docs: Add release notes for 10.0.5

In preparation for the 10.0.5 release, of course.

VERSION

@@ -1 +1 @@
-10.0.2
+10.0.5

bin/.cherry-ignore

@@ -11,3 +11,29 @@ f0f202e6b764be803470e27cba9102f14361ae22 glx: conditionaly build dri3 and presen
 # the 10.0 branch.  See
 # http://lists.freedesktop.org/archives/mesa-stable/2013-December/000500.html
 a057b837ddd1c725a7504eedc53c6df05a012773 egl: add HAVE_LIBDRM define, fix EGL X11 platform
+
+# Author requested skipping due to regressions
+# Picking it would require at least also picking:
+# 73c3c7e3, 3e0e9e3b, c59a605c
+b2d1c579bb84a88179072a6a783f8827e218db55 glcpp: Set extension defines after resolving the GLSL version.
+
+# These patches depend on other code not in stable branch.
+# (at least 3b22146dc714b6090f7423abbc4df53d7d1fdaa9)
+e190709119d8eb85c67bfbad5be699d39ad0118e mesa: Ensure that transform feedback refers to the correct program.
+43e77215b13b2f86e461cd8a62b542fc6854dd1c i965/gen7: Use to the correct program when uploading transform feedback state.
+
+# Author requested to ignore these four (since they depend on commits not in
+# stable).
+3313cc269bd428ca96a132d86da5fddc0f27386a i965: Add an option to ignore sample qualifier
+a92e5f7cf63d496ad7830b5cea4bbab287c25b8e i965: Use sample barycentric coordinates with per sample shading
+f5cfb4ae21df8eebfc6b86c0ce858b1c0a9160dd i965: Ignore 'centroid' interpolation qualifier in case of persample shading
+dc2f94bc786768329973403248820a2e5249f102 i965: Ignore 'centroid' interpolation qualifier in case of persample shading
+
+# This depends on the clear_buffer_object extensions work which is not in 10.0
+# (See commit 5f7bc0c75904a40da0973329badea8497e53a26a on other branches)
+aff7c5e78ab133866a90f67613508735c9b75094
+
+# These patches are fixing code not present in 10.0
+f34d75d6f69f4c0bf391e0adf1fd469601b01b04
+e8d85034dad37177fce780ee3e09501e60be6e81
+a61d859519d520b849c11ad5c1c1972870abd956

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='^\([[:space:]]*NOTE: .*[Cc]andidate.*10\.0\|CC:.*10\.0.*mesa-stable\)' HEAD..origin/master |\
 while read sha
 do
 	# Check to see whether the patch is on the ignore list.

common.py

@@ -100,4 +100,4 @@ def AddOptions(opts):
 	opts.Add(BoolOption('quiet', 'DEPRECATED: profile build', 'yes'))
 	opts.Add(BoolOption('texture_float', 'enable floating-point textures and renderbuffers', 'no'))
 	if host_platform == 'windows':
-		opts.Add(EnumOption('MSVC_VERSION', 'MS Visual C++ version', None, allowed_values=('7.1', '8.0', '9.0', '10.0', '11.0')))
+		opts.Add(EnumOption('MSVC_VERSION', 'MS Visual C++ version', None, allowed_values=('7.1', '8.0', '9.0', '10.0', '11.0', '12.0')))

configure.ac

@@ -227,6 +227,12 @@ esac
 AC_SUBST([VISIBILITY_CFLAGS])
 AC_SUBST([VISIBILITY_CXXFLAGS])
 
+dnl
+dnl Optional flags, check for compiler support
+dnl
+AX_CHECK_COMPILE_FLAG([-msse4.1], [SSE41_SUPPORTED=1], [SSE41_SUPPORTED=0])
+AM_CONDITIONAL([SSE41_SUPPORTED], [test x$SSE41_SUPPORTED = x1])
+
 dnl
 dnl Hacks to enable 32 or 64 bit build
 dnl
@@ -993,7 +999,7 @@ if test "x$enable_dri" = xyes; then
     DRI_DIRS=`echo "$DRI_DIRS" | $SED 's/  */ /g'`
 
     # Check for expat
-    PKG_CHECK_EXISTS([EXPAT], [have_expat=yes], [have_expat=no])
+    PKG_CHECK_EXISTS([expat], [have_expat=yes], [have_expat=no])
     if test "x$have_expat" = "xyes"; then
        PKG_CHECK_MODULES([EXPAT], [expat], [],
                          AC_MSG_ERROR([Expat required for DRI.]))
@@ -1527,9 +1533,9 @@ if test "x$enable_gallium_llvm" = xauto; then
 fi
 if test "x$enable_gallium_llvm" = xyes; then
     if test "x$llvm_prefix" != x; then
-        AC_PATH_PROG([LLVM_CONFIG], [llvm-config], [no], ["$llvm_prefix/bin"])
+        AC_PATH_TOOL([LLVM_CONFIG], [llvm-config], [no], ["$llvm_prefix/bin"])
     else
-        AC_PATH_PROG([LLVM_CONFIG], [llvm-config], [no])
+        AC_PATH_TOOL([LLVM_CONFIG], [llvm-config], [no])
     fi
 
     if test "x$LLVM_CONFIG" != xno; then

docs/relnotes/10.0.2.html

@@ -31,6 +31,9 @@ because compatibility contexts not supported.
 
 <h2>MD5 checksums</h2>
 <pre>
+de7d14baf0101b697c140d2f47ef27e9  MesaLib-10.0.2.tar.gz
+8544c0ab3e438a08b5103421ea15b6d2  MesaLib-10.0.2.tar.bz2
+181b0d6c1afca38e98a930d0e564ed90  MesaLib-10.0.2.zip
 </pre>
 
 

docs/relnotes/10.0.3.html

@@ -0,0 +1,206 @@
+<!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 10.0.3 Release Notes / (February 3, 2014)</h1>
+
+<p>
+Mesa 10.0.3 is a bug fix release which fixes bugs found since the 10.0.2 release.
+</p>
+<p>
+Mesa 10.0.3 implements the OpenGL 3.3 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 3.3.  OpenGL
+3.3 is <strong>only</strong> available if requested at context creation
+because compatibility contexts not supported.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+5f9f463ef08129f6762106b434910adb  MesaLib-10.0.3.tar.bz2
+fb3997b6500e153bc32370cb3fc4ca9e  MesaLib-10.0.3.tar.gz
+a07b4b6b9eb449b88a6cb5061e51c331  MesaLib-10.0.3.zip
+</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=72708">Bug 72708</a> - Master fails to build with older gcc due to -msse4.1</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=72926">Bug 72926</a> - [REGRESSION,swrast] Memory-related crash with anti-aliasing enabled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73096">Bug 73096</a> - Query GL_RGBA_SIGNED_COMPONENTS_EXT missing</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73100">Bug 73100</a> - Please use AC_PATH_TOOL instead of AC_PATH_PROG for llvm-config</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73418">Bug 73418</a> - OpenCL hangs graphics on CAYMAN</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73473">Bug 73473</a> - Potential crash bug in src/gallium/auxiliary/rtasm/rtasm_execmem.c</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73915">Bug 73915</a> - sample shading + centroid broken since f5cfb4a</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73956">Bug 73956</a> - SIGSEGV when passing GL_NONE to glReadBuffer</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74026">Bug 74026</a> - Compiler rejects chained assignments involving array dereferences</li>
+
+</ul>
+
+<h2>Changes</h2>
+
+<p>The full set of changes can be viewed by using the following git command:</p>
+
+<pre>
+  git log mesa-10.0.2..mesa-10.0.3
+</pre>
+
+<p>Aaron Watry (2):</p>
+<ul>
+  <li>radeon: Move gfx/dma cs cleanup to r600_common_context_cleanup</li>
+  <li>st/dri: prevent leak of dri option default values</li>
+</ul>
+
+<p>Andreas Fänger (1):</p>
+<ul>
+  <li>swrast: fix delayed texel buffer allocation regression for OpenMP</li>
+</ul>
+
+<p>Anuj Phogat (3):</p>
+<ul>
+  <li>glsl: Disable ARB_texture_rectangle in shader version 100.</li>
+  <li>i965: Use sample barycentric coordinates with per sample shading</li>
+  <li>i965: Ignore 'centroid' interpolation qualifier in case of persample shading</li>
+</ul>
+
+<p>Brian Paul (3):</p>
+<ul>
+  <li>mesa: implement missing glGet(GL_RGBA_SIGNED_COMPONENTS_EXT) query</li>
+  <li>st/mesa: fix glReadBuffer(GL_NONE) segfault</li>
+  <li>draw: fix incorrect vertex size computation in LLVM drawing code</li>
+</ul>
+
+<p>Carl Worth (5):</p>
+<ul>
+  <li>Add md5sums for 10.0.2. release.</li>
+  <li>cherry-ignore: Ignore several patches not yet ready for the stable branch</li>
+  <li>Drop another couple of patches.</li>
+  <li>cherry-ignore: Ignore 4 patches at teh request of the author, (Anuj).</li>
+  <li>Update version to 10.0.3</li>
+</ul>
+
+<p>Chad Versace (1):</p>
+<ul>
+  <li>i965/gen6/blorp: Emit more flushes to workaround hangs</li>
+</ul>
+
+<p>Chris Forbes (1):</p>
+<ul>
+  <li>i965: fold offset into coord for textureOffset(gsampler2DRect)</li>
+</ul>
+
+<p>Emil Velikov (5):</p>
+<ul>
+  <li>mesa: use signed temporary variable to store _ColorDrawBufferIndexes</li>
+  <li>st/mesa: use signed temporary variable to store _ColorDrawBufferIndexes</li>
+  <li>nv50: access only the available amount of textures</li>
+  <li>nv50: access only the available amount of constbuf</li>
+  <li>gallium/rtasm: handle mmap failures appropriately</li>
+</ul>
+
+<p>Eric Anholt (2):</p>
+<ul>
+  <li>i965: Fix handling of MESA_pack_invert in blit (PBO) readpixels.</li>
+  <li>i965: Don't do the temporary-and-blit-copy for INVALIDATE_RANGE maps.</li>
+</ul>
+
+<p>Ian Romanick (2):</p>
+<ul>
+  <li>mesa: Add COMPRESSED_RGBA_S3TC_DXT1_EXT to COMPRESSED_TEXTURE_FORMATS for GLES</li>
+  <li>radeon / r200: Pass the API into _mesa_initialize_context</li>
+</ul>
+
+<p>Ilia Mirkin (2):</p>
+<ul>
+  <li>mesa: fix GL_COLOR_SUM enum for drivers without ARB_vertex_program</li>
+  <li>st/vdpau: don't return a device if the screen doesn't support NPOT</li>
+</ul>
+
+<p>José Fonseca (1):</p>
+<ul>
+  <li>mesa: Use IROUND instead of roundf.</li>
+</ul>
+
+<p>Kenneth Graunke (2):</p>
+<ul>
+  <li>glsl: Rename "expr" to "lhs_expr" in vector_extract munging code.</li>
+  <li>glsl: Fix chained assignments of vector channels.</li>
+</ul>
+
+<p>Lauri Kasanen (1):</p>
+<ul>
+  <li>mesa: Fix build to properly check for supported compiler flags</li>
+</ul>
+
+<p>Marek Olšák (2):</p>
+<ul>
+  <li>st/mesa: use sRGB formats for MSAA resolving if destination is sRGB</li>
+  <li>gallium/util: util_format_srgb should not return FORMAT_NONE for sRGB formats</li>
+</ul>
+
+<p>Matt Turner (2):</p>
+<ul>
+  <li>glcpp: Define GL_EXT_shader_integer_mix in both GL and ES.</li>
+  <li>glx: Update glxext.h to revision 24777.</li>
+</ul>
+
+<p>Michał Górny (1):</p>
+<ul>
+  <li>Use AC_PATH_TOOL instead of AC_PATH_PROG for llvm-config.</li>
+</ul>
+
+<p>Paul Berry (1):</p>
+<ul>
+  <li>i965: Ensure that all necessary state is re-emitted if we run out of aperture.</li>
+</ul>
+
+<p>Paul Seidler (1):</p>
+<ul>
+  <li>build: move ARCH_LIBS definition outside of ASM definition</li>
+</ul>
+
+<p>Thomas Sondergaard (4):</p>
+<ul>
+  <li>mesa: Preliminary support for MSVC_VERSION=12.0</li>
+  <li>mesa: Fix compile error with MSVC 2013</li>
+  <li>mesa: Work around internal compiler error</li>
+  <li>mesa: Namespace qualify fma to override ambiguity with fma from math.h</li>
+</ul>
+
+<p>Tom Stellard (1):</p>
+<ul>
+  <li>r600g/compute: Emit DEALLOC_STATE on cayman after dispatching a compute shader.</li>
+</ul>
+
+</div>
+</body>
+</html>

