0);
       *(vbuf++) = fui(z);
 
-      *(vbuf++) = fui(16384 << nv50_miptree(dst)->ms_x);
+      *(vbuf++) = fui(32768 << nv50_miptree(dst)->ms_x);
       *(vbuf++) = fui(0.0f);
       *(vbuf++) = fui(x1);
       *(vbuf++) = fui(y0);
       *(vbuf++) = fui(z);
 
       *(vbuf++) = fui(0.0f);
-      *(vbuf++) = fui(16384 << nv50_miptree(dst)->ms_y);
+      *(vbuf++) = fui(32768 << nv50_miptree(dst)->ms_y);
       *(vbuf++) = fui(x0);
       *(vbuf++) = fui(y1);
       *(vbuf++) = fui(z);

src/gallium/drivers/nouveau/nvc0/nvc0_vbo.c

@@ -379,7 +379,7 @@ nvc0_validate_vertex_buffers_shared(struct nvc0_context *nvc0)
    unsigned b;
    const uint32_t mask = nvc0->vbo_user;
 
-   PUSH_SPACE(push, nvc0->num_vtxbufs * 8);
+   PUSH_SPACE(push, nvc0->num_vtxbufs * 8 + nvc0->vertex->num_elements);
    for (b = 0; b < nvc0->num_vtxbufs; ++b) {
       struct pipe_vertex_buffer *vb = &nvc0->vtxbuf[b];
       struct nv04_resource *buf;
@@ -414,6 +414,7 @@ nvc0_validate_vertex_buffers_shared(struct nvc0_context *nvc0)
    /* If there are more elements than buffers, we might not have unset
     * fetching on the later elements.
     */
+   PUSH_SPACE(push, nvc0->vertex->num_elements - b);
    for (; b < nvc0->vertex->num_elements; ++b)
       IMMED_NVC0(push, NVC0_3D(VERTEX_ARRAY_FETCH(b)), 0);
 
@@ -687,7 +688,7 @@ nvc0_draw_elements_inline_u32_short(struct nouveau_pushbuf *push,
 
    if (count & 1) {
       count--;
-      PUSH_SPACE(push, 1);
+      PUSH_SPACE(push, 2);
       BEGIN_NVC0(push, NVC0_3D(VB_ELEMENT_U32), 1);
       PUSH_DATA (push, *map++);
    }
@@ -817,6 +818,8 @@ nvc0_draw_indirect(struct nvc0_context *nvc0, const struct pipe_draw_info *info)
    unsigned size;
    const uint32_t offset = buf->offset + info->indirect_offset;
 
+   PUSH_SPACE(push, 7);
+
    /* must make FIFO wait for engines idle before continuing to process */
    if (buf->fence_wr && !nouveau_fence_signalled(buf->fence_wr))
       IMMED_NVC0(push, SUBC_3D(NV10_SUBCHAN_REF_CNT), 0);
@@ -901,6 +904,7 @@ nvc0_draw_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
    if (info->mode == PIPE_PRIM_PATCHES &&
        nvc0->state.patch_vertices != info->vertices_per_patch) {
       nvc0->state.patch_vertices = info->vertices_per_patch;
+      PUSH_SPACE(push, 1);
       IMMED_NVC0(push, NVC0_3D(PATCH_VERTICES), nvc0->state.patch_vertices);
    }
 
@@ -933,11 +937,14 @@ nvc0_draw_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
    }
 
    if (nvc0->cb_dirty) {
+      PUSH_SPACE(push, 1);
       IMMED_NVC0(push, NVC0_3D(MEM_BARRIER), 0x1011);
       nvc0->cb_dirty = false;
    }
 
    for (s = 0; s < 5; ++s) {
+      PUSH_SPACE(push, nvc0->num_textures[s] * 2);
+
       for (int i = 0; i < nvc0->num_textures[s]; ++i) {
          struct nv50_tic_entry *tic = nv50_tic_entry(nvc0->textures[s][i]);
          struct pipe_resource *res;

src/gallium/drivers/r600/evergreen_state.c

@@ -962,13 +962,6 @@ void evergreen_init_color_surface_rat(struct r600_context *rctx,
 		MAX2(64, rctx->screen->b.tiling_info.group_bytes / block_size);
 	unsigned pitch = align(pipe_buffer->width0, pitch_alignment);
 
-	/* XXX: This is copied from evergreen_init_color_surface().  I don't
-	 * know why this is necessary.
-	 */
-	if (pipe_buffer->usage == PIPE_USAGE_STAGING) {
-		endian = ENDIAN_NONE;
-	}
-
 	surf->cb_color_base = r600_resource(pipe_buffer)->gpu_address >> 8;
 
 	surf->cb_color_pitch = (pitch / 8) - 1;
@@ -1119,11 +1112,7 @@ void evergreen_init_color_surface(struct r600_context *rctx,
 	swap = r600_translate_colorswap(surf->base.format);
 	assert(swap != ~0);
 
-	if (rtex->resource.b.b.usage == PIPE_USAGE_STAGING) {
-		endian = ENDIAN_NONE;
-	} else {
-		endian = r600_colorformat_endian_swap(format);
-	}
+	endian = r600_colorformat_endian_swap(format);
 
 	/* blend clamp should be set for all NORM/SRGB types */
 	if (ntype == V_028C70_NUMBER_UNORM || ntype == V_028C70_NUMBER_SNORM ||

src/gallium/drivers/r600/r600_isa.h

@@ -102,13 +102,13 @@ enum alu_op_flags
 
 	/* condition codes - 3 bits */
 	AF_CC_SHIFT = 29,
-	AF_CC_MASK	= (7 << AF_CC_SHIFT),
-	AF_CC_E		= (0 << AF_CC_SHIFT),
-	AF_CC_GT	= (1 << AF_CC_SHIFT),
-	AF_CC_GE	= (2 << AF_CC_SHIFT),
-	AF_CC_NE	= (3 << AF_CC_SHIFT),
-	AF_CC_LT	= (4 << AF_CC_SHIFT),
-	AF_CC_LE	= (5 << AF_CC_SHIFT),
+	AF_CC_MASK	= (7U << AF_CC_SHIFT),
+	AF_CC_E		= (0U << AF_CC_SHIFT),
+	AF_CC_GT	= (1U << AF_CC_SHIFT),
+	AF_CC_GE	= (2U << AF_CC_SHIFT),
+	AF_CC_NE	= (3U << AF_CC_SHIFT),
+	AF_CC_LT	= (4U << AF_CC_SHIFT),
+	AF_CC_LE	= (5U << AF_CC_SHIFT),
 };
 
 /* flags for FETCH instructions (TEX/VTX) */
@@ -165,7 +165,7 @@ struct alu_op_info
 	 * (0 if instruction doesn't exist for chip class) */
 	int	slots[4];
 	/* flags (mostly autogenerated from instruction name) */
-	int	flags;
+	unsigned int flags;
 };
 
 /* FETCH instruction info */

src/gallium/drivers/r600/r600_state.c

@@ -930,11 +930,7 @@ static void r600_init_color_surface(struct r600_context *rctx,
 	swap = r600_translate_colorswap(surf->base.format);
 	assert(swap != ~0);
 
-	if (rtex->resource.b.b.usage == PIPE_USAGE_STAGING) {
-		endian = ENDIAN_NONE;
-	} else {
-		endian = r600_colorformat_endian_swap(format);
-	}
+	endian = r600_colorformat_endian_swap(format);
 
 	/* set blend bypass according to docs if SINT/UINT or
 	   8/24 COLOR variants */

src/gallium/drivers/r600/r600_state_common.c

@@ -643,21 +643,21 @@ static void r600_set_sampler_views(struct pipe_context *pipe, unsigned shader,
 		if (rviews[i]) {
 			struct r600_texture *rtex =
 				(struct r600_texture*)rviews[i]->base.texture;
+			bool is_buffer = rviews[i]->base.texture->target == PIPE_BUFFER;
 
-			if (rviews[i]->base.texture->target != PIPE_BUFFER) {
-				if (rtex->is_depth && !rtex->is_flushing_texture) {
-					dst->views.compressed_depthtex_mask |= 1 << i;
-				} else {
-					dst->views.compressed_depthtex_mask &= ~(1 << i);
-				}
-
-				/* Track compressed colorbuffers. */
-				if (rtex->cmask.size) {
-					dst->views.compressed_colortex_mask |= 1 << i;
-				} else {
-					dst->views.compressed_colortex_mask &= ~(1 << i);
-				}
+			if (!is_buffer && rtex->is_depth && !rtex->is_flushing_texture) {
+				dst->views.compressed_depthtex_mask |= 1 << i;
+			} else {
+				dst->views.compressed_depthtex_mask &= ~(1 << i);
 			}
+
+			/* Track compressed colorbuffers. */
+			if (!is_buffer && rtex->cmask.size) {
+				dst->views.compressed_colortex_mask |= 1 << i;
+			} else {
+				dst->views.compressed_colortex_mask &= ~(1 << i);
+			}
+
 			/* Changing from array to non-arrays textures and vice versa requires
 			 * updating TEX_ARRAY_OVERRIDE in sampler states on R6xx-R7xx. */
 			if (rctx->b.chip_class <= R700 &&
@@ -2517,6 +2517,13 @@ uint32_t r600_colorformat_endian_swap(uint32_t colorformat)
 