docs/relnotes/10.0.4.html

@@ -0,0 +1,191 @@
+<!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 10.0.4 Release Notes / (March 12, 2014)</h1>
+
+<p>
+Mesa 10.0.4 is a bug fix release which fixes bugs found since the 10.0.3 release.
+</p>
+<p>
+Mesa 10.0.4 implements the OpenGL 3.3 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 3.3.  OpenGL
+3.3 is <strong>only</strong> available if requested at context creation
+because compatibility contexts not supported.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+5a3c5b90776ec8a9fcd777c99e0607e2  MesaLib-10.0.4.tar.gz
+8b148869d2620b0720c8a8d2b7eb3e38  MesaLib-10.0.4.tar.bz2
+da2418d25bfbc273660af7e755fb367e  MesaLib-10.0.4.zip
+</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=71870">Bug 71870</a> - Metro: Last Light rendering issues</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=72895">Bug 72895</a> - Missing trees in flightgear 2.12.1 with mesa 10.0.1</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74251">Bug 74251</a> - Segfault in st_finalize_texture with Texture Buffer</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74723">Bug 74723</a> - main/shaderapi.c:407: detach_shader: Assertion `shProg-&gt;Shaders[j]-&gt;Type == 0x8B31 || shProg-&gt;Shaders[j]-&gt;Type == 0x8B30' failed.</li>
+
+</ul>
+
+<h2>Changes</h2>
+
+<p>The full set of changes can be viewed by using the following git command:</p>
+
+<pre>
+  git log mesa-10.0.3..mesa-10.0.4
+</pre>
+
+<p>Anuj Phogat (4):</p>
+<ul>
+  <li>mesa: Generate correct error code in glDrawBuffers()</li>
+  <li>mesa: Add GL_TEXTURE_CUBE_MAP_ARRAY to legal_get_tex_level_parameter_target()</li>
+  <li>glsl: Fix condition to generate shader link error</li>
+  <li>i965: Fix the region's pitch condition to use blitter</li>
+</ul>
+
+<p>Brian Paul (8):</p>
+<ul>
+  <li>r200: move driContextSetFlags(ctx) call after ctx var is initialized</li>
+  <li>radeon: move driContextSetFlags(ctx) call after ctx var is initialized</li>
+  <li>gallium/auxiliary/indices: replace free() with FREE()</li>
+  <li>draw: fix incorrect color of flat-shaded clipped lines</li>
+  <li>st/mesa: avoid sw fallback for getting/decompressing textures</li>
+  <li>mesa: update assertion in detach_shader() for geom shaders</li>
+  <li>mesa: do depth/stencil format conversion in glGetTexImage</li>
+  <li>softpipe: use 64-bit arithmetic in softpipe_resource_layout()</li>
+</ul>
+
+<p>Carl Worth (4):</p>
+<ul>
+  <li>docs: Add md5sums for 10.0.3 release</li>
+  <li>main: Avoid double-free of shader Label</li>
+  <li>get-pick-list: Update to only find patches nominated for the 10.0 branch</li>
+  <li>Update version to 10.0.4</li>
+</ul>
+
+<p>Chris Forbes (1):</p>
+<ul>
+  <li>i965: Validate (and resolve) all the bound textures.</li>
+</ul>
+
+<p>Christian König (1):</p>
+<ul>
+  <li>radeon/uvd: fix feedback buffer handling v2</li>
+</ul>
+
+<p>Daniel Kurtz (1):</p>
+<ul>
+  <li>glsl: Add locking to builtin_builder singleton</li>
+</ul>
+
+<p>Emil Velikov (3):</p>
+<ul>
+  <li>dri/nouveau: Pass the API into _mesa_initialize_context</li>
+  <li>nv50: correctly calculate the number of vertical blocks during transfer map</li>
+  <li>dri/i9*5: correctly calculate the amount of system memory</li>
+</ul>
+
+<p>Fredrik Höglund (3):</p>
+<ul>
+  <li>mesa: Preserve the NewArrays state when copying a VAO</li>
+  <li>glx: Fix the default values for GLXFBConfig attributes</li>
+  <li>glx: Fix the GLXFBConfig attrib sort priorities</li>
+</ul>
+
+<p>Hans (2):</p>
+<ul>
+  <li>util: don't define isfinite(), isnan() for MSVC &gt;= 1800</li>
+  <li>mesa: don't define c99 math functions for MSVC &gt;= 1800</li>
+</ul>
+
+<p>Ian Romanick (6):</p>
+<ul>
+  <li>meta: Release resources used by decompress_texture_image</li>
+  <li>meta: Release resources used by _mesa_meta_DrawPixels</li>
+  <li>meta: Fallback to software for GetTexImage of compressed GL_TEXTURE_CUBE_MAP_ARRAY</li>
+  <li>meta: Consistenly use non-Apple VAO functions</li>
+  <li>glcpp: Only warn for macro names containing __</li>
+  <li>glsl: Only warn for macro names containing __</li>
+</ul>
+
+<p>Ilia Mirkin (3):</p>
+<ul>
+  <li>nv30: report 8 maximum inputs</li>
+  <li>nouveau/video: make sure that firmware is present when checking caps</li>
+  <li>nouveau: fix chipset checks for nv1a by using the oclass instead</li>
+</ul>
+
+<p>Julien Cristau (1):</p>
+<ul>
+  <li>glx/dri2: fix build failure on HURD</li>
+</ul>
+
+<p>Kenneth Graunke (2):</p>
+<ul>
+  <li>glsl: Don't lose precision qualifiers when encountering "centroid".</li>
+  <li>i965: Create a hardware context before initializing state module.</li>
+</ul>
+
+<p>Kusanagi Kouichi (1):</p>
+<ul>
+  <li>targets/vdpau: Always use c++ to link</li>
+</ul>
+
+<p>Marek Olšák (1):</p>
+<ul>
+  <li>st/mesa: fix crash when a shader uses a TBO and it's not bound</li>
+</ul>
+
+<p>Matt Turner (1):</p>
+<ul>
+  <li>glsl: Initialize ubo_binding_mask flags to zero.</li>
+</ul>
+
+<p>Paul Berry (2):</p>
+<ul>
+  <li>glsl: Make condition_to_hir() callable from outside ast_iteration_statement.</li>
+  <li>glsl: Fix continue statements in do-while loops.</li>
+</ul>
+
+<p>Tom Stellard (1):</p>
+<ul>
+  <li>r600g/compute: PIPE_CAP_COMPUTE should be false for pre-evergreen GPUs</li>
+</ul>
+
+<p>Topi Pohjolainen (1):</p>
+<ul>
+  <li>i965/blorp: do not use unnecessary hw-blending support</li>
+</ul>
+
+</div>
+</body>
+</html>

docs/relnotes/10.0.5.html

@@ -0,0 +1,170 @@
+<!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 10.0.5 Release Notes / (April 18, 2014)</h1>
+
+<p>
+Mesa 10.0.5 is a bug fix release which fixes bugs found since the 10.0.4 release.
+</p>
+<p>
+Mesa 10.0.5 implements the OpenGL 3.3 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 3.3.  OpenGL
+3.3 is <strong>only</strong> available if requested at context creation
+because compatibility contexts not supported.
+</p>
+
+
+<h2>MD5 checksums</h2>
+<pre>
+</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=58660">Bug 58660</a> - CAYMAN broken with HyperZ on</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=64471">Bug 64471</a> - Radeon HD6570 lockup in Brütal Legend with HyperZ</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=66352">Bug 66352</a> - GPU lockup in L4D2 on TURKS with HyperZ</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=68799">Bug 68799</a> - [APITRACE] Hyper-Z lockup with Falcon BMS 4.32u6 on CAYMAN</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=71547">Bug 71547</a> - compilation failure :#error &quot;SSE4.1 instruction set not enabled&quot;</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=72685">Bug 72685</a> - [radeonsi hyperz] Artifacts in Unigine Sanctuary</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=73088">Bug 73088</a> - [HyperZ] Juniper (6770): Gone Home / Unigine Heaven 4.0 lock up system after several minutes of use</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74428">Bug 74428</a> - hyperz causes gpu hang in Counter-strike: Source</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74803">Bug 74803</a> - [r600g] HyperZ broken on RV630 (Cogs shadows are broken)</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74863">Bug 74863</a> - [r600g] HyperZ broken on RV770 and CYPRESS (Left 4 Dead 2 trees corruption) bisected!</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74892">Bug 74892</a> - HyperZ GPU lockup with radeonsi 7970M PITCAIRN and Distance Alpha game</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74988">Bug 74988</a> - Buffer overrun (segfault) decompressing ETC2 texture in GLBenchmark 3.0 Manhattan</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=75279">Bug 75279</a> - XCloseDisplay() takes one minute around nouveau_dri.so, freezing Firefox startup</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=77102">Bug 77102</a> - gallium nouveau has no profile in vdpau and libva</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=77207">Bug 77207</a> - [ivb/hsw] batch overwritten with garbage</li>
+
+</ul>
+
+<h2>Changes</h2>
+
+<p>The full set of changes can be viewed by using the following git command:</p>
+
+<pre>
+  git log mesa-10.0.4..mesa-10.0.5
+</pre>
+
+<p>Alex Deucher (1):</p>
+<ul>
+  <li>radeon: reverse DBG_NO_HYPERZ logic</li>
+</ul>
+
+<p>Brian Paul (9):</p>
+<ul>
+  <li>mesa: add unpacking code for MESA_FORMAT_Z32_FLOAT_S8X24_UINT</li>
+  <li>mesa: fix copy &amp; paste bugs in pack_ubyte_SARGB8()</li>
+  <li>mesa: fix copy &amp; paste bugs in pack_ubyte_SRGB8()</li>
+  <li>mesa: fix unpack_Z32_FLOAT_X24S8() / unpack_Z32_FLOAT() mix-up</li>
+  <li>st/mesa: add null pointer checking in query object functions</li>
+  <li>mesa: fix glMultiDrawArrays inside a display list</li>
+  <li>cso: fix sampler view count in cso_set_sampler_views()</li>
+  <li>svga: replace sampler assertion with conditional</li>
+  <li>svga: move LIST_INITHEAD(dirty_buffers) earlier in svga_context_create()</li>
+</ul>
+
+<p>Carl Worth (3):</p>
+<ul>
+  <li>docs: Add md5sums for the 10.0.4 release.</li>
+  <li>Ignore patches which don't apply.</li>
+  <li>Update version to 10.0.5</li>
+</ul>
+
+<p>Christian König (2):</p>
+<ul>
+  <li>st/mesa: recreate sampler view on context change v3</li>
+  <li>st/mesa: fix sampler view handling with shared textures v4</li>
+</ul>
+
+<p>Courtney Goeltzenleuchter (1):</p>
+<ul>
+  <li>mesa: add bounds checking to eliminate buffer overrun</li>
+</ul>
+
+<p>Emil Velikov (2):</p>
+<ul>
+  <li>mesa: return v.value_int64 when the requested type is TYPE_INT64</li>
+  <li>glx: drop obsolete _XUnlock_Mutex in __glXInitialize error path</li>
+</ul>
+
+<p>Eric Anholt (1):</p>
+<ul>
+  <li>i965: Fix buffer overruns in MSAA MCS buffer clearing.</li>
+</ul>
+
+<p>Ilia Mirkin (6):</p>
+<ul>
+  <li>nouveau: fix fence waiting logic in screen destroy</li>
+  <li>nv50: adjust blit_3d handling of ms output textures</li>
+  <li>mesa/main: condition GL_DEPTH_STENCIL on ARB_depth_texture</li>
+  <li>nouveau: add forgotten GL_COMPRESSED_INTENSITY to texture format list</li>
+  <li>nouveau: there may not have been a texture if the fbo was incomplete</li>
+  <li>nouveau: fix firmware check on nvd7/nvd9</li>
+</ul>
+
+<p>Johannes Nixdorf (1):</p>
+<ul>
+  <li>configure.ac: fix the detection of expat with pkg-config</li>
+</ul>
+
+<p>Jonathan Gray (1):</p>
+<ul>
+  <li>gallium: add endian detection for OpenBSD</li>
+</ul>
+
+<p>José Fonseca (1):</p>
+<ul>
+  <li>draw: Duplicate TGSI tokens in draw_pipe_pstipple module.</li>
+</ul>
+
+<p>Matt Turner (1):</p>
+<ul>
+  <li>mesa: Wrap SSE4.1 code in #ifdef __SSE4_1__.</li>
+</ul>
+
+<p>Paul Berry (1):</p>
+<ul>
+  <li>i965/gen7: Prefer vertical alignment of 4 when possible.</li>
+</ul>
+
+</div>
+</body>
+</html>

include/GL/glxext.h

@@ -33,10 +33,10 @@ extern "C" {
 ** used to make the header, and the header can be found at
 **   http://www.opengl.org/registry/
 **
-** Khronos $Revision: 23422 $ on $Date: 2013-10-08 15:40:45 -0700 (Tue, 08 Oct 2013) $
+** Khronos $Revision: 24777 $ on $Date: 2014-01-14 14:02:32 -0800 (Tue, 14 Jan 2014) $
 */
 
-#define GLX_GLXEXT_VERSION 20131008
+#define GLX_GLXEXT_VERSION 20140114
 
 /* Generated C header for:
  * API: glx
@@ -49,6 +49,7 @@ extern "C" {
 
 #ifndef GLX_VERSION_1_3
 #define GLX_VERSION_1_3 1
+typedef XID GLXContextID;
 typedef struct __GLXFBConfigRec *GLXFBConfig;
 typedef XID GLXWindow;
 typedef XID GLXPbuffer;
@@ -272,7 +273,6 @@ __GLXextFuncPtr glXGetProcAddressARB (const GLubyte *procName);
 
 #ifndef GLX_EXT_import_context
 #define GLX_EXT_import_context 1
-typedef XID GLXContextID;
 #define GLX_SHARE_CONTEXT_EXT             0x800A
 #define GLX_VISUAL_ID_EXT                 0x800B
 #define GLX_SCREEN_EXT                    0x800C
@@ -433,6 +433,14 @@ void glXCopyImageSubDataNV (Display *dpy, GLXContext srcCtx, GLuint srcName, GLe
 #endif
 #endif /* GLX_NV_copy_image */
 
+#ifndef GLX_NV_delay_before_swap
+#define GLX_NV_delay_before_swap 1
+typedef Bool ( *PFNGLXDELAYBEFORESWAPNVPROC) (Display *dpy, GLXDrawable drawable, GLfloat seconds);
+#ifdef GLX_GLXEXT_PROTOTYPES
+Bool glXDelayBeforeSwapNV (Display *dpy, GLXDrawable drawable, GLfloat seconds);
+#endif
+#endif /* GLX_NV_delay_before_swap */
+
 #ifndef GLX_NV_float_buffer
 #define GLX_NV_float_buffer 1
 #define GLX_FLOAT_COMPONENTS_NV           0x20B0

m4/ax_check_compile_flag.m4

@@ -0,0 +1,72 @@
+# ===========================================================================
+#   http://www.gnu.org/software/autoconf-archive/ax_check_compile_flag.html
+# ===========================================================================
+#
+# SYNOPSIS
+#
+#   AX_CHECK_COMPILE_FLAG(FLAG, [ACTION-SUCCESS], [ACTION-FAILURE], [EXTRA-FLAGS])
+#
+# DESCRIPTION
+#
+#   Check whether the given FLAG works with the current language's compiler
+#   or gives an error.  (Warnings, however, are ignored)
+#
+#   ACTION-SUCCESS/ACTION-FAILURE are shell commands to execute on
+#   success/failure.
+#
+#   If EXTRA-FLAGS is defined, it is added to the current language's default
+#   flags (e.g. CFLAGS) when the check is done.  The check is thus made with
+#   the flags: "CFLAGS EXTRA-FLAGS FLAG".  This can for example be used to
+#   force the compiler to issue an error when a bad flag is given.
+#
+#   NOTE: Implementation based on AX_CFLAGS_GCC_OPTION. Please keep this
+#   macro in sync with AX_CHECK_{PREPROC,LINK}_FLAG.
+#
+# LICENSE
+#
+#   Copyright (c) 2008 Guido U. Draheim <guidod@gmx.de>
+#   Copyright (c) 2011 Maarten Bosmans <mkbosmans@gmail.com>
+#
+#   This program is free software: you can redistribute it and/or modify it
+#   under the terms of the GNU General Public License as published by the
+#   Free Software Foundation, either version 3 of the License, or (at your
+#   option) any later version.
+#
+#   This program is distributed in the hope that it will be useful, but
+#   WITHOUT ANY WARRANTY; without even the implied warranty of
+#   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General
+#   Public License for more details.
+#
+#   You should have received a copy of the GNU General Public License along
+#   with this program. If not, see <http://www.gnu.org/licenses/>.
+#
+#   As a special exception, the respective Autoconf Macro's copyright owner
+#   gives unlimited permission to copy, distribute and modify the configure
+#   scripts that are the output of Autoconf when processing the Macro. You
+#   need not follow the terms of the GNU General Public License when using
+#   or distributing such scripts, even though portions of the text of the
+#   Macro appear in them. The GNU General Public License (GPL) does govern
+#   all other use of the material that constitutes the Autoconf Macro.
+#
+#   This special exception to the GPL applies to versions of the Autoconf
+#   Macro released by the Autoconf Archive. When you make and distribute a
+#   modified version of the Autoconf Macro, you may extend this special
+#   exception to the GPL to apply to your modified version as well.
+
+#serial 2
+
+AC_DEFUN([AX_CHECK_COMPILE_FLAG],
+[AC_PREREQ(2.59)dnl for _AC_LANG_PREFIX
+AS_VAR_PUSHDEF([CACHEVAR],[ax_cv_check_[]_AC_LANG_ABBREV[]flags_$4_$1])dnl
+AC_CACHE_CHECK([whether _AC_LANG compiler accepts $1], CACHEVAR, [
+  ax_check_save_flags=$[]_AC_LANG_PREFIX[]FLAGS
+  _AC_LANG_PREFIX[]FLAGS="$[]_AC_LANG_PREFIX[]FLAGS $4 $1"
+  AC_COMPILE_IFELSE([AC_LANG_PROGRAM()],
+    [AS_VAR_SET(CACHEVAR,[yes])],
+    [AS_VAR_SET(CACHEVAR,[no])])
+  _AC_LANG_PREFIX[]FLAGS=$ax_check_save_flags])
+AS_IF([test x"AS_VAR_GET(CACHEVAR)" = xyes],
+  [m4_default([$2], :)],
+  [m4_default([$3], :)])
+AS_VAR_POPDEF([CACHEVAR])dnl
+])dnl AX_CHECK_COMPILE_FLAGS

src/gallium/auxiliary/cso_cache/cso_context.c

@@ -1182,11 +1182,12 @@ cso_set_sampler_views(struct cso_context *ctx,
       pipe_sampler_view_reference(&info->views[i], NULL);
    }
 