 		/* 32-bit buffers. */
 		case V_0280A0_COLOR_8_8_8_8:
+			/*
+			 * No need to do endian swaps on four 8-bits components,
+			 * as mesa<-->pipe formats conversion take into account
+			 * the endianess
+			 */
+			return ENDIAN_NONE;
+
 		case V_0280A0_COLOR_2_10_10_10:
 		case V_0280A0_COLOR_8_24:
 		case V_0280A0_COLOR_24_8:

src/gallium/drivers/r600/sb/sb_expr.cpp

@@ -598,9 +598,13 @@ bool expr_handler::fold_assoc(alu_node *n) {
 
 	unsigned op = n->bc.op;
 	bool allow_neg = false, cur_neg = false;
+	bool distribute_neg = false;
 
 	switch(op) {
 	case ALU_OP2_ADD:
+		distribute_neg = true;
+		allow_neg = true;
+		break;
 	case ALU_OP2_MUL:
 	case ALU_OP2_MUL_IEEE:
 		allow_neg = true;
@@ -632,7 +636,7 @@ bool expr_handler::fold_assoc(alu_node *n) {
 		if (v1->is_const()) {
 			literal arg = v1->get_const_value();
 			apply_alu_src_mod(a->bc, 1, arg);
-			if (cur_neg)
+			if (cur_neg && distribute_neg)
 				arg.f = -arg.f;
 
 			if (a == n)
@@ -660,7 +664,7 @@ bool expr_handler::fold_assoc(alu_node *n) {
 		if (v0->is_const()) {
 			literal arg = v0->get_const_value();
 			apply_alu_src_mod(a->bc, 0, arg);
-			if (cur_neg)
+			if (cur_neg && distribute_neg)
 				arg.f = -arg.f;
 
 			if (last_arg == 0) {

src/gallium/drivers/radeon/r600_buffer_common.c

@@ -286,7 +286,8 @@ static void *r600_buffer_transfer_map(struct pipe_context *ctx,
 		usage |= PIPE_TRANSFER_UNSYNCHRONIZED;
 	}
 	else if ((usage & PIPE_TRANSFER_DISCARD_RANGE) &&
-		 !(usage & PIPE_TRANSFER_UNSYNCHRONIZED) &&
+		 !(usage & (PIPE_TRANSFER_UNSYNCHRONIZED |
+			    PIPE_TRANSFER_PERSISTENT)) &&
 		 !(rscreen->debug_flags & DBG_NO_DISCARD_RANGE) &&
 		 r600_can_dma_copy_buffer(rctx, box->x, 0, box->width)) {
 		assert(usage & PIPE_TRANSFER_WRITE);
@@ -313,7 +314,8 @@ static void *r600_buffer_transfer_map(struct pipe_context *ctx,
 	}
 	/* Using a staging buffer in GTT for larger reads is much faster. */
 	else if ((usage & PIPE_TRANSFER_READ) &&
-		 !(usage & PIPE_TRANSFER_WRITE) &&
+		 !(usage & (PIPE_TRANSFER_WRITE |
+			    PIPE_TRANSFER_PERSISTENT)) &&
 		 rbuffer->domains == RADEON_DOMAIN_VRAM &&
 		 r600_can_dma_copy_buffer(rctx, 0, box->x, box->width)) {
 		struct r600_resource *staging;

src/gallium/drivers/radeon/r600_pipe_common.h

@@ -629,7 +629,7 @@ r600_get_sampler_view_priority(struct r600_resource *res)
 	} while (0);
 
 #define R600_ERR(fmt, args...) \
-	fprintf(stderr, "EE %s:%d %s - "fmt, __FILE__, __LINE__, __func__, ##args)
+	fprintf(stderr, "EE %s:%d %s - " fmt, __FILE__, __LINE__, __func__, ##args)
 
 /* For MSAA sample positions. */
 #define FILL_SREG(s0x, s0y, s1x, s1y, s2x, s2y, s3x, s3y)  \

src/gallium/drivers/radeon/r600_texture.c

@@ -529,6 +529,16 @@ static unsigned r600_texture_get_htile_size(struct r600_common_screen *rscreen,
 	    rscreen->info.drm_major == 2 && rscreen->info.drm_minor < 38)
 		return 0;
 
+	/* Overalign HTILE on P2 configs to work around GPU hangs in
+	 * piglit/depthstencil-render-miplevels 585.
+	 *
+	 * This has been confirmed to help Kabini & Stoney, where the hangs
+	 * are always reproducible. I think I have seen the test hang
+	 * on Carrizo too, though it was very rare there.
+	 */
+	if (rscreen->chip_class >= CIK && num_pipes < 4)
+		num_pipes = 4;
+
 	switch (num_pipes) {
 	case 1:
 		cl_width = 32;
@@ -1219,8 +1229,8 @@ unsigned r600_translate_colorswap(enum pipe_format format)
 			return V_0280A0_SWAP_STD_REV; /* WZYX */
 		else if (HAS_SWIZZLE(1,Y) && HAS_SWIZZLE(2,X))
 			return V_0280A0_SWAP_ALT; /* ZYXW */
-		else if (HAS_SWIZZLE(1,X) && HAS_SWIZZLE(2,Y))
-			return V_0280A0_SWAP_ALT_REV; /* WXYZ */
+		else if (HAS_SWIZZLE(1,Z) && HAS_SWIZZLE(2,W))
+			return V_0280A0_SWAP_ALT_REV; /* YZWX */
 		break;
 	}
 	return ~0U;
@@ -1328,6 +1338,11 @@ void evergreen_do_fast_color_clear(struct r600_common_context *rctx,
 {
 	int i;
 
+	/* This function is broken in BE, so just disable this path for now */
+#ifdef PIPE_ARCH_BIG_ENDIAN
+	return;
+#endif
+
 	if (rctx->render_cond)
 		return;
 

src/gallium/drivers/radeon/radeon_video.c

@@ -237,6 +237,7 @@ int rvid_get_video_param(struct pipe_screen *screen,
 	case PIPE_VIDEO_CAP_SUPPORTED:
 		switch (codec) {
 		case PIPE_VIDEO_FORMAT_MPEG12:
+			return profile != PIPE_VIDEO_PROFILE_MPEG1;
 		case PIPE_VIDEO_FORMAT_MPEG4:
 		case PIPE_VIDEO_FORMAT_MPEG4_AVC:
 			if (rscreen->family < CHIP_PALM)
@@ -257,7 +258,7 @@ int rvid_get_video_param(struct pipe_screen *screen,
 	case PIPE_VIDEO_CAP_MAX_WIDTH:
 		return (rscreen->family < CHIP_TONGA) ? 2048 : 4096;
 	case PIPE_VIDEO_CAP_MAX_HEIGHT:
-		return (rscreen->family < CHIP_TONGA) ? 1152 : 2304;
+		return (rscreen->family < CHIP_TONGA) ? 1152 : 4096;
 	case PIPE_VIDEO_CAP_PREFERED_FORMAT:
 		return PIPE_FORMAT_NV12;
 	case PIPE_VIDEO_CAP_PREFERS_INTERLACED:

src/gallium/drivers/radeonsi/si_state.c

@@ -1862,11 +1862,7 @@ static void si_initialize_color_surface(struct si_context *sctx,
 	}
 	assert(format != V_028C70_COLOR_INVALID);
 	swap = r600_translate_colorswap(surf->base.format);
-	if (rtex->resource.b.b.usage == PIPE_USAGE_STAGING) {
-		endian = V_028C70_ENDIAN_NONE;
-	} else {
-		endian = si_colorformat_endian_swap(format);
-	}
+	endian = si_colorformat_endian_swap(format);
 
 	/* blend clamp should be set for all NORM/SRGB types */
 	if (ntype == V_028C70_NUMBER_UNORM ||

src/gallium/drivers/softpipe/sp_tex_sample.c

@@ -2209,6 +2209,7 @@ img_filter_2d_ewa(const struct sp_sampler_view *sp_sview,
                   const float t[TGSI_QUAD_SIZE],
                   const float p[TGSI_QUAD_SIZE],
                   const uint faces[TGSI_QUAD_SIZE],
+                  const int8_t *offset,
                   unsigned level,
                   const float dudx, const float dvdx,
                   const float dudy, const float dvdy,
@@ -2268,6 +2269,8 @@ img_filter_2d_ewa(const struct sp_sampler_view *sp_sview,
    /* F *= formScale; */ /* no need to scale F as we don't use it below here */
 
    args.level = level;
+   args.offset = offset;
+
    for (j = 0; j < TGSI_QUAD_SIZE; j++) {
       /* Heckbert MS thesis, p. 59; scan over the bounding box of the ellipse
        * and incrementally update the value of Ax^2+Bxy*Cy^2; when this
@@ -2431,6 +2434,8 @@ mip_filter_linear_aniso(const struct sp_sampler_view *sp_sview,
    const float dvdy = (t[QUAD_TOP_LEFT]     - t[QUAD_BOTTOM_LEFT]) * t_to_v;
    struct img_filter_args args;
 
+   args.offset = filt_args->offset;
+
    if (filt_args->control == TGSI_SAMPLER_LOD_BIAS ||
        filt_args->control == TGSI_SAMPLER_LOD_NONE ||
        /* XXX FIXME */
@@ -2503,8 +2508,8 @@ mip_filter_linear_aniso(const struct sp_sampler_view *sp_sview,
        * seem to be worth the extra running time.
        */
       img_filter_2d_ewa(sp_sview, sp_samp, min_filter, mag_filter,
-                        s, t, p, filt_args->faces, level0,
-                        dudx, dvdx, dudy, dvdy, rgba);
+                        s, t, p, filt_args->faces, filt_args->offset,
+                        level0, dudx, dvdx, dudy, dvdy, rgba);
    }
 
    if (DEBUG_TEX) {

src/gallium/state_trackers/clover/Makefile.am

@@ -41,6 +41,7 @@ libclllvm_la_CXXFLAGS = \
 	$(VISIBILITY_CXXFLAGS) \
 	$(LLVM_CXXFLAGS) \
 	$(DEFINES) \
+	$(LIBELF_CFLAGS) \
 	-DLIBCLC_INCLUDEDIR=\"$(LIBCLC_INCLUDEDIR)/\" \
 	-DLIBCLC_LIBEXECDIR=\"$(LIBCLC_LIBEXECDIR)/\" \
 	-DCLANG_RESOURCE_DIR=\"$(CLANG_RESOURCE_DIR)\"

src/gallium/state_trackers/omx/vid_dec.c

@@ -140,7 +140,7 @@ static OMX_ERRORTYPE vid_dec_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam
 
    r = omx_base_filter_Constructor(comp, name);
    if (r)
-	return r;
+      return r;
 
    priv->profile = PIPE_VIDEO_PROFILE_UNKNOWN;
 
@@ -268,7 +268,7 @@ static OMX_ERRORTYPE vid_dec_SetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i
       r = checkHeader(param, sizeof(OMX_PARAM_COMPONENTROLETYPE));
       if (r)
          return r;
- 
+
       if (!strcmp((char *)role->cRole, OMX_VID_DEC_MPEG2_ROLE)) {
          priv->profile = PIPE_VIDEO_PROFILE_MPEG2_MAIN;
       } else if (!strcmp((char *)role->cRole, OMX_VID_DEC_AVC_ROLE)) {
@@ -321,7 +321,7 @@ static OMX_ERRORTYPE vid_dec_GetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i
          strcpy((char *)role->cRole, OMX_VID_DEC_MPEG2_ROLE);
       else if (priv->profile == PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH)
          strcpy((char *)role->cRole, OMX_VID_DEC_AVC_ROLE);
- 
+
       break;
    }
 