-   info->nr_views = count;
-
    /* bind the new sampler views */
-   ctx->pipe->set_sampler_views(ctx->pipe, shader_stage, 0, count,
+   ctx->pipe->set_sampler_views(ctx->pipe, shader_stage, 0,
+                                MAX2(info->nr_views, count),
                                 info->views);
+
+   info->nr_views = count;
 }
 
 

src/gallium/auxiliary/draw/draw_pipe_clip.c

@@ -588,7 +588,12 @@ do_clip_line( struct draw_stage *stage,
 
    if (v0->clipmask) {
       interp( clipper, stage->tmp[0], t0, v0, v1, viewport_index );
-      copy_flat(stage, stage->tmp[0], v0);
+      if (stage->draw->rasterizer->flatshade_first) {
+         copy_flat(stage, stage->tmp[0], v0);  /* copy v0 color to tmp[0] */
+      }
+      else {
+         copy_flat(stage, stage->tmp[0], v1);  /* copy v1 color to tmp[0] */
+      }
       newprim.v[0] = stage->tmp[0];
    }
    else {
@@ -597,6 +602,12 @@ do_clip_line( struct draw_stage *stage,
 
    if (v1->clipmask) {
       interp( clipper, stage->tmp[1], t1, v1, v0, viewport_index );
+      if (stage->draw->rasterizer->flatshade_first) {
+         copy_flat(stage, stage->tmp[1], v0);  /* copy v0 color to tmp[1] */
+      }
+      else {
+         copy_flat(stage, stage->tmp[1], v1);  /* copy v1 color to tmp[1] */
+      }
       newprim.v[1] = stage->tmp[1];
    }
    else {

src/gallium/auxiliary/draw/draw_pipe_pstipple.c

@@ -673,7 +673,7 @@ pstip_create_fs_state(struct pipe_context *pipe,
    struct pstip_fragment_shader *pstipfs = CALLOC_STRUCT(pstip_fragment_shader);
 
    if (pstipfs) {
-      pstipfs->state = *fs;
+      pstipfs->state.tokens = tgsi_dup_tokens(fs->tokens);
 
       /* pass-through */
       pstipfs->driver_fs = pstip->driver_create_fs_state(pstip->pipe, fs);
@@ -707,6 +707,7 @@ pstip_delete_fs_state(struct pipe_context *pipe, void *fs)
    if (pstipfs->pstip_fs)
       pstip->driver_delete_fs_state(pstip->pipe, pstipfs->pstip_fs);
 
+   FREE((void*)pstipfs->state.tokens);
    FREE(pstipfs);
 }
 

src/gallium/auxiliary/draw/draw_pt_fetch_shade_pipeline.c

@@ -72,8 +72,8 @@ static void fetch_pipeline_prepare( struct draw_pt_middle_end *middle,
 
    const unsigned gs_out_prim = (gs ? gs->output_primitive :
                                  u_assembled_prim(prim));
-   unsigned nr = MAX2( vs->info.num_inputs,
-		       draw_total_vs_outputs(draw) );
+   unsigned nr_vs_outputs = draw_total_vs_outputs(draw);
+   unsigned nr = MAX2(vs->info.num_inputs, nr_vs_outputs);
 
    if (gs) {
       nr = MAX2(nr, gs->info.num_outputs + 1);
@@ -129,6 +129,9 @@ static void fetch_pipeline_prepare( struct draw_pt_middle_end *middle,
    /* No need to prepare the shader.
     */
    vs->prepare(vs, draw);
+
+   /* Make sure that the vertex size didn't change at any point above */
+   assert(nr_vs_outputs == draw_total_vs_outputs(draw));
 }
 
 

src/gallium/auxiliary/draw/draw_pt_fetch_shade_pipeline_llvm.c

@@ -141,19 +141,11 @@ llvm_middle_end_prepare( struct draw_pt_middle_end *middle,
    struct draw_geometry_shader *gs = draw->gs.geometry_shader;
    const unsigned out_prim = gs ? gs->output_primitive :
       u_assembled_prim(in_prim);
-   const unsigned nr = MAX2(vs->info.num_inputs,
-                            draw_total_vs_outputs(draw));
+   unsigned nr;
 
    fpme->input_prim = in_prim;
    fpme->opt = opt;
 
-   /* Always leave room for the vertex header whether we need it or
-    * not.  It's hard to get rid of it in particular because of the
-    * viewport code in draw_pt_post_vs.c.
-    */
-   fpme->vertex_size = sizeof(struct vertex_header) + nr * 4 * sizeof(float);
-
-
    draw_pt_post_vs_prepare( fpme->post_vs,
                             draw->clip_xy,
                             draw->clip_z,
@@ -177,6 +169,30 @@ llvm_middle_end_prepare( struct draw_pt_middle_end *middle,
       *max_vertices = 4096;
    }
 
+   /* Get the number of float[4] attributes per vertex.
+    * Note: this must be done after draw_pt_emit_prepare() since that
+    * can effect the vertex size.
+    */
+   nr = MAX2(vs->info.num_inputs, draw_total_vs_outputs(draw));
+
+   /* Always leave room for the vertex header whether we need it or
+    * not.  It's hard to get rid of it in particular because of the
+    * viewport code in draw_pt_post_vs.c.
+    */
+   fpme->vertex_size = sizeof(struct vertex_header) + nr * 4 * sizeof(float);
+
+   /* Get the number of float[4] attributes per vertex.
+    * Note: this must be done after draw_pt_emit_prepare() since that
+    * can effect the vertex size.
+    */
+   nr = MAX2(vs->info.num_inputs, draw_total_vs_outputs(draw));
+
+   /* Always leave room for the vertex header whether we need it or
+    * not.  It's hard to get rid of it in particular because of the
+    * viewport code in draw_pt_post_vs.c.
+    */
+   fpme->vertex_size = sizeof(struct vertex_header) + nr * 4 * sizeof(float);
+
    /* return even number */
    *max_vertices = *max_vertices & ~1;
 

src/gallium/auxiliary/indices/u_primconvert.c

@@ -74,7 +74,7 @@ void
 util_primconvert_destroy(struct primconvert_context *pc)
 {
    util_primconvert_save_index_buffer(pc, NULL);
-   free(pc);
+   FREE(pc);
 }
 
 void

src/gallium/auxiliary/rtasm/rtasm_execmem.c

@@ -69,7 +69,7 @@ static struct mem_block *exec_heap = NULL;
 static unsigned char *exec_mem = NULL;
 
 
-static void
+static int
 init_heap(void)
 {
    if (!exec_heap)
@@ -79,6 +79,8 @@ init_heap(void)
       exec_mem = (unsigned char *) mmap(0, EXEC_HEAP_SIZE, 
 					PROT_EXEC | PROT_READ | PROT_WRITE, 
 					MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+
+   return (exec_mem != MAP_FAILED);
 }
 
 
@@ -90,7 +92,8 @@ rtasm_exec_malloc(size_t size)
 
    pipe_mutex_lock(exec_mutex);
 
-   init_heap();
+   if (!init_heap())
+      goto bail;
 
    if (exec_heap) {
       size = (size + 31) & ~31;  /* next multiple of 32 bytes */
@@ -101,7 +104,8 @@ rtasm_exec_malloc(size_t size)
       addr = exec_mem + block->ofs;
    else 
       debug_printf("rtasm_exec_malloc failed\n");
-   
+
+bail:
    pipe_mutex_unlock(exec_mutex);
    
    return addr;

src/gallium/auxiliary/util/u_format.h

@@ -872,6 +872,9 @@ util_format_get_component_bits(enum pipe_format format,
 static INLINE enum pipe_format
 util_format_srgb(enum pipe_format format)
 {
+   if (util_format_is_srgb(format))
+      return format;
+
    switch (format) {
    case PIPE_FORMAT_L8_UNORM:
       return PIPE_FORMAT_L8_SRGB;

src/gallium/auxiliary/util/u_math.h

@@ -112,10 +112,13 @@ static INLINE float logf( float f )
 #define logf(x) ((float)log((double)(x)))
 #endif /* logf */
 
+#if _MSC_VER < 1800
 #define isfinite(x) _finite((double)(x))
 #define isnan(x) _isnan((double)(x))
+#endif /* _MSC_VER < 1800 */
 #endif /* _MSC_VER < 1400 && !defined(__cplusplus) */
 
+#if _MSC_VER < 1800
 static INLINE double log2( double x )
 {
    const double invln2 = 1.442695041;
@@ -133,6 +136,7 @@ roundf(float x)
 {
    return x >= 0.0f ? floorf(x + 0.5f) : ceilf(x - 0.5f);
 }
+#endif
 
 #define INFINITY (DBL_MAX + DBL_MAX)
 #define NAN (INFINITY - INFINITY)

src/gallium/drivers/nouveau/nouveau_screen.h

@@ -49,6 +49,11 @@ struct nouveau_screen {
 
 	boolean hint_buf_keep_sysmem_copy;
 
+	struct {
+		unsigned profiles_checked;
+		unsigned profiles_present;
+	} firmware_info;
+
 #ifdef NOUVEAU_ENABLE_DRIVER_STATISTICS
    union {
       uint64_t v[29];

src/gallium/drivers/nouveau/nouveau_vp3_video.c

@@ -21,6 +21,7 @@
  */
 
 #include <sys/mman.h>
+#include <sys/stat.h>
 #include <stdio.h>
 #include <fcntl.h>
 
@@ -350,6 +351,77 @@ nouveau_vp3_load_firmware(struct nouveau_vp3_decoder *dec,
    return 0;
 }
 
+static int
+firmware_present(struct pipe_screen *pscreen, enum pipe_video_profile profile)
+{
+   struct nouveau_screen *screen = nouveau_screen(pscreen);
+   int chipset = screen->device->chipset;
+   int vp3 = chipset < 0xa3 || chipset == 0xaa || chipset == 0xac;
+   int vp5 = chipset >= 0xd0;
+   int ret;
+
+   /* For all chipsets, try to create a BSP objects. Assume that if firmware
+    * is present for it, firmware is also present for VP/PPP */
+   if (!(screen->firmware_info.profiles_checked & 1)) {
+      struct nouveau_object *channel = NULL, *bsp = NULL;
+      struct nv04_fifo nv04_data = {.vram = 0xbeef0201, .gart = 0xbeef0202};
+      struct nvc0_fifo nvc0_args = {};
+      struct nve0_fifo nve0_args = {.engine = NVE0_FIFO_ENGINE_BSP};
+      void *data = NULL;
+      int size, oclass;
+      if (chipset < 0xc0)
+         oclass = 0x85b1;
+      else if (chipset < 0xe0)
+         oclass = 0x90b1;
+      else
+         oclass = 0x95b1;
+
+      if (chipset < 0xc0) {
+         data = &nv04_data;
+         size = sizeof(nv04_data);
+      } else if (chipset < 0xe0) {
+         data = &nvc0_args;
+         size = sizeof(nvc0_args);
+      } else {
+         data = &nve0_args;
+         size = sizeof(nve0_args);
+      }
+
+      /* kepler must have its own channel, so just do this for everyone */
+      nouveau_object_new(&screen->device->object, 0,
+                         NOUVEAU_FIFO_CHANNEL_CLASS,
+                         data, size, &channel);
+
+      if (channel) {
+         nouveau_object_new(channel, 0, oclass, NULL, 0, &bsp);
+         if (bsp)
+            screen->firmware_info.profiles_present |= 1;
+         nouveau_object_del(&bsp);
+         nouveau_object_del(&channel);
+      }
+      screen->firmware_info.profiles_checked |= 1;
+   }
+
+   if (!(screen->firmware_info.profiles_present & 1))
+      return 0;
+
+   /* For vp3/vp4 chipsets, make sure that the relevant firmware is present */
+   if (!vp5 && !(screen->firmware_info.profiles_checked & (1 << profile))) {
+      char path[PATH_MAX];
+      struct stat s;
+      if (vp3)
+         vp3_getpath(profile, path);
+      else
+         vp4_getpath(profile, path);
+      ret = stat(path, &s);
+      if (!ret && s.st_size > 1000)
+         screen->firmware_info.profiles_present |= (1 << profile);
+      screen->firmware_info.profiles_checked |= (1 << profile);
+   }
+
+   return vp5 || (screen->firmware_info.profiles_present & (1 << profile));
+}
+
 int
 nouveau_vp3_screen_get_video_param(struct pipe_screen *pscreen,
                                    enum pipe_video_profile profile,
@@ -363,8 +435,10 @@ nouveau_vp3_screen_get_video_param(struct pipe_screen *pscreen,
    switch (param) {
    case PIPE_VIDEO_CAP_SUPPORTED:
       /* VP3 does not support MPEG4, VP4+ do. */
-      return profile >= PIPE_VIDEO_PROFILE_MPEG1 && (
-            !vp3 || codec != PIPE_VIDEO_FORMAT_MPEG4);
+      return entrypoint == PIPE_VIDEO_ENTRYPOINT_BITSTREAM &&
+         profile >= PIPE_VIDEO_PROFILE_MPEG1 &&
+         (!vp3 || codec != PIPE_VIDEO_FORMAT_MPEG4) &&
+         firmware_present(pscreen, profile);
    case PIPE_VIDEO_CAP_NPOT_TEXTURES:
       return 1;
    case PIPE_VIDEO_CAP_MAX_WIDTH:

src/gallium/drivers/nouveau/nv30/nv30_screen.c

@@ -218,7 +218,7 @@ nv30_screen_get_shader_param(struct pipe_screen *pscreen, unsigned shader,
       case PIPE_SHADER_CAP_MAX_CONTROL_FLOW_DEPTH:
          return 0;
       case PIPE_SHADER_CAP_MAX_INPUTS:
-         return (eng3d->oclass >= NV40_3D_CLASS) ? 12 : 10;
+         return 8; /* should be possible to do 10 with nv4x */
       case PIPE_SHADER_CAP_MAX_CONSTS:
          return (eng3d->oclass >= NV40_3D_CLASS) ? 224 : 32;
       case PIPE_SHADER_CAP_MAX_CONST_BUFFERS:
@@ -298,10 +298,16 @@ nv30_screen_destroy(struct pipe_screen *pscreen)
 {
    struct nv30_screen *screen = nv30_screen(pscreen);
 