@@ -419,6 +419,7 @@ static OMX_ERRORTYPE vid_dec_DecodeBuffer(omx_base_PortType *port, OMX_BUFFERHEA
    priv->in_buffers[i] = buf;
    priv->sizes[i] = buf->nFilledLen;
    priv->inputs[i] = buf->pBuffer;
+   priv->timestamps[i] = buf->nTimeStamp;
 
    while (priv->num_in_buffers > (!!(buf->nFlags & OMX_BUFFERFLAG_EOS) ? 0 : 1)) {
       bool eos = !!(priv->in_buffers[0]->nFlags & OMX_BUFFERFLAG_EOS);
@@ -469,12 +470,13 @@ static OMX_ERRORTYPE vid_dec_DecodeBuffer(omx_base_PortType *port, OMX_BUFFERHEA
          priv->in_buffers[0] = priv->in_buffers[1];
          priv->sizes[0] = priv->sizes[1] - delta;
          priv->inputs[0] = priv->inputs[1] + delta;
+         priv->timestamps[0] = priv->timestamps[1];
       }
 
       if (r)
          return r;
    }
- 
+
    return OMX_ErrorNone;
 }
 
@@ -513,7 +515,7 @@ static void vid_dec_FillOutput(vid_dec_PrivateType *priv, struct pipe_video_buff
 
    box.width = def->nFrameWidth / 2;
    box.height = def->nFrameHeight / 2;
- 
+
    src = priv->pipe->transfer_map(priv->pipe, views[1]->texture, 0,
                                   PIPE_TRANSFER_READ, &box, &transfer);
    util_copy_rect(dst, views[1]->texture->format, def->nStride, 0, 0,
@@ -526,9 +528,13 @@ static void vid_dec_FrameDecoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE*
 {
    vid_dec_PrivateType *priv = comp->pComponentPrivate;
    bool eos = !!(input->nFlags & OMX_BUFFERFLAG_EOS);
+   OMX_TICKS timestamp;
 
-   if (!input->pInputPortPrivate)
-      input->pInputPortPrivate = priv->Flush(priv);
+   if (!input->pInputPortPrivate) {
+      input->pInputPortPrivate = priv->Flush(priv, &timestamp);
+      if (timestamp != OMX_VID_DEC_TIMESTAMP_INVALID)
+         input->nTimeStamp = timestamp;
+   }
 
    if (input->pInputPortPrivate) {
       if (output->pInputPortPrivate) {
@@ -539,6 +545,7 @@ static void vid_dec_FrameDecoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE*
          vid_dec_FillOutput(priv, input->pInputPortPrivate, output);
       }
       output->nFilledLen = output->nAllocLen;
+      output->nTimeStamp = input->nTimeStamp;
    }
 
    if (eos && input->pInputPortPrivate)

src/gallium/state_trackers/omx/vid_dec.h

@@ -59,6 +59,8 @@
 #define OMX_VID_DEC_AVC_NAME "OMX.mesa.video_decoder.avc"
 #define OMX_VID_DEC_AVC_ROLE "video_decoder.avc"
 
+#define OMX_VID_DEC_TIMESTAMP_INVALID ((OMX_TICKS) -1)
+
 struct vl_vlc;
 
 DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType)
@@ -69,7 +71,7 @@ DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType)
    struct pipe_video_codec *codec; \
    void (*Decode)(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left); \
    void (*EndFrame)(vid_dec_PrivateType *priv); \
-   struct pipe_video_buffer *(*Flush)(vid_dec_PrivateType *priv); \
+   struct pipe_video_buffer *(*Flush)(vid_dec_PrivateType *priv, OMX_TICKS *timestamp); \
    struct pipe_video_buffer *target, *shadow; \
    union { \
       struct { \
@@ -100,6 +102,9 @@ DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType)
    OMX_BUFFERHEADERTYPE *in_buffers[2]; \
    const void *inputs[2]; \
    unsigned sizes[2]; \
+   OMX_TICKS timestamps[2]; \
+   OMX_TICKS timestamp; \
+   bool first_buf_in_frame; \
    bool frame_finished; \
    bool frame_started; \
    unsigned bytes_left; \

src/gallium/state_trackers/omx/vid_dec_h264.c

@@ -45,6 +45,7 @@
 struct dpb_list {
    struct list_head list;
    struct pipe_video_buffer *buffer;
+   OMX_TICKS timestamp;
    unsigned poc;
 };
 
@@ -82,7 +83,7 @@ static const uint8_t Default_8x8_Inter[64] = {
 
 static void vid_dec_h264_Decode(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left);
 static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv);
-static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv);
+static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp);
 
 void vid_dec_h264_Init(vid_dec_PrivateType *priv)
 {
@@ -91,9 +92,10 @@ void vid_dec_h264_Init(vid_dec_PrivateType *priv)
    priv->Decode = vid_dec_h264_Decode;
    priv->EndFrame = vid_dec_h264_EndFrame;
    priv->Flush = vid_dec_h264_Flush;
-   
+
    LIST_INITHEAD(&priv->codec_data.h264.dpb_list);
    priv->picture.h264.field_order_cnt[0] = priv->picture.h264.field_order_cnt[1] = INT_MAX;
+   priv->first_buf_in_frame = true;
 }
 
 static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv)
@@ -104,6 +106,9 @@ static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv)
       return;
 
    vid_dec_NeedTarget(priv);
+   if (priv->first_buf_in_frame)
+      priv->timestamp = priv->timestamps[0];
+   priv->first_buf_in_frame = false;
 
    priv->picture.h264.num_ref_frames = priv->picture.h264.pps->sps->max_num_ref_frames;
 
@@ -127,7 +132,8 @@ static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv)
    priv->frame_started = true;
 }
 
-static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv)
+static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv,
+                                                    OMX_TICKS *timestamp)
 {
    struct dpb_list *entry, *result = NULL;
    struct pipe_video_buffer *buf;
@@ -146,6 +152,8 @@ static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv)
       return NULL;
 
    buf = result->buffer;
+   if (timestamp)
+      *timestamp = result->timestamp;
 
    --priv->codec_data.h264.dpb_num;
    LIST_DEL(&result->list);
@@ -159,6 +167,7 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv)
    struct dpb_list *entry;
    struct pipe_video_buffer *tmp;
    bool top_field_first;
+   OMX_TICKS timestamp;
 
    if (!priv->frame_started)
       return;
@@ -181,7 +190,9 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv)
    if (!entry)
       return;
 
+   priv->first_buf_in_frame = true;
    entry->buffer = priv->target;
+   entry->timestamp = priv->timestamp;
    entry->poc = MIN2(priv->picture.h264.field_order_cnt[0], priv->picture.h264.field_order_cnt[1]);
    LIST_ADDTAIL(&entry->list, &priv->codec_data.h264.dpb_list);
    ++priv->codec_data.h264.dpb_num;
@@ -192,7 +203,8 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv)
       return;
 
    tmp = priv->in_buffers[0]->pInputPortPrivate;
-   priv->in_buffers[0]->pInputPortPrivate = vid_dec_h264_Flush(priv);
+   priv->in_buffers[0]->pInputPortPrivate = vid_dec_h264_Flush(priv, &timestamp);
+   priv->in_buffers[0]->nTimeStamp = timestamp;
    priv->target = tmp;
    priv->frame_finished = priv->in_buffers[0]->pInputPortPrivate != NULL;
 }
@@ -829,7 +841,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp,
          priv->picture.h264.field_order_cnt[0] = expectedPicOrderCnt + priv->codec_data.h264.delta_pic_order_cnt[0];
          priv->picture.h264.field_order_cnt[1] = priv->picture.h264.field_order_cnt[0] +
             sps->offset_for_top_to_bottom_field + priv->codec_data.h264.delta_pic_order_cnt[1];
-         
+
       } else if (!priv->picture.h264.bottom_field_flag)
          priv->picture.h264.field_order_cnt[0] = expectedPicOrderCnt + priv->codec_data.h264.delta_pic_order_cnt[0];
       else
@@ -859,7 +871,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp,
       if (!priv->picture.h264.field_pic_flag) {
          priv->picture.h264.field_order_cnt[0] = tempPicOrderCnt;
          priv->picture.h264.field_order_cnt[1] = tempPicOrderCnt;
-         
+
       } else if (!priv->picture.h264.bottom_field_flag)
          priv->picture.h264.field_order_cnt[0] = tempPicOrderCnt;
       else
@@ -876,7 +888,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp,
 
    priv->picture.h264.num_ref_idx_l0_active_minus1 = pps->num_ref_idx_l0_default_active_minus1;
    priv->picture.h264.num_ref_idx_l1_active_minus1 = pps->num_ref_idx_l1_default_active_minus1;
- 
+
    if (slice_type == PIPE_H264_SLICE_TYPE_P ||
        slice_type == PIPE_H264_SLICE_TYPE_SP ||
        slice_type == PIPE_H264_SLICE_TYPE_B) {

src/gallium/state_trackers/omx/vid_dec_mpeg12.c

@@ -61,7 +61,7 @@ static uint8_t default_non_intra_matrix[64] = {
 
 static void vid_dec_mpeg12_Decode(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left);
 static void vid_dec_mpeg12_EndFrame(vid_dec_PrivateType *priv);
-static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv);
+static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp);
 
 void vid_dec_mpeg12_Init(vid_dec_PrivateType *priv)
 {
@@ -131,10 +131,12 @@ static void vid_dec_mpeg12_EndFrame(vid_dec_PrivateType *priv)
    priv->in_buffers[0]->pInputPortPrivate = done;
 }
 
-static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv)
+static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp)
 {
    struct pipe_video_buffer *result = priv->picture.mpeg12.ref[1];
    priv->picture.mpeg12.ref[1] = NULL;
+   if (timestamp)
+      *timestamp = OMX_VID_DEC_TIMESTAMP_INVALID;
    return result;
 }
 

src/gallium/state_trackers/omx/vid_enc.c

@@ -179,7 +179,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam
    if (!screen->get_video_param(screen, PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH,
                                 PIPE_VIDEO_ENTRYPOINT_ENCODE, PIPE_VIDEO_CAP_SUPPORTED))
       return OMX_ErrorBadParameter;
- 
+
    priv->stacked_frames_num = screen->get_video_param(screen,
                                 PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH,
                                 PIPE_VIDEO_ENTRYPOINT_ENCODE,
@@ -242,7 +242,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam
 
    port->Port_AllocateBuffer = vid_enc_AllocateOutBuffer;
    port->Port_FreeBuffer = vid_enc_FreeOutBuffer;
- 
+
    priv->bitrate.eControlRate = OMX_Video_ControlRateDisable;
    priv->bitrate.nTargetBitrate = 0;
 