-   if (screen->base.fence.current &&
-       screen->base.fence.current->state >= NOUVEAU_FENCE_STATE_EMITTED) {
-      nouveau_fence_wait(screen->base.fence.current);
-      nouveau_fence_ref (NULL, &screen->base.fence.current);
+   if (screen->base.fence.current) {
+      struct nouveau_fence *current = NULL;
+
+      /* nouveau_fence_wait will create a new current fence, so wait on the
+       * _current_ one, and remove both.
+       */
+      nouveau_fence_ref(screen->base.fence.current, &current);
+      nouveau_fence_wait(current);
+      nouveau_fence_ref(NULL, &current);
+      nouveau_fence_ref(NULL, &screen->base.fence.current);
    }
 
    nouveau_object_del(&screen->query);

src/gallium/drivers/nouveau/nv50/nv50_context.c

@@ -165,7 +165,7 @@ nv50_invalidate_resource_storage(struct nouveau_context *ctx,
    }
 
    if (res->bind & PIPE_BIND_SAMPLER_VIEW) {
-      for (s = 0; s < 5; ++s) {
+      for (s = 0; s < 3; ++s) {
       for (i = 0; i < nv50->num_textures[s]; ++i) {
          if (nv50->textures[s][i] &&
              nv50->textures[s][i]->texture == res) {
@@ -179,7 +179,7 @@ nv50_invalidate_resource_storage(struct nouveau_context *ctx,
    }
 
    if (res->bind & PIPE_BIND_CONSTANT_BUFFER) {
-      for (s = 0; s < 5; ++s) {
+      for (s = 0; s < 3; ++s) {
       for (i = 0; i < nv50->num_vtxbufs; ++i) {
          if (!nv50->constbuf[s][i].user &&
              nv50->constbuf[s][i].u.buf == res) {

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

@@ -283,8 +283,15 @@ nv50_screen_destroy(struct pipe_screen *pscreen)
    struct nv50_screen *screen = nv50_screen(pscreen);
 
    if (screen->base.fence.current) {
-      nouveau_fence_wait(screen->base.fence.current);
-      nouveau_fence_ref (NULL, &screen->base.fence.current);
+      struct nouveau_fence *current = NULL;
+
+      /* nouveau_fence_wait will create a new current fence, so wait on the
+       * _current_ one, and remove both.
+       */
+      nouveau_fence_ref(screen->base.fence.current, &current);
+      nouveau_fence_wait(current);
+      nouveau_fence_ref(NULL, &current);
+      nouveau_fence_ref(NULL, &screen->base.fence.current);
    }
    if (screen->base.pushbuf)
       screen->base.pushbuf->user_priv = NULL;

src/gallium/drivers/nouveau/nv50/nv50_surface.c

@@ -935,6 +935,7 @@ nv50_blit_3d(struct nv50_context *nv50, const struct pipe_blit_info *info)
    float x0, x1, y0, y1, z;
    float dz;
    float x_range, y_range;
+   float tri_x, tri_y;
 
    blit->mode = nv50_blit_select_mode(info);
    blit->color_mask = nv50_blit_derive_color_mask(info);
@@ -954,11 +955,14 @@ nv50_blit_3d(struct nv50_context *nv50, const struct pipe_blit_info *info)
    x_range = (float)info->src.box.width / (float)info->dst.box.width;
    y_range = (float)info->src.box.height / (float)info->dst.box.height;
 
+   tri_x = 16384 << nv50_miptree(dst)->ms_x;
+   tri_y = 16384 << nv50_miptree(dst)->ms_y;
+
    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 + tri_x * x_range;
+   y1 = y0 + tri_y * y_range;
 
    x0 *= (float)(1 << nv50_miptree(src)->ms_x);
    x1 *= (float)(1 << nv50_miptree(src)->ms_x);
@@ -1027,7 +1031,7 @@ nv50_blit_3d(struct nv50_context *nv50, const struct pipe_blit_info *info)
       PUSH_DATAf(push, y0);
       PUSH_DATAf(push, z);
       BEGIN_NV04(push, NV50_3D(VTX_ATTR_2F_X(0)), 2);
-      PUSH_DATAf(push, 16384 << nv50_miptree(dst)->ms_x);
+      PUSH_DATAf(push, tri_x);
       PUSH_DATAf(push, 0.0f);
       BEGIN_NV04(push, NV50_3D(VTX_ATTR_3F_X(1)), 3);
       PUSH_DATAf(push, x0);
@@ -1035,7 +1039,7 @@ nv50_blit_3d(struct nv50_context *nv50, const struct pipe_blit_info *info)
       PUSH_DATAf(push, z);
       BEGIN_NV04(push, NV50_3D(VTX_ATTR_2F_X(0)), 2);
       PUSH_DATAf(push, 0.0f);
-      PUSH_DATAf(push, 16384 << nv50_miptree(dst)->ms_y);
+      PUSH_DATAf(push, tri_y);
       BEGIN_NV04(push, NV50_3D(VERTEX_END_GL), 1);
       PUSH_DATA (push, 0);
    }

src/gallium/drivers/nouveau/nv50/nv50_transfer.c

@@ -278,7 +278,7 @@ nv50_miptree_transfer_map(struct pipe_context *pctx,
 
    if (util_format_is_plain(res->format)) {
       tx->nblocksx = box->width << mt->ms_x;
-      tx->nblocksy = box->height << mt->ms_x;
+      tx->nblocksy = box->height << mt->ms_y;
    } else {
       tx->nblocksx = util_format_get_nblocksx(res->format, box->width);
       tx->nblocksy = util_format_get_nblocksy(res->format, box->height);

src/gallium/drivers/nouveau/nv50/nv84_video.c

@@ -741,16 +741,80 @@ error:
    return NULL;
 }
 
+#define FIRMWARE_BSP_KERN  0x01
+#define FIRMWARE_VP_KERN   0x02
+#define FIRMWARE_BSP_H264  0x04
+#define FIRMWARE_VP_MPEG2  0x08
+#define FIRMWARE_VP_H264_1 0x10
+#define FIRMWARE_VP_H264_2 0x20
+#define FIRMWARE_PRESENT(val, fw) (val & FIRMWARE_ ## fw)
+
+static int
+firmware_present(struct pipe_screen *pscreen, enum pipe_video_format codec)
+{
+   struct nouveau_screen *screen = nouveau_screen(pscreen);
+   struct nouveau_object *obj = NULL;
+   struct stat s;
+   int checked = screen->firmware_info.profiles_checked;
+   int present, ret;
+
+   if (!FIRMWARE_PRESENT(checked, VP_KERN)) {
+      nouveau_object_new(screen->channel, 0, 0x7476, NULL, 0, &obj);
+      if (obj)
+         screen->firmware_info.profiles_present |= FIRMWARE_VP_KERN;
+      nouveau_object_del(&obj);
+      screen->firmware_info.profiles_checked |= FIRMWARE_VP_KERN;
+   }
+
+   if (codec == PIPE_VIDEO_FORMAT_MPEG4_AVC) {
+      if (!FIRMWARE_PRESENT(checked, BSP_KERN)) {
+         nouveau_object_new(screen->channel, 0, 0x74b0, NULL, 0, &obj);
+         if (obj)
+            screen->firmware_info.profiles_present |= FIRMWARE_BSP_KERN;
+         nouveau_object_del(&obj);
+         screen->firmware_info.profiles_checked |= FIRMWARE_BSP_KERN;
+      }
+
+      if (!FIRMWARE_PRESENT(checked, VP_H264_1)) {
+         ret = stat("/lib/firmware/nouveau/nv84_vp-h264-1", &s);
+         if (!ret && s.st_size > 1000)
+            screen->firmware_info.profiles_present |= FIRMWARE_VP_H264_1;
+         screen->firmware_info.profiles_checked |= FIRMWARE_VP_H264_1;
+      }
+
+      /* should probably check the others, but assume that 1 means all */
+
+      present = screen->firmware_info.profiles_present;
+      return FIRMWARE_PRESENT(present, VP_KERN) &&
+         FIRMWARE_PRESENT(present, BSP_KERN) &&
+         FIRMWARE_PRESENT(present, VP_H264_1);
+   } else {
+      if (!FIRMWARE_PRESENT(checked, VP_MPEG2)) {
+         ret = stat("/lib/firmware/nouveau/nv84_vp-mpeg12", &s);
+         if (!ret && s.st_size > 1000)
+            screen->firmware_info.profiles_present |= FIRMWARE_VP_MPEG2;
+         screen->firmware_info.profiles_checked |= FIRMWARE_VP_MPEG2;
+      }
+      present = screen->firmware_info.profiles_present;
+      return FIRMWARE_PRESENT(present, VP_KERN) &&
+         FIRMWARE_PRESENT(present, VP_MPEG2);
+   }
+}
+
 int
 nv84_screen_get_video_param(struct pipe_screen *pscreen,
                             enum pipe_video_profile profile,
                             enum pipe_video_entrypoint entrypoint,
                             enum pipe_video_cap param)
 {
+   enum pipe_video_format codec;
+
    switch (param) {
    case PIPE_VIDEO_CAP_SUPPORTED:
-      return u_reduce_video_profile(profile) == PIPE_VIDEO_FORMAT_MPEG4_AVC ||
-         u_reduce_video_profile(profile) == PIPE_VIDEO_FORMAT_MPEG12;
+      codec = u_reduce_video_profile(profile);
+      return (codec == PIPE_VIDEO_FORMAT_MPEG4_AVC ||
+              codec == PIPE_VIDEO_FORMAT_MPEG12) &&
+         firmware_present(pscreen, codec);
    case PIPE_VIDEO_CAP_NPOT_TEXTURES:
       return 1;
    case PIPE_VIDEO_CAP_MAX_WIDTH:

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

@@ -331,7 +331,14 @@ nvc0_screen_destroy(struct pipe_screen *pscreen)
    struct nvc0_screen *screen = nvc0_screen(pscreen);
 
    if (screen->base.fence.current) {
-      nouveau_fence_wait(screen->base.fence.current);
+      struct nouveau_fence *current = NULL;
+
+      /* nouveau_fence_wait will create a new current fence, so wait on the
+       * _current_ one, and remove both.
+       */
+      nouveau_fence_ref(screen->base.fence.current, &current);
+      nouveau_fence_wait(current);
+      nouveau_fence_ref(NULL, &current);
       nouveau_fence_ref(NULL, &screen->base.fence.current);
    }
    if (screen->base.pushbuf)

src/gallium/drivers/r600/evergreen_compute.c

@@ -489,7 +489,14 @@ static void compute_emit_cs(struct r600_context *ctx, const uint *block_layout,
 	ctx->b.flags = 0;
 
 	if (ctx->b.chip_class >= CAYMAN) {
-		ctx->skip_surface_sync_on_next_cs_flush = true;
+		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
+		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_CS_PARTIAL_FLUSH) | EVENT_INDEX(4);
+		/* DEALLOC_STATE prevents the GPU from hanging when a
+		 * SURFACE_SYNC packet is emitted some time after a DISPATCH_DIRECT
+		 * with any of the CB*_DEST_BASE_ENA or DB_DEST_BASE_ENA bits set.
+		 */
+		cs->buf[cs->cdw++] = PKT3C(PKT3_DEALLOC_STATE, 0, 0);
+		cs->buf[cs->cdw++] = 0;
 	}
 
 #if 0

src/gallium/drivers/r600/evergreend.h

@@ -63,6 +63,7 @@
 #define R600_TEXEL_PITCH_ALIGNMENT_MASK        0x7
 
 #define PKT3_NOP                               0x10
+#define PKT3_DEALLOC_STATE                     0x14
 #define PKT3_DISPATCH_DIRECT                   0x15
 #define PKT3_DISPATCH_INDIRECT                 0x16
 #define PKT3_INDIRECT_BUFFER_END               0x17

src/gallium/drivers/r600/r600_hw_context.c

@@ -293,7 +293,7 @@ void r600_flush_emit(struct r600_context *rctx)
 				S_0085F0_SMX_ACTION_ENA(1);
 	}
 
-	if (cp_coher_cntl && !rctx->skip_surface_sync_on_next_cs_flush) {
+	if (cp_coher_cntl) {
 		cs->buf[cs->cdw++] = PKT3(PKT3_SURFACE_SYNC, 3, 0);
 		cs->buf[cs->cdw++] = cp_coher_cntl;   /* CP_COHER_CNTL */
 		cs->buf[cs->cdw++] = 0xffffffff;      /* CP_COHER_SIZE */
@@ -354,8 +354,6 @@ void r600_context_flush(struct r600_context *ctx, unsigned flags)
 
 	/* Flush the CS. */
 	ctx->b.ws->cs_flush(ctx->b.rings.gfx.cs, flags, ctx->screen->cs_count++);
-
-	ctx->skip_surface_sync_on_next_cs_flush = false;
 }
 
 void r600_begin_new_cs(struct r600_context *ctx)

src/gallium/drivers/r600/r600_pipe.c

@@ -43,7 +43,7 @@
 
 static const struct debug_named_value r600_debug_options[] = {
 	/* features */
-	{ "nohyperz", DBG_NO_HYPERZ, "Disable Hyper-Z" },
+	{ "hyperz", DBG_HYPERZ, "Enable Hyper-Z" },
 #if defined(R600_USE_LLVM)
 	{ "nollvm", DBG_NO_LLVM, "Disable the LLVM shader compiler" },
 #endif
@@ -192,13 +192,6 @@ static void r600_destroy_context(struct pipe_context *context)
 
 	r600_release_command_buffer(&rctx->start_cs_cmd);
 
-	if (rctx->b.rings.gfx.cs) {
-		rctx->b.ws->cs_destroy(rctx->b.rings.gfx.cs);
-	}
-	if (rctx->b.rings.dma.cs) {
-		rctx->b.ws->cs_destroy(rctx->b.rings.dma.cs);
-	}
-
 	FREE(rctx->start_compute_cs_cmd.buf);
 
 	r600_common_context_cleanup(&rctx->b);
@@ -420,7 +413,6 @@ static int r600_get_param(struct pipe_screen* pscreen, enum pipe_cap param)
 	case PIPE_CAP_VERTEX_ELEMENT_SRC_OFFSET_4BYTE_ALIGNED_ONLY:
 	case PIPE_CAP_USER_INDEX_BUFFERS:
 	case PIPE_CAP_USER_CONSTANT_BUFFERS:
-	case PIPE_CAP_COMPUTE:
 	case PIPE_CAP_START_INSTANCE:
 	case PIPE_CAP_MAX_DUAL_SOURCE_RENDER_TARGETS:
 	case PIPE_CAP_TEXTURE_BUFFER_OBJECTS:
@@ -429,6 +421,9 @@ static int r600_get_param(struct pipe_screen* pscreen, enum pipe_cap param)
 	case PIPE_CAP_TEXTURE_MULTISAMPLE:
 		return 1;
 
+	case PIPE_CAP_COMPUTE:
+		return rscreen->b.chip_class > R700;
+
 	case PIPE_CAP_TGSI_TEXCOORD:
 		return 0;
 
@@ -899,8 +894,8 @@ struct pipe_screen *r600_screen_create(struct radeon_winsys *ws)
 		rscreen->b.debug_flags |= DBG_COMPUTE;
 	if (debug_get_bool_option("R600_DUMP_SHADERS", FALSE))
 		rscreen->b.debug_flags |= DBG_FS | DBG_VS | DBG_GS | DBG_PS | DBG_CS;
-	if (!debug_get_bool_option("R600_HYPERZ", TRUE))
-		rscreen->b.debug_flags |= DBG_NO_HYPERZ;
+	if (debug_get_bool_option("R600_HYPERZ", FALSE))
+		rscreen->b.debug_flags |= DBG_HYPERZ;
 	if (!debug_get_bool_option("R600_LLVM", TRUE))
 		rscreen->b.debug_flags |= DBG_NO_LLVM;
 

src/gallium/drivers/r600/r600_pipe.h

@@ -507,16 +507,6 @@ struct r600_context {
 
 	void				*sb_context;
 	struct r600_isa		*isa;
-
-	/* Work-around for flushing problems with compute shaders on Cayman:
-	 * Emitting a SURFACE_SYNC packet with any of the CB*_DEST_BASE_ENA
-	 * or DB_DEST_BASE_ENA bits set after dispatching a compute shader
-	 * hangs the GPU.
-	 *
-	 * Setting this to true will prevent r600_flush_emit() from emitting
-	 * a SURFACE_SYNC packet.  This field will be cleared by
-	 * by r600_context_flush() after flushing the command stream. */
-	boolean				skip_surface_sync_on_next_cs_flush;
 };
 
 static INLINE void r600_emit_command_buffer(struct radeon_winsys_cs *cs,

src/gallium/drivers/radeon/r600_pipe_common.c

@@ -234,6 +234,13 @@ bool r600_common_context_init(struct r600_common_context *rctx,
 
 void r600_common_context_cleanup(struct r600_common_context *rctx)
 {
+	if (rctx->rings.gfx.cs) {
+		rctx->ws->cs_destroy(rctx->rings.gfx.cs);
+	}
+	if (rctx->rings.dma.cs) {
+		rctx->ws->cs_destroy(rctx->rings.dma.cs);
+	}
+
 	if (rctx->allocator_so_filled_size) {
 		u_suballocator_destroy(rctx->allocator_so_filled_size);
 	}

src/gallium/drivers/radeon/r600_pipe_common.h

@@ -73,7 +73,7 @@
 #define DBG_PS			(1 << 11)
 #define DBG_CS			(1 << 12)
 /* features */
-#define DBG_NO_HYPERZ		(1 << 13)
+#define DBG_HYPERZ		(1 << 13)
 /* The maximum allowed bit is 15. */
 
 struct r600_common_context;

src/gallium/drivers/radeon/r600_texture.c

@@ -542,7 +542,7 @@ r600_texture_create_object(struct pipe_screen *screen,
 	if (rtex->is_depth &&
 	    !(base->flags & (R600_RESOURCE_FLAG_TRANSFER |
 			     R600_RESOURCE_FLAG_FLUSHED_DEPTH)) &&
-	    !(rscreen->debug_flags & DBG_NO_HYPERZ)) {
+	    (rscreen->debug_flags & DBG_HYPERZ)) {
 		if (rscreen->chip_class >= SI) {
 			/* XXX implement Hyper-Z for SI.
 			 * Reuse the CMASK allocator, which is almost the same as HTILE. */

src/gallium/drivers/radeon/radeon_uvd.c

@@ -58,6 +58,9 @@
 #define NUM_H264_REFS 17
 #define NUM_VC1_REFS 5
 
+#define FB_BUFFER_OFFSET 0x1000
+#define FB_BUFFER_SIZE 2048
+
 /* UVD buffer representation */
 struct ruvd_buffer
 {
@@ -81,6 +84,7 @@ struct ruvd_decoder {
 
 	struct ruvd_buffer		msg_fb_buffers[NUM_BUFFERS];
 	struct ruvd_msg			*msg;
+	uint32_t			*fb;
 
 	struct ruvd_buffer		bs_buffers[NUM_BUFFERS];
 	void*				bs_ptr;
@@ -131,16 +135,21 @@ static void send_cmd(struct ruvd_decoder *dec, unsigned cmd,
 	set_reg(dec, RUVD_GPCOM_VCPU_CMD, cmd << 1);
 }
 
-/* map the next available message buffer */
-static void map_msg_buf(struct ruvd_decoder *dec)
+/* map the next available message/feedback buffer */
+static void map_msg_fb_buf(struct ruvd_decoder *dec)
 {
 	struct ruvd_buffer* buf;
+	uint8_t *ptr;
 
-	/* grap the current message buffer */
+	/* grab the current message/feedback buffer */
 	buf = &dec->msg_fb_buffers[dec->cur_buffer];
 
-	/* copy the message into it */
-	dec->msg = dec->ws->buffer_map(buf->cs_handle, dec->cs, PIPE_TRANSFER_WRITE);
+	/* and map it for CPU access */
+	ptr = dec->ws->buffer_map(buf->cs_handle, dec->cs, PIPE_TRANSFER_WRITE);
+
+	/* calc buffer offsets */
+	dec->msg = (struct ruvd_msg *)ptr;
+	dec->fb = (uint32_t *)(ptr + FB_BUFFER_OFFSET);
 }
 
 /* unmap and send a message command to the VCPU */
@@ -148,8 +157,8 @@ static void send_msg_buf(struct ruvd_decoder *dec)
 {
 	struct ruvd_buffer* buf;
 
-	/* ignore the request if message buffer isn't mapped */
-	if (!dec->msg)
+	/* ignore the request if message/feedback buffer isn't mapped */
+	if (!dec->msg || !dec->fb)
 		return;
 
 	/* grap the current message buffer */
@@ -157,6 +166,8 @@ static void send_msg_buf(struct ruvd_decoder *dec)
 
 	/* unmap the buffer */
 	dec->ws->buffer_unmap(buf->cs_handle);
+	dec->msg = NULL;
+	dec->fb = NULL;
 
 	/* and send it to the hardware */
 	send_cmd(dec, RUVD_CMD_MSG_BUFFER, buf->cs_handle, 0,
@@ -644,7 +655,7 @@ static void ruvd_destroy(struct pipe_video_codec *decoder)
 
 	assert(decoder);
 
-	map_msg_buf(dec);
+	map_msg_fb_buf(dec);
 	memset(dec->msg, 0, sizeof(*dec->msg));
 	dec->msg->size = sizeof(*dec->msg);
 	dec->msg->msg_type = RUVD_MSG_DESTROY;
@@ -773,7 +784,7 @@ static void ruvd_end_frame(struct pipe_video_codec *decoder,
 	memset(dec->bs_ptr, 0, bs_size - dec->bs_size);
 	dec->ws->buffer_unmap(bs_buf->cs_handle);
 
-	map_msg_buf(dec);
+	map_msg_fb_buf(dec);
 	dec->msg->size = sizeof(*dec->msg);
 	dec->msg->msg_type = RUVD_MSG_DECODE;
 	dec->msg->stream_handle = dec->stream_handle;
@@ -813,6 +824,10 @@ static void ruvd_end_frame(struct pipe_video_codec *decoder,
 
 	dec->msg->body.decode.db_surf_tile_config = dec->msg->body.decode.dt_surf_tile_config;
 	dec->msg->body.decode.extension_support = 0x1;
+
+	/* set at least the feedback buffer size */
+	dec->fb[0] = FB_BUFFER_SIZE;
+
 	send_msg_buf(dec);
 
 	send_cmd(dec, RUVD_CMD_DPB_BUFFER, dec->dpb.cs_handle, 0,
@@ -822,7 +837,7 @@ static void ruvd_end_frame(struct pipe_video_codec *decoder,
 	send_cmd(dec, RUVD_CMD_DECODING_TARGET_BUFFER, dt, 0,
 		 RADEON_USAGE_WRITE, RADEON_DOMAIN_VRAM);
 	send_cmd(dec, RUVD_CMD_FEEDBACK_BUFFER, msg_fb_buf->cs_handle,
-		 0x1000, RADEON_USAGE_WRITE, RADEON_DOMAIN_GTT);
+		 FB_BUFFER_OFFSET, RADEON_USAGE_WRITE, RADEON_DOMAIN_GTT);
 	set_reg(dec, RUVD_ENGINE_CNTL, 1);
 
 	flush(dec);
@@ -898,7 +913,8 @@ struct pipe_video_codec *ruvd_create_decoder(struct pipe_context *context,
 
 	bs_buf_size = width * height * 512 / (16 * 16);
 	for (i = 0; i < NUM_BUFFERS; ++i) {
-		unsigned msg_fb_size = align(sizeof(struct ruvd_msg), 0x1000) + 0x1000;
+		unsigned msg_fb_size = FB_BUFFER_OFFSET + FB_BUFFER_SIZE;
+		STATIC_ASSERT(sizeof(struct ruvd_msg) <= FB_BUFFER_OFFSET);
 		if (!create_buffer(dec, &dec->msg_fb_buffers[i], msg_fb_size)) {
 			RUVD_ERR("Can't allocated message buffers.\n");
 			goto error;
@@ -920,7 +936,7 @@ struct pipe_video_codec *ruvd_create_decoder(struct pipe_context *context,
 
 	clear_buffer(dec, &dec->dpb);
 
-	map_msg_buf(dec);
+	map_msg_fb_buf(dec);
 	dec->msg->size = sizeof(*dec->msg);
 	dec->msg->msg_type = RUVD_MSG_CREATE;
 	dec->msg->stream_handle = dec->stream_handle;

src/gallium/drivers/softpipe/sp_quad_blend.c

@@ -860,8 +860,8 @@ clamp_colors(float (*quadColor)[4])
 {
    unsigned i, j;
 
-   for (j = 0; j < TGSI_QUAD_SIZE; j++) {
-      for (i = 0; i < 4; i++) {
+   for (i = 0; i < 4; i++) {
+      for (j = 0; j < TGSI_QUAD_SIZE; j++) {
          quadColor[i][j] = CLAMP(quadColor[i][j], 0.0F, 1.0F);
       }
    }

src/gallium/drivers/softpipe/sp_texture.c

@@ -60,7 +60,7 @@ softpipe_resource_layout(struct pipe_screen *screen,
    unsigned width = pt->width0;
    unsigned height = pt->height0;
    unsigned depth = pt->depth0;
-   unsigned buffer_size = 0;
+   uint64_t buffer_size = 0;
 
    for (level = 0; level <= pt->last_level; level++) {
       unsigned slices;
@@ -76,8 +76,8 @@ softpipe_resource_layout(struct pipe_screen *screen,
 
       spr->level_offset[level] = buffer_size;
 
-      buffer_size += (util_format_get_nblocksy(pt->format, height) *
-                      slices * spr->stride[level]);
+      buffer_size += (uint64_t) util_format_get_nblocksy(pt->format, height) *
+                     slices * spr->stride[level];
 
       width  = u_minify(width, 1);
       height = u_minify(height, 1);

src/gallium/drivers/svga/svga_context.c

@@ -90,6 +90,8 @@ struct pipe_context *svga_context_create( struct pipe_screen *screen,
    if (svga == NULL)
       goto no_svga;
 
+   LIST_INITHEAD(&svga->dirty_buffers);
+
    svga->pipe.screen = screen;
    svga->pipe.priv = priv;
    svga->pipe.destroy = svga_destroy;
@@ -155,8 +157,6 @@ struct pipe_context *svga_context_create( struct pipe_screen *screen,
 
    svga->dirty = ~0;
 
-   LIST_INITHEAD(&svga->dirty_buffers);
-
    return &svga->pipe;
 
 no_state:

src/gallium/drivers/svga/svga_state_fs.c

@@ -199,15 +199,43 @@ make_fs_key(const struct svga_context *svga,
    if (svga->curr.blend->need_white_fragments) {
       key->white_fragments = 1;
    }
-   
+
+#ifdef DEBUG
+   /*
+    * We expect a consistent set of samplers and sampler views.
+    * Do some debug checks/warnings here.
+    */
+   {
+      static boolean warned = FALSE;
+      unsigned i, n = MAX2(svga->curr.num_sampler_views,
+                           svga->curr.num_samplers);
+      /* Only warn once to prevent too much debug output */
+      if (!warned) {
+         if (svga->curr.num_sampler_views != svga->curr.num_samplers) {
+            debug_printf("svga: mismatched number of sampler views (%u) "
+                         "vs. samplers (%u)\n",
+                         svga->curr.num_sampler_views,
+                         svga->curr.num_samplers);
+         }
+         for (i = 0; i < n; i++) {
+            if ((svga->curr.sampler_views[i] == NULL) !=
+                (svga->curr.sampler[i] == NULL))
+               debug_printf("sampler_view[%u] = %p but sampler[%u] = %p\n",
+                            i, svga->curr.sampler_views[i],
+                            i, svga->curr.sampler[i]);
+         }
+         warned = TRUE;
+      }
+   }
+#endif
+
    /* XXX: want to limit this to the textures that the shader actually
     * refers to.
     *
     * SVGA_NEW_TEXTURE_BINDING | SVGA_NEW_SAMPLER
     */
    for (i = 0; i < svga->curr.num_sampler_views; i++) {
-      if (svga->curr.sampler_views[i]) {
-         assert(svga->curr.sampler[i]);
+      if (svga->curr.sampler_views[i] && svga->curr.sampler[i]) {
          assert(svga->curr.sampler_views[i]->texture);
          key->tex[i].texture_target = svga->curr.sampler_views[i]->texture->target;
          if (!svga->curr.sampler[i]->normalized_coords) {

src/gallium/drivers/svga/svga_state_tss.c

@@ -85,7 +85,7 @@ update_tss_binding(struct svga_context *svga,
       struct pipe_sampler_view *sv = svga->curr.sampler_views[i];
 
       /* get min max lod */
-      if (sv) {
+      if (sv && s) {
          min_lod = MAX2(0, (s->view_min_lod + sv->u.tex.first_level));
          max_lod = MIN2(s->view_max_lod + sv->u.tex.first_level,
                         sv->texture->last_level);

src/gallium/include/pipe/p_config.h

@@ -157,6 +157,16 @@
 # define PIPE_ARCH_BIG_ENDIAN
 #endif
 
+#elif defined(__OpenBSD__)
+#include <sys/types.h>
+#include <machine/endian.h>
+
+#if _BYTE_ORDER == _LITTLE_ENDIAN
+# define PIPE_ARCH_LITTLE_ENDIAN
+#elif _BYTE_ORDER == _BIG_ENDIAN
+# define PIPE_ARCH_BIG_ENDIAN
+#endif
+
 #else
 