@@ -253,7 +253,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam
    priv->profile_level.eProfile = OMX_VIDEO_AVCProfileBaseline;
    priv->profile_level.eLevel = OMX_VIDEO_AVCLevel42;
 
-   priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; 
+   priv->force_pic_type.IntraRefreshVOP = OMX_FALSE;
    priv->frame_num = 0;
    priv->pic_order_cnt = 0;
    priv->restricted_b_frames = debug_get_bool_option("OMX_USE_RESTRICTED_B_FRAMES", FALSE);
@@ -380,7 +380,7 @@ static OMX_ERRORTYPE vid_enc_SetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i
 
          port = (omx_base_video_PortType *)priv->ports[OMX_BASE_FILTER_OUTPUTPORT_INDEX];
          port->sPortParam.nBufferSize = framesize * 512 / (16*16);
-      
+
          priv->frame_rate = def->format.video.xFramerate;
 
          priv->callbacks->EventHandler(comp, priv->callbackData, OMX_EventPortSettingsChanged,
@@ -532,10 +532,10 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx,
    vid_enc_PrivateType *priv = comp->pComponentPrivate;
    OMX_ERRORTYPE r;
    int i;
- 
+
    if (!config)
       return OMX_ErrorBadParameter;
-                         
+
    switch(idx) {
    case OMX_IndexConfigVideoIntraVOPRefresh: {
       OMX_CONFIG_INTRAREFRESHVOPTYPE *type = config;
@@ -543,9 +543,9 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx,
       r = checkHeader(config, sizeof(OMX_CONFIG_INTRAREFRESHVOPTYPE));
       if (r)
          return r;
-      
+
       priv->force_pic_type = *type;
-      
+
       break;
    }
    case OMX_IndexConfigCommonScale: {
@@ -568,11 +568,11 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx,
       priv->scale = *scale;
       if (priv->scale.xWidth != 0xffffffff && priv->scale.xHeight != 0xffffffff) {
          struct pipe_video_buffer templat = {};
- 
+
          templat.buffer_format = PIPE_FORMAT_NV12;
          templat.chroma_format = PIPE_VIDEO_CHROMA_FORMAT_420;
-         templat.width = priv->scale.xWidth; 
-         templat.height = priv->scale.xHeight; 
+         templat.width = priv->scale.xWidth;
+         templat.height = priv->scale.xHeight;
          templat.interlaced = false;
          for (i = 0; i < OMX_VID_ENC_NUM_SCALING_BUFFERS; ++i) {
             priv->scale_buffer[i] = priv->s_pipe->create_video_buffer(priv->s_pipe, &templat);
@@ -615,7 +615,7 @@ static OMX_ERRORTYPE vid_enc_GetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx,
    default:
       return omx_base_component_GetConfig(handle, idx, config);
    }
-   
+
    return OMX_ErrorNone;
 }
 
@@ -1007,10 +1007,10 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic
    switch (priv->bitrate.eControlRate) {
    case OMX_Video_ControlRateVariable:
       rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_VARIABLE;
-      break; 
+      break;
    case OMX_Video_ControlRateConstant:
       rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_CONSTANT;
-      break; 
+      break;
    case OMX_Video_ControlRateVariableSkipFrames:
       rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_VARIABLE_SKIP;
       break;
@@ -1020,8 +1020,8 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic
    default:
       rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_DISABLE;
       break;
-   } 
-      
+   }
+
    rate_ctrl->frame_rate_den = OMX_VID_ENC_CONTROL_FRAME_RATE_DEN_DEFAULT;
    rate_ctrl->frame_rate_num = ((priv->frame_rate) >> 16) * rate_ctrl->frame_rate_den;
 
@@ -1032,7 +1032,7 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic
          rate_ctrl->target_bitrate = priv->bitrate.nTargetBitrate;
       else
          rate_ctrl->target_bitrate = OMX_VID_ENC_BITRATE_MAX;
-      rate_ctrl->peak_bitrate = rate_ctrl->target_bitrate;    
+      rate_ctrl->peak_bitrate = rate_ctrl->target_bitrate;
       if (rate_ctrl->target_bitrate < OMX_VID_ENC_BITRATE_MEDIAN)
          rate_ctrl->vbv_buffer_size = MIN2((rate_ctrl->target_bitrate * 2.75), OMX_VID_ENC_BITRATE_MEDIAN);
       else
@@ -1048,7 +1048,7 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic
       rate_ctrl->peak_bits_picture_integer = rate_ctrl->target_bits_picture;
       rate_ctrl->peak_bits_picture_fraction = 0;
    }
-   
+
    picture->quant_i_frames = priv->quant.nQpI;
    picture->quant_p_frames = priv->quant.nQpP;
    picture->quant_b_frames = priv->quant.nQpB;
@@ -1066,7 +1066,7 @@ static void enc_HandleTask(omx_base_PortType *port, struct encode_task *task,
    unsigned size = priv->ports[OMX_BASE_FILTER_OUTPUTPORT_INDEX]->sPortParam.nBufferSize;
    struct pipe_video_buffer *vbuf = task->buf;
    struct pipe_h264_enc_picture_desc picture = {};
- 
+
    /* -------------- scale input image --------- */
    enc_ScaleInput(port, &vbuf, &size);
    priv->s_pipe->flush(priv->s_pipe, NULL, 0);
@@ -1157,7 +1157,7 @@ static OMX_ERRORTYPE vid_enc_EncodeFrame(omx_base_PortType *port, OMX_BUFFERHEAD
        priv->force_pic_type.IntraRefreshVOP) {
       enc_ClearBframes(port, inp);
       picture_type = PIPE_H264_ENC_PICTURE_TYPE_IDR;
-      priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; 
+      priv->force_pic_type.IntraRefreshVOP = OMX_FALSE;
       priv->frame_num = 0;
    } else if (priv->codec->profile == PIPE_VIDEO_PROFILE_MPEG4_AVC_BASELINE ||
               !(priv->pic_order_cnt % OMX_VID_ENC_P_PERIOD_DEFAULT) ||
@@ -1166,7 +1166,7 @@ static OMX_ERRORTYPE vid_enc_EncodeFrame(omx_base_PortType *port, OMX_BUFFERHEAD
    } else {
       picture_type = PIPE_H264_ENC_PICTURE_TYPE_B;
    }
-   
+
    task->pic_order_cnt = priv->pic_order_cnt++;
 
    if (picture_type == PIPE_H264_ENC_PICTURE_TYPE_B) {
@@ -1242,7 +1242,7 @@ static void vid_enc_BufferEncoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE*
    output->pBuffer = priv->t_pipe->transfer_map(priv->t_pipe, outp->bitstream, 0,
                                                 PIPE_TRANSFER_READ_WRITE,
                                                 &box, &outp->transfer);
- 
+
    /* ------------- get size of result ----------------- */
 
    priv->codec->get_feedback(priv->codec, task->feedback, &size);

src/gallium/targets/dri/Android.mk

@@ -28,9 +28,10 @@ include $(CLEAR_VARS)
 LOCAL_MODULE := gallium_dri
 
 ifeq ($(MESA_LOLLIPOP_BUILD),true)
-LOCAL_MODULE_RELATIVE_PATH := $(notdir $(MESA_DRI_MODULE_PATH))
+LOCAL_MODULE_RELATIVE_PATH := $(MESA_DRI_MODULE_REL_PATH)
 else
 LOCAL_MODULE_PATH := $(MESA_DRI_MODULE_PATH)
+LOCAL_UNSTRIPPED_PATH := $(MESA_DRI_MODULE_UNSTRIPPED_PATH)
 endif
 
 LOCAL_SRC_FILES := target.c

src/gallium/targets/opencl/Makefile.am

@@ -2,9 +2,6 @@ include $(top_srcdir)/src/gallium/Automake.inc
 
 lib_LTLIBRARIES = lib@OPENCL_LIBNAME@.la
 
-AM_CPPFLAGS = \
-        $(LIBELF_CFLAGS)
-
 lib@OPENCL_LIBNAME@_la_LDFLAGS = \
 	$(LLVM_LDFLAGS) \
 	-no-undefined \

src/gallium/winsys/svga/drm/vmw_screen_ioctl.c

@@ -52,6 +52,7 @@
 #include <unistd.h>
 
 #define VMW_MAX_DEFAULT_TEXTURE_SIZE   (128 * 1024 * 1024)
+#define VMW_FENCE_TIMEOUT_SECONDS 60
 
 struct vmw_region
 {
@@ -721,7 +722,7 @@ vmw_ioctl_fence_finish(struct vmw_winsys_screen *vws,
    memset(&arg, 0, sizeof(arg));
 
    arg.handle = handle;
-   arg.timeout_us = 10*1000000;
+   arg.timeout_us = VMW_FENCE_TIMEOUT_SECONDS*1000000;
    arg.lazy = 0;
    arg.flags = vflags;
 

src/gallium/winsys/svga/drm/vmw_surface.c

@@ -170,6 +170,8 @@ vmw_svga_winsys_surface_unmap(struct svga_winsys_context *swc,
       *rebind = vsrf->rebind;
       vsrf->rebind = FALSE;
       vmw_svga_winsys_buffer_unmap(&vsrf->screen->base, vsrf->buf);
+   } else {
+      *rebind = FALSE;
    }
    pipe_mutex_unlock(vsrf->mutex);
 }

src/glsl/Android.gen.mk

@@ -58,6 +58,8 @@ define glsl_local-y-to-c-and-h
 	$(hide) $(YACC) -o $@ -p "glcpp_parser_" $<
 endef
 
+YACC_HEADER_SUFFIX := .hpp
+
 define local-yy-to-cpp-and-h
 	@mkdir -p $(dir $@)
 	@echo "Mesa Yacc: $(PRIVATE_MODULE) <= $<"

src/glsl/ast_function.cpp

@@ -558,7 +558,8 @@ done:
 	    state->symbols->add_global_function(f);
 	    emit_function(state, f);
 	 }
-	 f->add_signature(sig->clone_prototype(f, NULL));
+	 sig = sig->clone_prototype(f, NULL);
+	 f->add_signature(sig);
       }
    }
    return sig;

src/glsl/linker.cpp

@@ -2117,6 +2117,7 @@ link_intrastage_shaders(void *mem_ctx,
 
       if (ok) {
          memcpy(linking_shaders, shader_list, num_shaders * sizeof(gl_shader *));
+         _mesa_glsl_initialize_builtin_functions();
          linking_shaders[num_shaders] = _mesa_glsl_get_builtin_function_shader();
 
          ok = link_function_calls(prog, linked, linking_shaders, num_shaders + 1);
@@ -2626,6 +2627,13 @@ assign_attribute_or_color_locations(gl_shader_program *prog,
 	 continue;
       }
 