 #if defined(PIPE_ARCH_X86) || defined(PIPE_ARCH_X86_64) || defined(PIPE_ARCH_ARM) || defined(PIPE_ARCH_AARCH64)

src/gallium/state_trackers/dri/common/dri_screen.c

@@ -360,6 +360,12 @@ dri_destroy_option_cache(struct dri_screen * screen)
    }
 
    free(screen->optionCache.values);
+
+   /* Default values are copied to screen->optionCache->values in
+    * initOptionCache. The info field, however, is a pointer copy, so don't free
+    * that twice.
+    */
+   free(screen->optionCacheDefaults.values);
 }
 
 void

src/gallium/state_trackers/vdpau/device.c

@@ -72,6 +72,11 @@ vdp_imp_device_create_x11(Display *display, int screen, VdpDevice *device,
       goto no_context;
    }
 
+   if (!pscreen->get_param(pscreen, PIPE_CAP_NPOT_TEXTURES)) {
+      ret = VDP_STATUS_NO_IMPLEMENTATION;
+      goto no_context;
+   }
+
    *device = vlAddDataHTAB(dev);
    if (*device == 0) {
       ret = VDP_STATUS_ERROR;

src/gallium/targets/r600/vdpau/Makefile.am

@@ -58,17 +58,13 @@ libvdpau_r600_la_LIBADD = \
 	$(LIBDRM_LIBS) \
 	$(RADEON_LIBS)
 
-if HAVE_MESA_LLVM
 libvdpau_r600_la_LINK = $(CXXLINK) $(libvdpau_r600_la_LDFLAGS)
 # Mention a dummy pure C++ file to trigger generation of the $(LINK) variable
 nodist_EXTRA_libvdpau_r600_la_SOURCES = dummy-cpp.cpp
 
+if HAVE_MESA_LLVM
 libvdpau_r600_la_LDFLAGS += $(LLVM_LDFLAGS)
 libvdpau_r600_la_LIBADD += $(LLVM_LIBS)
-else
-libvdpau_r600_la_LINK = $(LINK) $(libvdpau_r600_la_LDFLAGS)
-# Mention a dummy pure C file to trigger generation of the $(LINK) variable
-nodist_EXTRA_libvdpau_r600_la_SOURCES = dummy-c.c
 endif
 
 # Provide compatibility with scripts for the old Mesa build system for

src/glsl/ast.h

@@ -846,14 +846,13 @@ public:
 
    ast_node *body;
 
-private:
    /**
     * Generate IR from the condition of a loop
     *
     * This is factored out of ::hir because some loops have the condition
     * test at the top (for and while), and others have it at the end (do-while).
     */
-   void condition_to_hir(class ir_loop *, struct _mesa_glsl_parse_state *);
+   void condition_to_hir(exec_list *, struct _mesa_glsl_parse_state *);
 };
 
 

src/glsl/ast_to_hir.cpp

@@ -734,14 +734,15 @@ do_assignment(exec_list *instructions, struct _mesa_glsl_parse_state *state,
 {
    void *ctx = state;
    bool error_emitted = (lhs->type->is_error() || rhs->type->is_error());
+   ir_rvalue *extract_channel = NULL;
 
    /* If the assignment LHS comes back as an ir_binop_vector_extract
     * expression, move it to the RHS as an ir_triop_vector_insert.
     */
    if (lhs->ir_type == ir_type_expression) {
-      ir_expression *const expr = lhs->as_expression();
+      ir_expression *const lhs_expr = lhs->as_expression();
 
-      if (unlikely(expr->operation == ir_binop_vector_extract)) {
+      if (unlikely(lhs_expr->operation == ir_binop_vector_extract)) {
          ir_rvalue *new_rhs =
             validate_assignment(state, lhs_loc, lhs->type,
                                 rhs, is_initializer);
@@ -749,12 +750,24 @@ do_assignment(exec_list *instructions, struct _mesa_glsl_parse_state *state,
          if (new_rhs == NULL) {
             return lhs;
          } else {
+            /* This converts:
+             * - LHS: (expression float vector_extract <vec> <channel>)
+             * - RHS: <scalar>
+             * into:
+             * - LHS: <vec>
+             * - RHS: (expression vec2 vector_insert <vec> <channel> <scalar>)
+             *
+             * The LHS type is now a vector instead of a scalar.  Since GLSL
+             * allows assignments to be used as rvalues, we need to re-extract
+             * the channel from assignment_temp when returning the rvalue.
+             */
+            extract_channel = lhs_expr->operands[1];
             rhs = new(ctx) ir_expression(ir_triop_vector_insert,
-                                         expr->operands[0]->type,
-                                         expr->operands[0],
+                                         lhs_expr->operands[0]->type,
+                                         lhs_expr->operands[0],
                                          new_rhs,
-                                         expr->operands[1]);
-            lhs = expr->operands[0]->clone(ctx, NULL);
+                                         extract_channel);
+            lhs = lhs_expr->operands[0]->clone(ctx, NULL);
          }
       }
    }
@@ -848,6 +861,11 @@ do_assignment(exec_list *instructions, struct _mesa_glsl_parse_state *state,
    if (!error_emitted)
       instructions->push_tail(new(ctx) ir_assignment(lhs, deref_var));
 
+   if (extract_channel) {
+      return new(ctx) ir_expression(ir_binop_vector_extract,
+                                    new(ctx) ir_dereference_variable(var),
+                                    extract_channel->clone(ctx, NULL));
+   }
    return new(ctx) ir_dereference_variable(var);
 }
 
@@ -2785,10 +2803,17 @@ validate_identifier(const char *identifier, YYLTYPE loc,
        *     "In addition, all identifiers containing two
        *      consecutive underscores (__) are reserved as
        *      possible future keywords."
+       *
+       * The intention is that names containing __ are reserved for internal
+       * use by the implementation, and names prefixed with GL_ are reserved
+       * for use by Khronos.  Names simply containing __ are dangerous to use,
+       * but should be allowed.
+       *
+       * A future version of the GLSL specification will clarify this.
        */
-      _mesa_glsl_error(&loc, state,
-                       "identifier `%s' uses reserved `__' string",
-                       identifier);
+      _mesa_glsl_warning(&loc, state,
+                         "identifier `%s' uses reserved `__' string",
+                         identifier);
    }
 }
 
@@ -3961,17 +3986,22 @@ ast_jump_statement::hir(exec_list *instructions,
 	 _mesa_glsl_error(& loc, state,
 			  "break may only appear in a loop or a switch");
       } else {
-	 /* For a loop, inline the for loop expression again,
-	  * since we don't know where near the end of
-	  * the loop body the normal copy of it
-	  * is going to be placed.
+	 /* For a loop, inline the for loop expression again, since we don't
+	  * know where near the end of the loop body the normal copy of it is
+	  * going to be placed.  Same goes for the condition for a do-while
+	  * loop.
 	  */
 	 if (state->loop_nesting_ast != NULL &&
-	     mode == ast_continue &&
-	     state->loop_nesting_ast->rest_expression) {
-	    state->loop_nesting_ast->rest_expression->hir(instructions,
-							  state);
-	 }
+	     mode == ast_continue) {
+            if (state->loop_nesting_ast->rest_expression) {
+               state->loop_nesting_ast->rest_expression->hir(instructions,
+                                                             state);
+            }
+            if (state->loop_nesting_ast->mode ==
+                ast_iteration_statement::ast_do_while) {
+               state->loop_nesting_ast->condition_to_hir(instructions, state);
+            }
+         }
 
 	 if (state->switch_state.is_switch_innermost &&
 	     mode == ast_break) {
@@ -4301,14 +4331,14 @@ ast_case_label::hir(exec_list *instructions,
 }
 
 void
-ast_iteration_statement::condition_to_hir(ir_loop *stmt,
+ast_iteration_statement::condition_to_hir(exec_list *instructions,
 					  struct _mesa_glsl_parse_state *state)
 {
    void *ctx = state;
 
    if (condition != NULL) {
       ir_rvalue *const cond =
-	 condition->hir(& stmt->body_instructions, state);
+	 condition->hir(instructions, state);
 
       if ((cond == NULL)
 	  || !cond->type->is_boolean() || !cond->type->is_scalar()) {
@@ -4329,7 +4359,7 @@ ast_iteration_statement::condition_to_hir(ir_loop *stmt,
 	    new(ctx) ir_loop_jump(ir_loop_jump::jump_break);
 
 	 if_stmt->then_instructions.push_tail(break_stmt);
-	 stmt->body_instructions.push_tail(if_stmt);
+	 instructions->push_tail(if_stmt);
       }
    }
 }
@@ -4364,7 +4394,7 @@ ast_iteration_statement::hir(exec_list *instructions,
    state->switch_state.is_switch_innermost = false;
 
    if (mode != ast_do_while)
-      condition_to_hir(stmt, state);
+      condition_to_hir(&stmt->body_instructions, state);
 
    if (body != NULL)
       body->hir(& stmt->body_instructions, state);
@@ -4373,7 +4403,7 @@ ast_iteration_statement::hir(exec_list *instructions,
       rest_expression->hir(& stmt->body_instructions, state);
 
    if (mode == ast_do_while)
-      condition_to_hir(stmt, state);
+      condition_to_hir(&stmt->body_instructions, state);
 
    if (mode != ast_do_while)
       state->symbols->pop_scope();

src/glsl/ast_type.cpp

@@ -123,6 +123,7 @@ ast_type_qualifier::merge_qualifier(YYLTYPE *loc,
    ubo_layout_mask.flags.q.shared = 1;
 
    ast_type_qualifier ubo_binding_mask;
+   ubo_binding_mask.flags.i = 0;
    ubo_binding_mask.flags.q.explicit_binding = 1;
    ubo_binding_mask.flags.q.explicit_offset = 1;
 

src/glsl/builtin_functions.cpp

@@ -3864,7 +3864,7 @@ builtin_builder::_fma(const glsl_type *type)
    ir_variable *c = in_var(type, "c");
    MAKE_SIG(type, gpu_shader5, 3, a, b, c);
 
-   body.emit(ret(fma(a, b, c)));
+   body.emit(ret(ir_builder::fma(a, b, c)));
 
    return sig;
 }
@@ -3997,6 +3997,7 @@ builtin_builder::_atomic_op(const char *intrinsic,
 
 /* The singleton instance of builtin_builder. */
 static builtin_builder builtins;
+_glthread_DECLARE_STATIC_MUTEX(builtins_lock);
 
 /**
  * External API (exposing the built-in module to the rest of the compiler):
@@ -4005,19 +4006,27 @@ static builtin_builder builtins;
 void
 _mesa_glsl_initialize_builtin_functions()
 {
+   _glthread_LOCK_MUTEX(builtins_lock);
    builtins.initialize();
+   _glthread_UNLOCK_MUTEX(builtins_lock);
 }
 
 void
 _mesa_glsl_release_builtin_functions()
 {
+   _glthread_LOCK_MUTEX(builtins_lock);
    builtins.release();
+   _glthread_UNLOCK_MUTEX(builtins_lock);
 }
 
 ir_function_signature *
 _mesa_glsl_find_builtin_function(_mesa_glsl_parse_state *state,
                                  const char *name, exec_list *actual_parameters)
 {
-   return builtins.find(state, name, actual_parameters);
+   ir_function_signature * s;
+   _glthread_LOCK_MUTEX(builtins_lock);
+   s = builtins.find(state, name, actual_parameters);
+   _glthread_UNLOCK_MUTEX(builtins_lock);
+   return s;
 }
 /** @} */

src/glsl/glcpp/glcpp-parse.y

@@ -1272,9 +1272,6 @@ glcpp_parser_create (const struct gl_extensions *extensions, int api)
 	      if (extensions->ARB_sample_shading)
 	         add_builtin_define(parser, "GL_ARB_sample_shading", 1);
 
-	      if (extensions->EXT_shader_integer_mix)
-	         add_builtin_define(parser, "GL_EXT_shader_integer_mix", 1);
-
 	      if (extensions->ARB_texture_gather)
 	         add_builtin_define(parser, "GL_ARB_texture_gather", 1);
 
@@ -1283,6 +1280,11 @@ glcpp_parser_create (const struct gl_extensions *extensions, int api)
 	   }
 	}
 
+	if (extensions != NULL) {
+	   if (extensions->EXT_shader_integer_mix)
+	      add_builtin_define(parser, "GL_EXT_shader_integer_mix", 1);
+	}
+
 	language_version = 110;
 	add_builtin_define(parser, "__VERSION__", language_version);
 
@@ -1826,11 +1828,27 @@ static void
 _check_for_reserved_macro_name (glcpp_parser_t *parser, YYLTYPE *loc,
 				const char *identifier)
 {
-	/* According to the GLSL specification, macro names starting with "__"
-	 * or "GL_" are reserved for future use.  So, don't allow them.
+	/* Section 3.3 (Preprocessor) of the GLSL 1.30 spec (and later) and
+	 * the GLSL ES spec (all versions) say:
+	 *
+	 *     "All macro names containing two consecutive underscores ( __ )
+	 *     are reserved for future use as predefined macro names. All
+	 *     macro names prefixed with "GL_" ("GL" followed by a single
+	 *     underscore) are also reserved."
+	 *
+	 * The intention is that names containing __ are reserved for internal
+	 * use by the implementation, and names prefixed with GL_ are reserved
+	 * for use by Khronos.  Since every extension adds a name prefixed
+	 * with GL_ (i.e., the name of the extension), that should be an
+	 * error.  Names simply containing __ are dangerous to use, but should
+	 * be allowed.
+	 *
+	 * A future version of the GLSL specification will clarify this.
 	 */
 	if (strstr(identifier, "__")) {
-		glcpp_error (loc, parser, "Macro names containing \"__\" are reserved.\n");
+		glcpp_warning(loc, parser,
+			      "Macro names containing \"__\" are reserved "
+			      "for use by the implementation.\n");
 	}
 	if (strncmp(identifier, "GL_", 3) == 0) {
 		glcpp_error (loc, parser, "Macro names starting with \"GL_\" are reserved.\n");

src/glsl/glcpp/tests/086-reserved-macro-names.c.expected

@@ -1,8 +1,8 @@
-0:1(10): preprocessor error: Macro names containing "__" are reserved.
+0:1(10): preprocessor warning: Macro names containing "__" are reserved for use by the implementation.
 
 0:2(9): preprocessor error: Macro names starting with "GL_" are reserved.
 
-0:3(9): preprocessor error: Macro names containing "__" are reserved.
+0:3(9): preprocessor warning: Macro names containing "__" are reserved for use by the implementation.
 
 
 

src/glsl/glsl_parser.yy

@@ -1530,7 +1530,7 @@ type_qualifier:
                           "just before storage qualifiers");
       }
       $$ = $1;
-      $$.flags.i |= $2.flags.i;
+      $$.merge_qualifier(&@1, state, $2);
    }
    | storage_qualifier type_qualifier
    {

src/glsl/glsl_parser_extras.cpp

@@ -292,6 +292,10 @@ _mesa_glsl_parse_state::process_version_directive(YYLTYPE *locp, int version,
       }
    }
 
+   if (this->es_shader) {
+      this->ARB_texture_rectangle_enable = false;
+   }
+
    this->language_version = version;
 
    bool supported = false;

src/glsl/ir_constant_expression.cpp

@@ -1402,7 +1402,7 @@ ir_expression::constant_expression_value(struct hash_table *variable_context)
          data.f[c] = ldexp(op[0]->value.f[c], op[1]->value.i[c]);
          /* Flush subnormal values to zero. */
          if (!isnormal(data.f[c]))
-            data.f[c] = copysign(0.0, op[0]->value.f[c]);
+            data.f[c] = copysign(0.0f, op[0]->value.f[c]);
       }
       break;
 

src/glsl/linker.cpp

@@ -2373,11 +2373,12 @@ link_shaders(struct gl_context *ctx, struct gl_shader_program *prog)
       goto done;
 
    /* OpenGL ES requires that a vertex shader and a fragment shader both be
-    * present in a linked program.  By checking prog->IsES, we also
-    * catch the GL_ARB_ES2_compatibility case.
+    * present in a linked program. GL_ARB_ES2_compatibility doesn't say
+    * anything about shader linking when one of the shaders (vertex or
+    * fragment shader) is absent. So, the extension shouldn't change the
+    * behavior specified in GLSL specification.
     */
-   if (!prog->InternalSeparateShader &&
-       (ctx->API == API_OPENGLES2 || prog->IsES)) {
+   if (!prog->InternalSeparateShader && ctx->API == API_OPENGLES2) {
       if (prog->_LinkedShaders[MESA_SHADER_VERTEX] == NULL) {
 	 linker_error(prog, "program lacks a vertex shader\n");
       } else if (prog->_LinkedShaders[MESA_SHADER_FRAGMENT] == NULL) {

src/glx/dri2_query_renderer.c

@@ -20,6 +20,9 @@
  * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
  * DEALINGS IN THE SOFTWARE.
  */
+
+#if defined(GLX_DIRECT_RENDERING) && !defined(GLX_USE_APPLEGL)
+
 #include "glxclient.h"
 #include "glx_error.h"
 #include "xf86drm.h"
@@ -95,3 +98,5 @@ dri2_query_renderer_string(struct glx_screen *base, int attribute,
 
    return psc->rendererQuery->queryString(psc->driScreen, dri_attribute, value);
 }
+
+#endif /* GLX_DIRECT_RENDERING */

src/glx/glxcmds.c

@@ -917,13 +917,10 @@ init_fbconfig_for_chooser(struct glx_config * config,
    if (fbconfig_style_tags) {
       config->rgbMode = GL_TRUE;
       config->doubleBufferMode = GLX_DONT_CARE;
-      /* allow any kind of drawable, including those for off-screen buffers */
-      config->drawableType = 0;
-   } else {
-       /* allow configs which support on-screen drawing */
-       config->drawableType = GLX_WINDOW_BIT;
+      config->renderType = GLX_RGBA_BIT;
    }
 
+   config->drawableType = GLX_WINDOW_BIT;
    config->visualRating = GLX_DONT_CARE;
    config->transparentPixel = GLX_NONE;
    config->transparentRed = GLX_DONT_CARE;
@@ -932,8 +929,6 @@ init_fbconfig_for_chooser(struct glx_config * config,
    config->transparentAlpha = GLX_DONT_CARE;
    config->transparentIndex = GLX_DONT_CARE;
 
-   /* Set GLX_RENDER_TYPE property to not expect any flags by default. */
-   config->renderType = 0;
    config->xRenderable = GLX_DONT_CARE;
    config->fbconfigID = (GLXFBConfigID) (GLX_DONT_CARE);
 
@@ -1102,7 +1097,7 @@ static int
 fbconfig_compare(struct glx_config **a, struct glx_config **b)
 {
    /* The order of these comparisons must NOT change.  It is defined by
-    * the GLX 1.3 spec and ARB_multisample.
+    * the GLX 1.4 specification.
     */
 
    PREFER_SMALLER(visualSelectGroup);
@@ -1131,6 +1126,9 @@ fbconfig_compare(struct glx_config **a, struct glx_config **b)
 
    PREFER_SMALLER(numAuxBuffers);
 
+   PREFER_SMALLER(sampleBuffers);
+   PREFER_SMALLER(samples);
+
    PREFER_LARGER_OR_ZERO(depthBits);
    PREFER_SMALLER(stencilBits);
 
@@ -1144,12 +1142,6 @@ fbconfig_compare(struct glx_config **a, struct glx_config **b)
 
    PREFER_SMALLER(visualType);
 
-   /* None of the multisample specs say where this comparison should happen,
-    * so I put it near the end.
-    */
-   PREFER_SMALLER(sampleBuffers);
-   PREFER_SMALLER(samples);
-
    /* None of the pbuffer or fbconfig specs say that this comparison needs
     * to happen at all, but it seems like it should.
     */

src/glx/glxext.c

@@ -823,7 +823,6 @@ __glXInitialize(Display * dpy)
    dpyPriv->codes = XInitExtension(dpy, __glXExtensionName);
    if (!dpyPriv->codes) {
       free(dpyPriv);
-      _XUnlockMutex(_Xglobal_lock);
       return NULL;
    }
 
@@ -839,7 +838,6 @@ __glXInitialize(Display * dpy)
 		     &dpyPriv->majorVersion, &dpyPriv->minorVersion)
        || (dpyPriv->majorVersion == 1 && dpyPriv->minorVersion < 1)) {
       free(dpyPriv);
-      _XUnlockMutex(_Xglobal_lock);
       return NULL;
    }
 
@@ -900,7 +898,7 @@ __glXInitialize(Display * dpy)
    dpyPriv->next = glx_displays;
    glx_displays = dpyPriv;
 
-    _XUnlockMutex(_Xglobal_lock);
+   _XUnlockMutex(_Xglobal_lock);
 
    return dpyPriv;
 }

src/mesa/Makefile.am

@@ -95,6 +95,12 @@ AM_CPPFLAGS = $(DEFINES) $(INCLUDE_DIRS)
 AM_CFLAGS = $(LLVM_CFLAGS) $(VISIBILITY_CFLAGS)
 AM_CXXFLAGS = $(LLVM_CFLAGS) $(VISIBILITY_CXXFLAGS)
 
+ARCH_LIBS =
+
+if SSE41_SUPPORTED
+ARCH_LIBS += libmesa_sse41.la
+endif
+
 MESA_ASM_FILES_FOR_ARCH =
 
 if HAVE_X86_ASM
@@ -103,8 +109,6 @@ noinst_PROGRAMS = gen_matypes
 gen_matypes_SOURCES = x86/gen_matypes.c
 BUILT_SOURCES += matypes.h
 
-ARCH_LIBS = libmesa_sse41.la
-
 if HAVE_X86_64_ASM
 MESA_ASM_FILES_FOR_ARCH += $(X86_64_FILES)
 AM_CPPFLAGS += -I$(builddir)/x86-64 -I$(srcdir)/x86-64

src/mesa/drivers/common/meta.c

@@ -355,6 +355,8 @@ static void cleanup_temp_texture(struct gl_context *ctx, struct temp_texture *te
 static void meta_glsl_clear_cleanup(struct gl_context *ctx, struct clear_state *clear);
 static void meta_glsl_generate_mipmap_cleanup(struct gl_context *ctx,
                                               struct gen_mipmap_state *mipmap);
+static void meta_decompress_cleanup(struct decompress_state *decompress);
+static void meta_drawpix_cleanup(struct drawpix_state *drawpix);
 
 static GLuint
 compile_shader_with_debug(struct gl_context *ctx, GLenum target, const GLcharARB *source)
@@ -449,6 +451,8 @@ _mesa_meta_free(struct gl_context *ctx)
    meta_glsl_clear_cleanup(ctx, &ctx->Meta->Clear);
    meta_glsl_generate_mipmap_cleanup(ctx, &ctx->Meta->Mipmap);
    cleanup_temp_texture(ctx, &ctx->Meta->TempTex);
+   meta_decompress_cleanup(&ctx->Meta->Decompress);
+   meta_drawpix_cleanup(&ctx->Meta->DrawPix);
    if (old_context)
       _mesa_make_current(old_context, old_context->WinSysDrawBuffer, old_context->WinSysReadBuffer);
    else
@@ -2489,7 +2493,24 @@ _mesa_meta_CopyPixels(struct gl_context *ctx, GLint srcX, GLint srcY,
    _mesa_meta_end(ctx);
 }
 
+static void
+meta_drawpix_cleanup(struct drawpix_state *drawpix)
+{
+   if (drawpix->ArrayObj != 0) {
+      _mesa_DeleteVertexArrays(1, &drawpix->ArrayObj);
+      drawpix->ArrayObj = 0;
+   }
 
+   if (drawpix->StencilFP != 0) {
+      _mesa_DeleteProgramsARB(1, &drawpix->StencilFP);
+      drawpix->StencilFP = 0;
+   }
+
+   if (drawpix->DepthFP != 0) {
+      _mesa_DeleteProgramsARB(1, &drawpix->DepthFP);
+      drawpix->DepthFP = 0;
+   }
+}
 
 /**
  * When the glDrawPixels() image size is greater than the max rectangle
@@ -2951,8 +2972,8 @@ _mesa_meta_Bitmap(struct gl_context *ctx,
       /* one-time setup */
 
       /* create vertex array object */
-      _mesa_GenVertexArraysAPPLE(1, &bitmap->ArrayObj);
-      _mesa_BindVertexArrayAPPLE(bitmap->ArrayObj);
+      _mesa_GenVertexArrays(1, &bitmap->ArrayObj);
+      _mesa_BindVertexArray(bitmap->ArrayObj);
 
       /* create vertex array buffer */
       _mesa_GenBuffers(1, &bitmap->VBO);
@@ -3317,8 +3338,8 @@ setup_ff_generate_mipmap(struct gl_context *ctx,
    if (mipmap->ArrayObj == 0) {
       /* one-time setup */
       /* create vertex array object */
-      _mesa_GenVertexArraysAPPLE(1, &mipmap->ArrayObj);
-      _mesa_BindVertexArrayAPPLE(mipmap->ArrayObj);
+      _mesa_GenVertexArrays(1, &mipmap->ArrayObj);
+      _mesa_BindVertexArray(mipmap->ArrayObj);
 
       /* create vertex array buffer */
       _mesa_GenBuffers(1, &mipmap->VBO);
@@ -3891,6 +3912,25 @@ _mesa_meta_CopyTexSubImage(struct gl_context *ctx, GLuint dims,
 }
 
 
+static void
+meta_decompress_cleanup(struct decompress_state *decompress)
+{
+   if (decompress->FBO != 0) {
+      _mesa_DeleteFramebuffers(1, &decompress->FBO);
+      _mesa_DeleteRenderbuffers(1, &decompress->RBO);
+   }
+
+   if (decompress->ArrayObj != 0) {
+      _mesa_DeleteVertexArrays(1, &decompress->ArrayObj);
+      _mesa_DeleteBuffers(1, &decompress->VBO);
+   }
+
+   if (decompress->Sampler != 0)
+      _mesa_DeleteSamplers(1, &decompress->Sampler);
+
+   memset(decompress, 0, sizeof(*decompress));
+}
+
 /**
  * Decompress a texture image by drawing a quad with the compressed
  * texture and reading the pixels out of the color buffer.
@@ -4125,7 +4165,8 @@ _mesa_meta_GetTexImage(struct gl_context *ctx,
     * unsigned, normalized values.  We could handle signed and unnormalized 
     * with floating point renderbuffers...
     */
-   if (_mesa_is_format_compressed(texImage->TexFormat) &&
+   if (texImage->TexObject->Target != GL_TEXTURE_CUBE_MAP_ARRAY
+       && _mesa_is_format_compressed(texImage->TexFormat) &&
        _mesa_get_format_datatype(texImage->TexFormat)
        == GL_UNSIGNED_NORMALIZED) {
       struct gl_texture_object *texObj = texImage->TexObject;

src/mesa/drivers/dri/i915/intel_screen.c

@@ -741,7 +741,7 @@ i915_query_renderer_integer(__DRIscreen *psp, int param, unsigned int *value)
          * (uint64_t) system_page_size;
 
       const unsigned system_memory_megabytes =
-         (unsigned) (system_memory_bytes / 1024);
+         (unsigned) (system_memory_bytes / (1024 * 1024));
 
       value[0] = MIN2(system_memory_megabytes, gpu_mappable_megabytes);
       return 0;

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

@@ -266,7 +266,6 @@ retry:
     */
    brw->state.dirty.brw = ~0;
    brw->state.dirty.cache = ~0;
-   brw->batch.need_workaround_flush = true;
    brw->ib.type = -1;
    intel_batchbuffer_clear_cache(brw);
 

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

@@ -699,12 +699,6 @@ brwCreateContext(gl_api api,
 
    intel_batchbuffer_init(brw);
 
-   brw_init_state(brw);
-
-   intelInitExtensions(ctx);
-
-   intel_fbo_init(brw);
-
    if (brw->gen >= 6) {
       /* Create a new hardware context.  Using a hardware context means that
        * our GPU state will be saved/restored on context switch, allowing us
@@ -722,6 +716,12 @@ brwCreateContext(gl_api api,
       }
    }
 
+   brw_init_state(brw);
+
+   intelInitExtensions(ctx);
+
+   intel_fbo_init(brw);
+
    brw_init_surface_formats(brw);
 
    if (brw->is_g4x || brw->gen >= 5) {

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

@@ -273,7 +273,7 @@ brw_predraw_resolve_buffers(struct brw_context *brw)
    /* Resolve depth buffer of each enabled depth texture, and color buffer of
     * each fast-clear-enabled color texture.
     */
-   for (int i = 0; i < BRW_MAX_TEX_UNIT; i++) {
+   for (int i = 0; i < ctx->Const.MaxCombinedTextureImageUnits; i++) {
       if (!ctx->Texture.Unit[i]._ReallyEnabled)
 	 continue;
       tex_obj = intel_texture_object(ctx->Texture.Unit[i]._Current);
@@ -445,6 +445,12 @@ retry:
 	    }
 	 }
       }
+
+      /* Now that we know we haven't run out of aperture space, we can safely
+       * reset the dirty bits.
+       */
+      if (brw->state.dirty.brw)
+         brw_clear_dirty_bits(brw);
    }
 
    if (brw->always_flush_batch)