+      if (num_attr >= ARRAY_SIZE(to_assign)) {
+         linker_error(prog, "too many %s (max %u)",
+                      target_index == MESA_SHADER_VERTEX ?
+                      "vertex shader inputs" : "fragment shader outputs",
+                      (unsigned)ARRAY_SIZE(to_assign));
+         return false;
+      }
       to_assign[num_attr].slots = slots;
       to_assign[num_attr].var = var;
       num_attr++;

src/glsl/nir/nir_lower_vec_to_movs.c

@@ -83,6 +83,8 @@ insert_mov(nir_alu_instr *vec, unsigned start_idx, nir_shader *shader)
       }
    }
 
+   unsigned channels_handled = mov->dest.write_mask;
+
    /* In some situations (if the vecN is involved in a phi-web), we can end
     * up with a mov from a register to itself.  Some of those channels may end
     * up doing nothing and there's no reason to have them as part of the mov.
@@ -103,7 +105,7 @@ insert_mov(nir_alu_instr *vec, unsigned start_idx, nir_shader *shader)
       ralloc_free(mov);
    }
 
-   return mov->dest.write_mask;
+   return channels_handled;
 }
 
 static bool

src/glx/dri2_glx.c

@@ -1102,9 +1102,14 @@ dri2BindExtensions(struct dri2_screen *psc, struct glx_display * priv,
       __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context");
       __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context_profile");
 
-      if ((mask & (1 << __DRI_API_GLES2)) != 0)
-	 __glXEnableDirectExtension(&psc->base,
-				    "GLX_EXT_create_context_es2_profile");
+      if ((mask & ((1 << __DRI_API_GLES) |
+                   (1 << __DRI_API_GLES2) |
+                   (1 << __DRI_API_GLES3))) != 0) {
+         __glXEnableDirectExtension(&psc->base,
+                                    "GLX_EXT_create_context_es_profile");
+         __glXEnableDirectExtension(&psc->base,
+                                    "GLX_EXT_create_context_es2_profile");
+      }
    }
 
    for (i = 0; extensions[i]; i++) {

src/glx/dri3_glx.c

@@ -665,9 +665,14 @@ dri3_bind_extensions(struct dri3_screen *psc, struct glx_display * priv,
    __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context");
    __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context_profile");
 
-   if ((mask & (1 << __DRI_API_GLES2)) != 0)
+   if ((mask & ((1 << __DRI_API_GLES) |
+                (1 << __DRI_API_GLES2) |
+                (1 << __DRI_API_GLES3))) != 0) {
+      __glXEnableDirectExtension(&psc->base,
+                                 "GLX_EXT_create_context_es_profile");
       __glXEnableDirectExtension(&psc->base,
                                  "GLX_EXT_create_context_es2_profile");
+   }
 
    for (i = 0; extensions[i]; i++) {
       /* when on a different gpu than the server, the server pixmaps

src/glx/dri_common.c

@@ -547,9 +547,18 @@ dri2_convert_glx_attribs(unsigned num_attribs, const uint32_t *attribs,
       case GLX_CONTEXT_COMPATIBILITY_PROFILE_BIT_ARB:
 	 *api = __DRI_API_OPENGL;
 	 break;
-      case GLX_CONTEXT_ES2_PROFILE_BIT_EXT:
-	 *api = __DRI_API_GLES2;
-	 break;
+      case GLX_CONTEXT_ES_PROFILE_BIT_EXT:
+         if (*major_ver >= 3)
+            *api = __DRI_API_GLES3;
+         else if (*major_ver == 2 && *minor_ver == 0)
+            *api = __DRI_API_GLES2;
+         else if (*major_ver == 1 && *minor_ver < 2)
+            *api = __DRI_API_GLES;
+         else {
+            *error = __DRI_CTX_ERROR_BAD_API;
+            return false;
+         }
+         break;
       default:
 	 *error = __DRI_CTX_ERROR_BAD_API;
 	 return false;
@@ -580,19 +589,6 @@ dri2_convert_glx_attribs(unsigned num_attribs, const uint32_t *attribs,
       return false;
    }
 
-   /* The GLX_EXT_create_context_es2_profile spec says:
-    *
-    *     "... If the version requested is 2.0, and the
-    *     GLX_CONTEXT_ES2_PROFILE_BIT_EXT bit is set in the
-    *     GLX_CONTEXT_PROFILE_MASK_ARB attribute (see below), then the context
-    *     returned will implement OpenGL ES 2.0. This is the only way in which
-    *     an implementation may request an OpenGL ES 2.0 context."
-    */
-   if (*api == __DRI_API_GLES2 && (*major_ver != 2 || *minor_ver != 0)) {
-      *error = __DRI_CTX_ERROR_BAD_API;
-      return false;
-   }
-
    *error = __DRI_CTX_ERROR_SUCCESS;
    return true;
 }

src/glx/drisw_glx.c

@@ -623,8 +623,10 @@ driswBindExtensions(struct drisw_screen *psc, const __DRIextension **extensions)
       __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context");
       __glXEnableDirectExtension(&psc->base, "GLX_ARB_create_context_profile");
 
-      /* DRISW version >= 2 implies support for OpenGL ES 2.0.
+      /* DRISW version >= 2 implies support for OpenGL ES.
        */
+      __glXEnableDirectExtension(&psc->base,
+				 "GLX_EXT_create_context_es_profile");
       __glXEnableDirectExtension(&psc->base,
 				 "GLX_EXT_create_context_es2_profile");
    }

src/glx/glxextensions.c

@@ -146,6 +146,7 @@ static const struct extension_info known_glx_extensions[] = {
    { GLX(EXT_fbconfig_packed_float),   VER(0,0), Y, Y, N, N },
    { GLX(EXT_framebuffer_sRGB),        VER(0,0), Y, Y, N, N },
    { GLX(EXT_create_context_es2_profile), VER(0,0), Y, N, N, N },
+   { GLX(EXT_create_context_es_profile), VER(0,0), Y, N, N, N },
    { GLX(MESA_copy_sub_buffer),        VER(0,0), Y, N, N, N },
    { GLX(MESA_multithread_makecurrent),VER(0,0), Y, N, Y, N },
    { GLX(MESA_query_renderer),         VER(0,0), Y, N, N, Y },

src/glx/glxextensions.h

@@ -45,6 +45,7 @@ enum
    EXT_import_context_bit,
    EXT_framebuffer_sRGB_bit,
    EXT_fbconfig_packed_float_bit,
+   EXT_create_context_es_profile_bit,
    EXT_create_context_es2_profile_bit,
    MESA_copy_sub_buffer_bit,
    MESA_depth_float_bit,

src/loader/pci_id_driver_map.h

@@ -53,6 +53,12 @@ static const int radeonsi_chip_ids[] = {
 #undef CHIPSET
 };
 
+static const int virtio_gpu_chip_ids[] = {
+#define CHIPSET(chip, name, family) chip,
+#include "pci_ids/virtio_gpu_pci_ids.h"
+#undef CHIPSET
+};
+
 static const int vmwgfx_chip_ids[] = {
 #define CHIPSET(chip, name, family) chip,
 #include "pci_ids/vmwgfx_pci_ids.h"
@@ -78,6 +84,7 @@ static const struct {
    { 0x1002, "radeonsi", radeonsi_chip_ids, ARRAY_SIZE(radeonsi_chip_ids), _LOADER_GALLIUM},
    { 0x10de, "nouveau_vieux", NULL, -1,  _LOADER_DRI, is_nouveau_vieux },
    { 0x10de, "nouveau", NULL, -1,  _LOADER_GALLIUM },
+   { 0x1af4, "virtio_gpu", virtio_gpu_chip_ids, ARRAY_SIZE(virtio_gpu_chip_ids), _LOADER_GALLIUM },
    { 0x15ad, "vmwgfx", vmwgfx_chip_ids, ARRAY_SIZE(vmwgfx_chip_ids), _LOADER_GALLIUM },
    { 0x0000, NULL, NULL, 0 },
 };

src/mesa/Android.libmesa_dricore.mk

@@ -48,9 +48,8 @@ endif # x86
 endif # MESA_ENABLE_ASM
 
 ifeq ($(ARCH_X86_HAVE_SSE4_1),true)
-LOCAL_SRC_FILES += \
-	main/streaming-load-memcpy.c \
-	main/sse_minmax.c
+LOCAL_WHOLE_STATIC_LIBRARIES := \
+	libmesa_sse41
 LOCAL_CFLAGS := \
 	-msse4.1 \
        -DUSE_SSE41
@@ -64,7 +63,7 @@ LOCAL_C_INCLUDES := \
 	$(MESA_TOP)/src/gallium/include \
 	$(MESA_TOP)/src/gallium/auxiliary
 
-LOCAL_WHOLE_STATIC_LIBRARIES := \
+LOCAL_WHOLE_STATIC_LIBRARIES += \
 	libmesa_program
 
 include $(LOCAL_PATH)/Android.gen.mk

src/mesa/Android.libmesa_sse41.mk

@@ -0,0 +1,44 @@
+# Copyright 2012 Intel Corporation
+# Copyright (C) 2010-2011 Chia-I Wu <olvaffe@gmail.com>
+# Copyright (C) 2010-2011 LunarG Inc.
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the "Software"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included
+# in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+# DEALINGS IN THE SOFTWARE.
+
+ifeq ($(ARCH_X86_HAVE_SSE4_1),true)
+
+LOCAL_PATH := $(call my-dir)
+
+include $(LOCAL_PATH)/Makefile.sources
+
+include $(CLEAR_VARS)
+
+LOCAL_MODULE := libmesa_sse41
+
+LOCAL_SRC_FILES += \
+	$(X86_SSE41_FILES)
+
+LOCAL_C_INCLUDES := \
+	$(MESA_TOP)/src/mapi \
+	$(MESA_TOP)/src/gallium/include \
+	$(MESA_TOP)/src/gallium/auxiliary
+
+include $(MESA_COMMON_MK)
+include $(BUILD_STATIC_LIBRARY)
+
+endif

src/mesa/Android.libmesa_st_mesa.mk

@@ -47,6 +47,8 @@ endif # x86
 endif # MESA_ENABLE_ASM
 
 ifeq ($(ARCH_X86_HAVE_SSE4_1),true)
+LOCAL_WHOLE_STATIC_LIBRARIES := \
+	libmesa_sse41
 LOCAL_CFLAGS := \
        -DUSE_SSE41
 endif
@@ -59,7 +61,7 @@ LOCAL_C_INCLUDES := \
 	$(MESA_TOP)/src/gallium/auxiliary \
 	$(MESA_TOP)/src/gallium/include
 
-LOCAL_WHOLE_STATIC_LIBRARIES := \
+LOCAL_WHOLE_STATIC_LIBRARIES += \
 	libmesa_program
 
 include $(LOCAL_PATH)/Android.gen.mk

src/mesa/Android.mk

@@ -24,5 +24,6 @@ include $(LOCAL_PATH)/Android.mesa_gen_matypes.mk
 include $(LOCAL_PATH)/Android.libmesa_glsl_utils.mk
 include $(LOCAL_PATH)/Android.libmesa_dricore.mk
 include $(LOCAL_PATH)/Android.libmesa_st_mesa.mk
+include $(LOCAL_PATH)/Android.libmesa_sse41.mk
 
 include $(LOCAL_PATH)/program/Android.mk

src/mesa/Makefile.sources

@@ -578,6 +578,10 @@ X86_64_FILES =		\
 	x86-64/x86-64.h	\
 	x86-64/xform4.S
 
+X86_SSE41_FILES = \
+	main/streaming-load-memcpy.c \
+	main/sse_minmax.c
+
 SPARC_FILES =			\
 	sparc/sparc.h		\
 	sparc/sparc_clip.S	\

src/mesa/drivers/common/meta_copy_image.c

@@ -203,7 +203,7 @@ _mesa_meta_CopyImageSubData_uncompressed(struct gl_context *ctx,
          src_tex_image = wrap_renderbuffer(ctx, src_renderbuffer);
       }
       if (!make_view(ctx, src_tex_image, &src_view_tex_image, &src_view_texture,
-                     dst_tex_image->InternalFormat))
+                     dst_internal_format))
          goto cleanup;
    }
 

src/mesa/drivers/dri/Android.mk

@@ -28,9 +28,6 @@ include $(LOCAL_PATH)/common/Makefile.sources
 #-----------------------------------------------
 # Variables common to all DRI drivers
 
-MESA_DRI_MODULE_PATH := $(TARGET_OUT_SHARED_LIBRARIES)/dri
-MESA_DRI_MODULE_UNSTRIPPED_PATH := $(TARGET_OUT_SHARED_LIBRARIES_UNSTRIPPED)/dri
-
 MESA_DRI_CFLAGS := \
 	-DHAVE_ANDROID_PLATFORM
 

src/mesa/drivers/dri/common/Android.mk

@@ -74,20 +74,24 @@ $(intermediates)/xmlpool/%.po: $(LOCAL_PATH)/xmlpool/%.po $(POT)
 		sed -i -e 's/charset=.*\\n/charset=UTF-8\\n/' $@; \
 	fi
 
-$(intermediates)/xmlpool/%/LC_MESSAGES/options.mo: $(intermediates)/xmlpool/%.po
+PRIVATE_SCRIPT := $(LOCAL_PATH)/xmlpool/gen_xmlpool.py
+PRIVATE_LOCALEDIR := $(intermediates)/xmlpool
+PRIVATE_TEMPLATE_HEADER := $(LOCAL_PATH)/xmlpool/t_options.h
+PRIVATE_MO_FILES := $(MESA_DRI_OPTIONS_LANGS:%=$(intermediates)/xmlpool/%/LC_MESSAGES/options.mo)
+
+LOCAL_GENERATED_SOURCES += $(PRIVATE_MO_FILES)
+
+$(PRIVATE_MO_FILES): $(intermediates)/xmlpool/%/LC_MESSAGES/options.mo: $(intermediates)/xmlpool/%.po
 	mkdir -p $(dir $@)
 	msgfmt -o $@ $<
 
-$(MESA_DRI_OPTIONS_H): PRIVATE_SCRIPT := $(LOCAL_PATH)/xmlpool/gen_xmlpool.py
-$(MESA_DRI_OPTIONS_H): PRIVATE_LOCALEDIR := $(intermediates)/xmlpool
-$(MESA_DRI_OPTIONS_H): PRIVATE_TEMPLATE_HEADER := $(LOCAL_PATH)/xmlpool/t_options.h
-$(MESA_DRI_OPTIONS_H): PRIVATE_MO_FILES := $(MESA_DRI_OPTIONS_LANGS:%=$(intermediates)/xmlpool/%/LC_MESSAGES/options.mo)
-.SECONDEXPANSION:
-$(MESA_DRI_OPTIONS_H): $$(PRIVATE_SCRIPT) $$(PRIVATE_TEMPLATE_HEADER) $$(PRIVATE_MO_FILES)
-	@mkdir -p $(dir $@)
-	$(hide) $(MESA_PYTHON2) $(PRIVATE_SCRIPT) $(PRIVATE_TEMPLATE_HEADER) \
+$(LOCAL_GENERATED_SOURCES): PRIVATE_PYTHON := $(MESA_PYTHON2)
+$(LOCAL_GENERATED_SOURCES): PRIVATE_CUSTOM_TOOL = $(PRIVATE_PYTHON) $^ $(PRIVATE_TEMPLATE_HEADER) \
 		$(PRIVATE_LOCALEDIR) $(MESA_DRI_OPTIONS_LANGS) > $@
 
+$(MESA_DRI_OPTIONS_H): $(PRIVATE_SCRIPT) $(PRIVATE_TEMPLATE_HEADER) $(PRIVATE_MO_FILES)
+	$(transform-generated-source)
+
 include $(MESA_COMMON_MK)
 include $(BUILD_STATIC_LIBRARY)
 

src/mesa/drivers/dri/i915/Android.mk

@@ -27,7 +27,7 @@ include $(CLEAR_VARS)
 
 LOCAL_MODULE := i915_dri
 ifeq ($(MESA_LOLLIPOP_BUILD),true)
-LOCAL_MODULE_RELATIVE_PATH := $(notdir $(MESA_DRI_MODULE_PATH))
+LOCAL_MODULE_RELATIVE_PATH := $(MESA_DRI_MODULE_REL_PATH)
 else
 LOCAL_MODULE_PATH := $(MESA_DRI_MODULE_PATH)
 LOCAL_UNSTRIPPED_PATH := $(MESA_DRI_MODULE_UNSTRIPPED_PATH)

src/mesa/drivers/dri/i965/Android.mk

@@ -27,7 +27,7 @@ include $(CLEAR_VARS)
 
 LOCAL_MODULE := i965_dri
 ifeq ($(MESA_LOLLIPOP_BUILD),true)
-LOCAL_MODULE_RELATIVE_PATH := $(notdir $(MESA_DRI_MODULE_PATH))
+LOCAL_MODULE_RELATIVE_PATH := $(MESA_DRI_MODULE_REL_PATH)
 else
 LOCAL_MODULE_PATH := $(MESA_DRI_MODULE_PATH)
 LOCAL_UNSTRIPPED_PATH := $(MESA_DRI_MODULE_UNSTRIPPED_PATH)

src/mesa/drivers/dri/i965/brw_blorp.cpp

@@ -322,8 +322,14 @@ brw_hiz_op_params::brw_hiz_op_params(struct intel_mipmap_tree *mt,
     * not 8. But commit 1f112cc increased the alignment from 4 to 8, which
     * prevents the clobbering.
     */
-   depth.width = ALIGN(depth.width, 8);
-   depth.height = ALIGN(depth.height, 4);
+   dst.num_samples = mt->num_samples;
+   if (dst.num_samples > 1) {
+      depth.width = ALIGN(mt->logical_width0, 8);
+      depth.height = ALIGN(mt->logical_height0, 4);
+   } else {
+      depth.width = ALIGN(depth.width, 8);
+      depth.height = ALIGN(depth.height, 4);
+   }
 
    x1 = depth.width;
    y1 = depth.height;

src/mesa/drivers/dri/i965/brw_fs.cpp

@@ -2374,8 +2374,10 @@ fs_visitor::opt_sampler_eot()
     * we have enough space, but it will make sure the dead code eliminator kills
     * the instruction that this will replace.
     */
-   if (tex_inst->header_size != 0)
+   if (tex_inst->header_size != 0) {
+      invalidate_live_intervals();
       return true;
+   }
 
    fs_reg send_header = ibld.vgrf(BRW_REGISTER_TYPE_F,
                                   load_payload->sources + 1);
@@ -2406,6 +2408,7 @@ fs_visitor::opt_sampler_eot()
    tex_inst->insert_before(cfg->blocks[cfg->num_blocks - 1], new_load_payload);
    tex_inst->src[0] = send_header;
 
+   invalidate_live_intervals();
    return true;
 }
 
@@ -5056,12 +5059,18 @@ fs_visitor::optimize()
 void
 fs_visitor::fixup_3src_null_dest()
 {
+   bool progress = false;
+
    foreach_block_and_inst_safe (block, fs_inst, inst, cfg) {
       if (inst->is_3src() && inst->dst.is_null()) {
          inst->dst = fs_reg(VGRF, alloc.allocate(dispatch_width / 8),
                             inst->dst.type);
+         progress = true;
       }
    }
+
+   if (progress)
+      invalidate_live_intervals();
 }
 
 void

src/mesa/drivers/dri/i965/brw_fs_cse.cpp

@@ -137,6 +137,8 @@ operands_match(const fs_inst *a, const fs_inst *b, bool *negate)
       ys[1].f = ys1_imm;
 
       *negate = (xs0_negate != xs1_negate) != (ys0_negate != ys1_negate);
+      if (*negate && (a->saturate || b->saturate))
+         return false;
       return ret;
    } else if (!a->is_commutative()) {
       bool match = true;

src/mesa/drivers/dri/i965/brw_vec4.cpp

@@ -1002,6 +1002,7 @@ vec4_visitor::opt_register_coalesce()
 
          if (is_nop_mov) {
             inst->remove(block);
+            progress = true;
             continue;
          }
       }

src/mesa/drivers/dri/i965/brw_vec4_copy_propagation.cpp

@@ -256,8 +256,8 @@ try_constant_propagate(const struct brw_device_info *devinfo,
 
 static bool
 try_copy_propagate(const struct brw_device_info *devinfo,
-                   vec4_instruction *inst,
-                   int arg, struct copy_entry *entry)
+                   vec4_instruction *inst, int arg,
+                   struct copy_entry *entry, int attributes_per_reg)
 {
    /* Build up the value we are propagating as if it were the source of a
     * single MOV
@@ -322,7 +322,8 @@ try_copy_propagate(const struct brw_device_info *devinfo,
    unsigned composed_swizzle = brw_compose_swizzle(inst->src[arg].swizzle,
                                                    value.swizzle);
    if (inst->is_3src() &&
-       value.file == UNIFORM &&
+       (value.file == UNIFORM ||
+        (value.file == ATTR && attributes_per_reg != 1)) &&
        !brw_is_single_value_swizzle(composed_swizzle))
       return false;
 
@@ -397,6 +398,11 @@ try_copy_propagate(const struct brw_device_info *devinfo,
 bool
 vec4_visitor::opt_copy_propagation(bool do_constant_prop)
 {
+   /* If we are in dual instanced or single mode, then attributes are going
+    * to be interleaved, so one register contains two attribute slots.
+    */
+   const int attributes_per_reg =
+      prog_data->dispatch_mode == DISPATCH_MODE_4X2_DUAL_OBJECT ? 1 : 2;
    bool progress = false;
    struct copy_entry entries[alloc.total_size];
 
@@ -467,7 +473,7 @@ vec4_visitor::opt_copy_propagation(bool do_constant_prop)
          if (do_constant_prop && try_constant_propagate(devinfo, inst, i, &entry))
             progress = true;
 
-	 if (try_copy_propagate(devinfo, inst, i, &entry))
+         if (try_copy_propagate(devinfo, inst, i, &entry, attributes_per_reg))
 	    progress = true;
       }
 

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

@@ -1320,6 +1320,11 @@ brw_upload_image_surfaces(struct brw_context *brw,
       }
 
       brw->ctx.NewDriverState |= BRW_NEW_SURFACES;
+      /* This may have changed the image metadata dependent on the context
+       * image unit state and passed to the program as uniforms, make sure
+       * that push and pull constants are reuploaded.
+       */
+      brw->NewGLState |= _NEW_PROGRAM_CONSTANTS;
    }
 }
 

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

@@ -140,9 +140,9 @@ copy_image_with_memcpy(struct brw_context *brw,
    _mesa_get_format_block_size(src_mt->format, &src_bw, &src_bh);
 
    assert(src_width % src_bw == 0);
-   assert(src_height % src_bw == 0);
+   assert(src_height % src_bh == 0);
    assert(src_x % src_bw == 0);
-   assert(src_y % src_bw == 0);
+   assert(src_y % src_bh == 0);
 
    /* If we are on the same miptree, same level, and same slice, then
     * intel_miptree_map won't let us map it twice.  We have to do things a
@@ -153,7 +153,7 @@ copy_image_with_memcpy(struct brw_context *brw,
 
    if (same_slice) {
       assert(dst_x % src_bw == 0);
-      assert(dst_y % src_bw == 0);
+      assert(dst_y % src_bh == 0);
 
       map_x1 = MIN2(src_x, dst_x);
       map_y1 = MIN2(src_y, dst_y);
@@ -212,6 +212,7 @@ intel_copy_image_sub_data(struct gl_context *ctx,
    struct brw_context *brw = brw_context(ctx);
    struct intel_mipmap_tree *src_mt, *dst_mt;
    unsigned src_level, dst_level;
+   GLuint bw, bh;
 
    if (_mesa_meta_CopyImageSubData_uncompressed(ctx,
                                                 src_image, src_renderbuffer,
@@ -275,6 +276,19 @@ intel_copy_image_sub_data(struct gl_context *ctx,
    intel_miptree_all_slices_resolve_depth(brw, dst_mt);
    intel_miptree_resolve_color(brw, dst_mt);
 
+   _mesa_get_format_block_size(src_mt->format, &bw, &bh);
+
+   /* It's legal to have a WxH that's smaller than a compressed block. This
+    * happens for example when you are using a higher level LOD. For this case,
+    * we still want to copy the entire block, or else the decompression will be
+    * incorrect.
+    */
+   if (src_width < bw)
+      src_width = ALIGN_NPOT(src_width, bw);
+
+   if (src_height < bh)
+      src_height = ALIGN_NPOT(src_height, bh);
+
    if (copy_image_with_blitter(brw, src_mt, src_level,
                                src_x, src_y, src_z,
                                dst_mt, dst_level,

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

@@ -51,7 +51,7 @@ intel_miptree_create_for_teximage(struct brw_context *brw,
       width <<= 1;
       if (height != 1)
          height <<= 1;
-      if (depth != 1)
+      if (intelObj->base.Target == GL_TEXTURE_3D)
          depth <<= 1;
    }
 

src/mesa/drivers/dri/i965/test_vec4_cmod_propagation.cpp

@@ -40,6 +40,7 @@ public:
    struct gl_context *ctx;
    struct gl_shader_program *shader_prog;
    struct brw_vertex_program *vp;
+   struct brw_vue_prog_data *prog_data;
    vec4_visitor *v;
 };
 
@@ -47,9 +48,13 @@ class cmod_propagation_vec4_visitor : public vec4_visitor
 {
 public:
    cmod_propagation_vec4_visitor(struct brw_compiler *compiler,
-                                 nir_shader *shader)
-      : vec4_visitor(compiler, NULL, NULL, NULL, shader, NULL,
-                     false, -1) {}
+                                 nir_shader *shader,
+                                 struct brw_vue_prog_data *prog_data)
+      : vec4_visitor(compiler, NULL, NULL, prog_data, shader, NULL,
+                     false, -1)
+      {
+         prog_data->dispatch_mode = DISPATCH_MODE_4X2_DUAL_OBJECT;
+      }
 
 protected:
    /* Dummy implementation for pure virtual methods */
@@ -96,13 +101,14 @@ void cmod_propagation_test::SetUp()
    ctx = (struct gl_context *)calloc(1, sizeof(*ctx));
    compiler = (struct brw_compiler *)calloc(1, sizeof(*compiler));
    devinfo = (struct brw_device_info *)calloc(1, sizeof(*devinfo));
+   prog_data = (struct brw_vue_prog_data *)calloc(1, sizeof(*prog_data));
    compiler->devinfo = devinfo;
 
    vp = ralloc(NULL, struct brw_vertex_program);
 
    nir_shader *shader = nir_shader_create(NULL, MESA_SHADER_VERTEX, NULL);
 
-   v = new cmod_propagation_vec4_visitor(compiler, shader);
+   v = new cmod_propagation_vec4_visitor(compiler, shader, prog_data);
 
    _mesa_init_gl_program(&vp->program.Base, GL_VERTEX_SHADER, 0);
 

src/mesa/drivers/dri/i965/test_vec4_copy_propagation.cpp

@@ -38,6 +38,7 @@ public:
    struct gl_context *ctx;
    struct gl_shader_program *shader_prog;
    struct brw_vertex_program *vp;
+   struct brw_vue_prog_data *prog_data;
    vec4_visitor *v;
 };
 
@@ -45,10 +46,12 @@ class copy_propagation_vec4_visitor : public vec4_visitor
 {
 public:
    copy_propagation_vec4_visitor(struct brw_compiler *compiler,
-                                 nir_shader *shader)
-      : vec4_visitor(compiler, NULL, NULL, NULL, shader, NULL,
+                                 nir_shader *shader,
+                                 struct brw_vue_prog_data *prog_data)
+      : vec4_visitor(compiler, NULL, NULL, prog_data, shader, NULL,
                      false /* no_spills */, -1)
    {
+      prog_data->dispatch_mode = DISPATCH_MODE_4X2_DUAL_OBJECT;
    }
 
 protected:
@@ -90,13 +93,14 @@ void copy_propagation_test::SetUp()
    ctx = (struct gl_context *)calloc(1, sizeof(*ctx));
    compiler = (struct brw_compiler *)calloc(1, sizeof(*compiler));
    devinfo = (struct brw_device_info *)calloc(1, sizeof(*devinfo));
+   prog_data = (struct brw_vue_prog_data *)calloc(1, sizeof(*prog_data));
    compiler->devinfo = devinfo;
 
    vp = ralloc(NULL, struct brw_vertex_program);
 
    nir_shader *shader = nir_shader_create(NULL, MESA_SHADER_VERTEX, NULL);
 
-   v = new copy_propagation_vec4_visitor(compiler, shader);
+   v = new copy_propagation_vec4_visitor(compiler, shader, prog_data);
 
    _mesa_init_gl_program(&vp->program.Base, GL_VERTEX_SHADER, 0);
 

src/mesa/drivers/dri/i965/test_vec4_register_coalesce.cpp

@@ -40,6 +40,7 @@ public:
    struct gl_context *ctx;
    struct gl_shader_program *shader_prog;
    struct brw_vertex_program *vp;
+   struct brw_vue_prog_data *prog_data;
    vec4_visitor *v;
 };
 
@@ -48,10 +49,12 @@ class register_coalesce_vec4_visitor : public vec4_visitor
 {
 public:
    register_coalesce_vec4_visitor(struct brw_compiler *compiler,
-                                  nir_shader *shader)
-      : vec4_visitor(compiler, NULL, NULL, NULL, shader, NULL,
+                                  nir_shader *shader,
+                                  struct brw_vue_prog_data *prog_data)
+      : vec4_visitor(compiler, NULL, NULL, prog_data, shader, NULL,
                      false /* no_spills */, -1)
    {
+      prog_data->dispatch_mode = DISPATCH_MODE_4X2_DUAL_OBJECT;
    }
 
 protected:
@@ -93,13 +96,14 @@ void register_coalesce_test::SetUp()
    ctx = (struct gl_context *)calloc(1, sizeof(*ctx));
    compiler = (struct brw_compiler *)calloc(1, sizeof(*compiler));
    devinfo = (struct brw_device_info *)calloc(1, sizeof(*devinfo));
+   prog_data = (struct brw_vue_prog_data *)calloc(1, sizeof(*prog_data));
    compiler->devinfo = devinfo;
 
    vp = ralloc(NULL, struct brw_vertex_program);
 
    nir_shader *shader = nir_shader_create(NULL, MESA_SHADER_VERTEX, NULL);
 
-   v = new register_coalesce_vec4_visitor(compiler, shader);
+   v = new register_coalesce_vec4_visitor(compiler, shader, prog_data);
 
    _mesa_init_gl_program(&vp->program.Base, GL_VERTEX_SHADER, 0);
 

src/mesa/main/errors.c

@@ -758,15 +758,11 @@ _mesa_set_debug_state_int(struct gl_context *ctx, GLenum pname, GLint val)
 GLint
 _mesa_get_debug_state_int(struct gl_context *ctx, GLenum pname)
 {
-   struct gl_debug_state *debug;
    GLint val;
 
-   mtx_lock(&ctx->DebugMutex);
-   debug = ctx->Debug;
-   if (!debug) {
-      mtx_unlock(&ctx->DebugMutex);
+   struct gl_debug_state *debug = _mesa_lock_debug_state(ctx);
+   if (!debug)
       return 0;
-   }
 
    switch (pname) {
    case GL_DEBUG_OUTPUT:
@@ -791,7 +787,7 @@ _mesa_get_debug_state_int(struct gl_context *ctx, GLenum pname)
       break;
    }
 
-   mtx_unlock(&ctx->DebugMutex);
+   _mesa_unlock_debug_state(ctx);
 
    return val;
 }
@@ -803,15 +799,11 @@ _mesa_get_debug_state_int(struct gl_context *ctx, GLenum pname)
 void *
 _mesa_get_debug_state_ptr(struct gl_context *ctx, GLenum pname)
 {
-   struct gl_debug_state *debug;
    void *val;
+   struct gl_debug_state *debug = _mesa_lock_debug_state(ctx);
 
-   mtx_lock(&ctx->DebugMutex);
-   debug = ctx->Debug;
-   if (!debug) {
-      mtx_unlock(&ctx->DebugMutex);
+   if (!debug)
       return NULL;
-   }
 
    switch (pname) {
    case GL_DEBUG_CALLBACK_FUNCTION_ARB:
@@ -826,7 +818,7 @@ _mesa_get_debug_state_ptr(struct gl_context *ctx, GLenum pname)
       break;
    }
 
-   mtx_unlock(&ctx->DebugMutex);
+   _mesa_unlock_debug_state(ctx);
 
    return val;
 }

src/mesa/main/fbobject.c

@@ -2795,6 +2795,7 @@ reuse_framebuffer_texture_attachment(struct gl_framebuffer *fb,
    dst_att->Complete = src_att->Complete;
    dst_att->TextureLevel = src_att->TextureLevel;
    dst_att->Zoffset = src_att->Zoffset;
+   dst_att->Layered = src_att->Layered;
 }
 
 

src/mesa/main/get.c

@@ -1022,6 +1022,8 @@ find_custom_value(struct gl_context *ctx, const struct value_desc *d, union valu
       }
       break;
    /* GL_KHR_DEBUG */
+   case GL_DEBUG_OUTPUT:
+   case GL_DEBUG_OUTPUT_SYNCHRONOUS:
    case GL_DEBUG_LOGGED_MESSAGES:
    case GL_DEBUG_NEXT_LOGGED_MESSAGE_LENGTH:
    case GL_DEBUG_GROUP_STACK_DEPTH:

src/mesa/main/get_hash_params.py

@@ -126,6 +126,8 @@ descriptor=[
   [ "MAX_TEXTURE_MAX_ANISOTROPY_EXT", "CONTEXT_FLOAT(Const.MaxTextureMaxAnisotropy), extra_EXT_texture_filter_anisotropic" ],
 
 # GL_KHR_debug (GL 4.3)/ GL_ARB_debug_output
+  [ "DEBUG_OUTPUT", "LOC_CUSTOM, TYPE_BOOLEAN, 0, NO_EXTRA" ],
+  [ "DEBUG_OUTPUT_SYNCHRONOUS", "LOC_CUSTOM, TYPE_BOOLEAN, 0, NO_EXTRA" ],
   [ "DEBUG_LOGGED_MESSAGES", "LOC_CUSTOM, TYPE_INT, 0, NO_EXTRA" ],
   [ "DEBUG_NEXT_LOGGED_MESSAGE_LENGTH", "LOC_CUSTOM, TYPE_INT, 0, NO_EXTRA" ],
   [ "MAX_DEBUG_LOGGED_MESSAGES", "CONST(MAX_DEBUG_LOGGED_MESSAGES), NO_EXTRA" ],
@@ -742,6 +744,7 @@ descriptor=[
   [ "DEPTH_CLAMP", "CONTEXT_BOOL(Transform.DepthClamp), extra_ARB_depth_clamp" ],
 
 # GL_ATI_fragment_shader
+  [ "FRAGMENT_SHADER_ATI", "CONTEXT_BOOL(ATIFragmentShader.Enabled), extra_ATI_fragment_shader" ],
   [ "NUM_FRAGMENT_REGISTERS_ATI", "CONST(6), extra_ATI_fragment_shader" ],
   [ "NUM_FRAGMENT_CONSTANTS_ATI", "CONST(8), extra_ATI_fragment_shader" ],
   [ "NUM_PASSES_ATI", "CONST(2), extra_ATI_fragment_shader" ],

src/mesa/main/image.c

@@ -670,7 +670,7 @@ _mesa_clip_drawpixels(const struct gl_context *ctx,
  * so that the image region is entirely within the window bounds.
  * Note: this is different from _mesa_clip_drawpixels() in that the
  * scissor box is ignored, and we use the bounds of the current readbuffer
- * surface.
+ * surface or the attached image.
  *
  * \return  GL_TRUE if region to read is in bounds
  *          GL_FALSE if region is completely out of bounds (nothing to read)
@@ -682,6 +682,18 @@ _mesa_clip_readpixels(const struct gl_context *ctx,
                       struct gl_pixelstore_attrib *pack)
 {
    const struct gl_framebuffer *buffer = ctx->ReadBuffer;
+   struct gl_renderbuffer *rb = buffer->_ColorReadBuffer;
+   GLsizei clip_width;
+   GLsizei clip_height;
+
+   if (rb) {
+      clip_width = rb->Width;
+      clip_height = rb->Height;
+   } else {
+      clip_width = buffer->Width;
+      clip_height = buffer->Height;
+   }
+
 
    if (pack->RowLength == 0) {
       pack->RowLength = *width;
@@ -694,8 +706,8 @@ _mesa_clip_readpixels(const struct gl_context *ctx,
       *srcX = 0;
    }
    /* right clipping */
-   if (*srcX + *width > (GLsizei) buffer->Width)
-      *width -= (*srcX + *width - buffer->Width);
+   if (*srcX + *width > clip_width)
+      *width -= (*srcX + *width - clip_width);
 
    if (*width <= 0)
       return GL_FALSE;
@@ -707,8 +719,8 @@ _mesa_clip_readpixels(const struct gl_context *ctx,
       *srcY = 0;
    }
    /* top clipping */
-   if (*srcY + *height > (GLsizei) buffer->Height)
-      *height -= (*srcY + *height - buffer->Height);
+   if (*srcY + *height > clip_height)
+      *height -= (*srcY + *height - clip_height);
 
    if (*height <= 0)
       return GL_FALSE;

src/mesa/main/objectlabel.c

@@ -288,16 +288,18 @@ void GLAPIENTRY
 _mesa_ObjectPtrLabel(const void *ptr, GLsizei length, const GLchar *label)
 {
    GET_CURRENT_CONTEXT(ctx);
-   struct gl_sync_object *const syncObj = (struct gl_sync_object *) ptr;
+   struct gl_sync_object *syncObj;
    const char *callerstr;
    char **labelPtr;
 
+   syncObj = _mesa_get_and_ref_sync(ctx, (void*)ptr, true);
+
    if (_mesa_is_desktop_gl(ctx))
       callerstr = "glObjectPtrLabel";
    else
       callerstr = "glObjectPtrLabelKHR";
 
-   if (!_mesa_validate_sync(ctx, syncObj)) {
+   if (!syncObj) {
       _mesa_error(ctx, GL_INVALID_VALUE, "%s (not a valid sync object)",
                   callerstr);
       return;
@@ -306,6 +308,7 @@ _mesa_ObjectPtrLabel(const void *ptr, GLsizei length, const GLchar *label)
    labelPtr = &syncObj->Label;
 
    set_label(ctx, labelPtr, label, length, callerstr);
+   _mesa_unref_sync_object(ctx, syncObj, 1);
 }
 
 void GLAPIENTRY
@@ -313,7 +316,7 @@ _mesa_GetObjectPtrLabel(const void *ptr, GLsizei bufSize, GLsizei *length,
                         GLchar *label)
 {
    GET_CURRENT_CONTEXT(ctx);
-   struct gl_sync_object *const syncObj = (struct gl_sync_object *) ptr;
+   struct gl_sync_object *syncObj;
    const char *callerstr;
    char **labelPtr;
 
@@ -328,7 +331,8 @@ _mesa_GetObjectPtrLabel(const void *ptr, GLsizei bufSize, GLsizei *length,
       return;
    }
 
-   if (!_mesa_validate_sync(ctx, syncObj)) {
+   syncObj = _mesa_get_and_ref_sync(ctx, (void*)ptr, true);
+   if (!syncObj) {
       _mesa_error(ctx, GL_INVALID_VALUE, "%s (not a valid sync object)",
                   callerstr);
       return;
@@ -337,4 +341,5 @@ _mesa_GetObjectPtrLabel(const void *ptr, GLsizei bufSize, GLsizei *length,
    labelPtr = &syncObj->Label;
 
    copy_label(*labelPtr, label, length, bufSize);
+   _mesa_unref_sync_object(ctx, syncObj, 1);
 }

src/mesa/main/readpix.c

@@ -993,6 +993,7 @@ _mesa_ReadnPixelsARB( GLint x, GLint y, GLsizei width, GLsizei height,
 {
    GLenum err = GL_NO_ERROR;
    struct gl_renderbuffer *rb;
+   struct gl_pixelstore_attrib clippedPacking;
 
    GET_CURRENT_CONTEXT(ctx);
 
@@ -1094,7 +1095,9 @@ _mesa_ReadnPixelsARB( GLint x, GLint y, GLsizei width, GLsizei height,
       }
    }
 
-   if (width == 0 || height == 0)
+   /* Do all needed clipping here, so that we can forget about it later */
+   clippedPacking = ctx->Pack;
+   if (!_mesa_clip_readpixels(ctx, &x, &y, &width, &height, &clippedPacking))
       return; /* nothing to do */
 
    if (!_mesa_validate_pbo_access(2, &ctx->Pack, width, height, 1,
@@ -1118,7 +1121,7 @@ _mesa_ReadnPixelsARB( GLint x, GLint y, GLsizei width, GLsizei height,
    }
 
    ctx->Driver.ReadPixels(ctx, x, y, width, height,
-			  format, type, &ctx->Pack, pixels);
+                          format, type, &clippedPacking, pixels);
 }
 
 void GLAPIENTRY