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

@@ -1008,7 +1008,7 @@ fs_visitor::emit_fragcoord_interpolation(ir_variable *ir)
 fs_inst *
 fs_visitor::emit_linterp(const fs_reg &attr, const fs_reg &interp,
                          glsl_interp_qualifier interpolation_mode,
-                         bool is_centroid)
+                         bool is_centroid, bool is_sample)
 {
    brw_wm_barycentric_interp_mode barycoord_mode;
    if (brw->gen >= 6) {
@@ -1017,6 +1017,11 @@ fs_visitor::emit_linterp(const fs_reg &attr, const fs_reg &interp,
             barycoord_mode = BRW_WM_PERSPECTIVE_CENTROID_BARYCENTRIC;
          else
             barycoord_mode = BRW_WM_NONPERSPECTIVE_CENTROID_BARYCENTRIC;
+      } else if (is_sample) {
+          if (interpolation_mode == INTERP_QUALIFIER_SMOOTH)
+            barycoord_mode = BRW_WM_PERSPECTIVE_SAMPLE_BARYCENTRIC;
+         else
+            barycoord_mode = BRW_WM_NONPERSPECTIVE_SAMPLE_BARYCENTRIC;
       } else {
          if (interpolation_mode == INTERP_QUALIFIER_SMOOTH)
             barycoord_mode = BRW_WM_PERSPECTIVE_PIXEL_BARYCENTRIC;
@@ -1094,7 +1099,9 @@ fs_visitor::emit_general_interpolation(ir_variable *ir)
 		*/
                struct brw_reg interp = interp_reg(location, k);
                emit_linterp(attr, fs_reg(interp), interpolation_mode,
-                            ir->centroid);
+                            ir->centroid && !c->key.persample_shading,
+                            c->key.persample_shading);
+
                if (brw->needs_unlit_centroid_workaround && ir->centroid) {
                   /* Get the pixel/sample mask into f0 so that we know
                    * which pixels are lit.  Then, for each channel that is
@@ -1103,7 +1110,8 @@ fs_visitor::emit_general_interpolation(ir_variable *ir)
                    */
                   emit(FS_OPCODE_MOV_DISPATCH_TO_FLAGS);
                   fs_inst *inst = emit_linterp(attr, fs_reg(interp),
-                                               interpolation_mode, false);
+                                               interpolation_mode,
+                                               false, false);
                   inst->predicate = BRW_PREDICATE_NORMAL;
                   inst->predicate_inverse = true;
                }

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

@@ -338,7 +338,7 @@ public:
    fs_reg *emit_fragcoord_interpolation(ir_variable *ir);
    fs_inst *emit_linterp(const fs_reg &attr, const fs_reg &interp,
                          glsl_interp_qualifier interpolation_mode,
-                         bool is_centroid);
+                         bool is_centroid, bool is_sample);
    fs_reg *emit_frontfacing_interpolation(ir_variable *ir);
    fs_reg *emit_samplepos_setup(ir_variable *ir);
    fs_reg *emit_sampleid_setup(ir_variable *ir);

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

@@ -55,7 +55,7 @@ brw_lower_unnormalized_offset_visitor::visit_leave(ir_texture *ir)
    if (!ir->offset)
       return visit_continue;
 
-   if (ir->op == ir_tg4) {
+   if (ir->op == ir_tg4 || ir->op == ir_tex) {
       if (ir->sampler->type->sampler_dimensionality != GLSL_SAMPLER_DIM_RECT)
          return visit_continue;
    }

src/mesa/drivers/dri/i965/brw_state.h

@@ -143,6 +143,7 @@ brw_depthbuffer_format(struct brw_context *brw);
  * brw_state.c
  */
 void brw_upload_state(struct brw_context *brw);
+void brw_clear_dirty_bits(struct brw_context *brw);
 void brw_init_state(struct brw_context *brw);
 void brw_destroy_state(struct brw_context *brw);
 

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

@@ -573,6 +573,20 @@ void brw_upload_state(struct brw_context *brw)
 	 fprintf(stderr, "\n");
       }
    }
+}
 
+
+/**
+ * Clear dirty bits to account for the fact that the state emitted by
+ * brw_upload_state() has been committed to the hardware.  This is a separate
+ * call from brw_upload_state() because it's possible that after the call to
+ * brw_upload_state(), we will discover that we've run out of aperture space,
+ * and need to rewind the batch buffer to the state it had before the
+ * brw_upload_state() call.
+ */
+void
+brw_clear_dirty_bits(struct brw_context *brw)
+{
+   struct brw_state_flags *state = &brw->state.dirty;
    memset(state, 0, sizeof(*state));
 }

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

@@ -47,7 +47,7 @@ void brw_validate_textures( struct brw_context *brw )
    struct gl_context *ctx = &brw->ctx;
    int i;
 
-   for (i = 0; i < BRW_MAX_TEX_UNIT; i++) {
+   for (i = 0; i < ctx->Const.MaxCombinedTextureImageUnits; i++) {
       struct gl_texture_unit *texUnit = &ctx->Texture.Unit[i];
 
       if (texUnit->_ReallyEnabled) {

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

@@ -105,11 +105,11 @@ intel_vertical_texture_alignment_unit(struct brw_context *brw,
     * | Depth Buffer                           |  2  |  2  |  2  |  4  |  4  |
     * | Separate Stencil Buffer                | N/A | N/A | N/A |  4  |  8  |
     * | Multisampled (4x or 8x) render target  | N/A | N/A | N/A |  4  |  4  |
-    * | All Others                             |  2  |  2  |  2  |  2  |  2  |
+    * | All Others                             |  2  |  2  |  2  |  *  |  *  |
     * +----------------------------------------------------------------------+
     *
-    * On SNB+, non-special cases can be overridden by setting the SURFACE_STATE
-    * "Surface Vertical Alignment" field to VALIGN_2 or VALIGN_4.
+    * Where "*" means either VALIGN_2 or VALIGN_4 depending on the setting of
+    * the SURFACE_STATE "Surface Vertical Alignment" field.
     */
    if (_mesa_is_format_compressed(format))
       return 4;
@@ -128,6 +128,25 @@ intel_vertical_texture_alignment_unit(struct brw_context *brw,
       return 4;
    }
 
+   if (brw->gen == 7) {
+      /* On Gen7, we prefer a vertical alignment of 4 when possible, because
+       * that allows Y tiled render targets.
+       *
+       * From the Ivy Bridge PRM, Vol4 Part1 2.12.2.1 (SURFACE_STATE for most
+       * messages), on p64, under the heading "Surface Vertical Alignment":
+       *
+       *     Value of 1 [VALIGN_4] is not supported for format YCRCB_NORMAL
+       *     (0x182), YCRCB_SWAPUVY (0x183), YCRCB_SWAPUV (0x18f), YCRCB_SWAPY
+       *     (0x190)
+       *
+       *     VALIGN_4 is not supported for surface format R32G32B32_FLOAT.
+       */
+      if (base_format == GL_YCBCR_MESA || format == MESA_FORMAT_RGB_FLOAT32)
+         return 2;
+
+      return 4;
+   }
+
    return 2;
 }
 

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

@@ -48,6 +48,7 @@
 static unsigned
 brw_compute_barycentric_interp_modes(struct brw_context *brw,
                                      bool shade_model_flat,
+                                     bool persample_shading,
                                      const struct gl_fragment_program *fprog)
 {
    unsigned barycentric_interp_modes = 0;
@@ -60,7 +61,9 @@ brw_compute_barycentric_interp_modes(struct brw_context *brw,
    for (attr = 0; attr < VARYING_SLOT_MAX; ++attr) {
       enum glsl_interp_qualifier interp_qualifier =
          fprog->InterpQualifier[attr];
-      bool is_centroid = fprog->IsCentroid & BITFIELD64_BIT(attr);
+      bool is_centroid = (fprog->IsCentroid & BITFIELD64_BIT(attr)) &&
+         !persample_shading;
+      bool is_sample = persample_shading;
       bool is_gl_Color = attr == VARYING_SLOT_COL0 || attr == VARYING_SLOT_COL1;
 
       /* Ignore unused inputs. */
@@ -81,8 +84,12 @@ brw_compute_barycentric_interp_modes(struct brw_context *brw,
          if (is_centroid) {
             barycentric_interp_modes |=
                1 << BRW_WM_NONPERSPECTIVE_CENTROID_BARYCENTRIC;
+         } else if (is_sample) {
+            barycentric_interp_modes |=
+               1 << BRW_WM_NONPERSPECTIVE_SAMPLE_BARYCENTRIC;
          }
-         if (!is_centroid || brw->needs_unlit_centroid_workaround) {
+         if ((!is_centroid && !is_sample) ||
+             brw->needs_unlit_centroid_workaround) {
             barycentric_interp_modes |=
                1 << BRW_WM_NONPERSPECTIVE_PIXEL_BARYCENTRIC;
          }
@@ -92,8 +99,12 @@ brw_compute_barycentric_interp_modes(struct brw_context *brw,
          if (is_centroid) {
             barycentric_interp_modes |=
                1 << BRW_WM_PERSPECTIVE_CENTROID_BARYCENTRIC;
+         } else if (is_sample) {
+            barycentric_interp_modes |=
+               1 << BRW_WM_PERSPECTIVE_SAMPLE_BARYCENTRIC;
          }
-         if (!is_centroid || brw->needs_unlit_centroid_workaround) {
+         if ((!is_centroid && !is_sample) ||
+             brw->needs_unlit_centroid_workaround) {
             barycentric_interp_modes |=
                1 << BRW_WM_PERSPECTIVE_PIXEL_BARYCENTRIC;
          }
@@ -170,6 +181,7 @@ bool do_wm_prog(struct brw_context *brw,
 
    c->prog_data.barycentric_interp_modes =
       brw_compute_barycentric_interp_modes(brw, c->key.flat_shade,
+                                           c->key.persample_shading,
                                            &fp->program);
 
    program = brw_wm_fs_emit(brw, c, &fp->program, prog, &program_size);
@@ -490,6 +502,9 @@ static void brw_wm_populate_key( struct brw_context *brw,
       (ctx->Multisample.SampleAlphaToCoverage || ctx->Color.AlphaEnabled);
 
    /* _NEW_BUFFERS _NEW_MULTISAMPLE */
+   key->persample_shading =
+      _mesa_get_min_invocations_per_fragment(ctx, &fp->program) > 1;
+
    key->compute_pos_offset =
       _mesa_get_min_invocations_per_fragment(ctx, &fp->program) > 1 &&
       fp->program.Base.SystemValuesRead & SYSTEM_BIT_SAMPLE_POS;

src/mesa/drivers/dri/i965/brw_wm.h

@@ -61,6 +61,7 @@ struct brw_wm_prog_key {
    uint8_t iz_lookup;
    GLuint stats_wm:1;
    GLuint flat_shade:1;
+   GLuint persample_shading:1;
    GLuint nr_color_regions:5;
    GLuint replicate_alpha:1;
    GLuint render_to_fbo:1;

src/mesa/drivers/dri/i965/gen6_blorp.cpp

@@ -254,26 +254,6 @@ gen6_blorp_emit_blend_state(struct brw_context *brw,
    blend->blend1.write_disable_b = params->color_write_disable[2];
    blend->blend1.write_disable_a = params->color_write_disable[3];
 
-   /* When blitting from an XRGB source to a ARGB destination, we need to
-    * interpret the missing channel as 1.0.  Blending can do that for us:
-    * we simply use the RGB values from the fragment shader ("source RGB"),
-    * but smash the alpha channel to 1.
-    */
-   if (params->src.mt &&
-       _mesa_get_format_bits(params->dst.mt->format, GL_ALPHA_BITS) > 0 &&
-       _mesa_get_format_bits(params->src.mt->format, GL_ALPHA_BITS) == 0) {
-      blend->blend0.blend_enable = 1;
-      blend->blend0.ia_blend_enable = 1;
-
-      blend->blend0.blend_func = BRW_BLENDFUNCTION_ADD;
-      blend->blend0.ia_blend_func = BRW_BLENDFUNCTION_ADD;
-
-      blend->blend0.source_blend_factor = BRW_BLENDFACTOR_SRC_COLOR;
-      blend->blend0.dest_blend_factor = BRW_BLENDFACTOR_ZERO;
-      blend->blend0.ia_source_blend_factor = BRW_BLENDFACTOR_ONE;
-      blend->blend0.ia_dest_blend_factor = BRW_BLENDFACTOR_ZERO;
-   }
-
    return cc_blend_state_offset;
 }
 
@@ -1010,20 +990,9 @@ gen6_blorp_emit_primitive(struct brw_context *brw,
    OUT_BATCH(0);
    OUT_BATCH(0);
    ADVANCE_BATCH();
-}
 
-static void
-gen6_emit_hiz_workaround(struct brw_context *brw, enum gen6_hiz_op hiz_op)
-{
-   /* This fixes a HiZ hang in WebGL Google Maps. A more minimal fix likely
-    * exists, but this gets the job done.
-    */
-   if (hiz_op == GEN6_HIZ_OP_DEPTH_RESOLVE ||
-       hiz_op == GEN6_HIZ_OP_HIZ_RESOLVE) {
-      brw->batch.need_workaround_flush = true;
-      intel_emit_post_sync_nonzero_flush(brw);
-      intel_emit_depth_stall_flushes(brw);
-   }
+   /* Only used on Sandybridge; harmless to set elsewhere. */
+   brw->batch.need_workaround_flush = true;
 }
 
 /**
@@ -1048,7 +1017,9 @@ gen6_blorp_exec(struct brw_context *brw,
 
    uint32_t prog_offset = params->get_wm_prog(brw, &prog_data);
 
-   gen6_emit_hiz_workaround(brw, params->hiz_op);
+   /* Emit workaround flushes when we switch from drawing to blorping. */
+   brw->batch.need_workaround_flush = true;
+
    gen6_emit_3dstate_multisample(brw, params->num_samples);
    gen6_emit_3dstate_sample_mask(brw, params->num_samples, 1.0, false, ~0u);
    gen6_blorp_emit_state_base_address(brw, params);

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

@@ -199,9 +199,9 @@ intel_miptree_blit(struct brw_context *brw,
     * As a result of these two limitations, we can only use the blitter to do
     * this copy when the region's pitch is less than 32k.
     */
-   if (src_mt->region->pitch > 32768 ||
-       dst_mt->region->pitch > 32768) {
-      perf_debug("Falling back due to >32k pitch\n");
+   if (src_mt->region->pitch >= 32768 ||
+       dst_mt->region->pitch >= 32768) {
+      perf_debug("Falling back due to >=32k pitch\n");
       return false;
    }
 

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

@@ -357,7 +357,8 @@ intel_bufferobj_map_range(struct gl_context * ctx,
     * BO, and we'll copy what they put in there out at unmap or
     * FlushRange time.
     */
-   if ((access & GL_MAP_INVALIDATE_RANGE_BIT) &&
+   if (!(access & GL_MAP_UNSYNCHRONIZED_BIT) &&
+       (access & GL_MAP_INVALIDATE_RANGE_BIT) &&
        drm_intel_bo_busy(intel_obj->buffer)) {
       if (access & GL_MAP_FLUSH_EXPLICIT_BIT) {
 	 intel_obj->range_map_buffer = malloc(length);

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

@@ -1252,7 +1252,7 @@ intel_miptree_alloc_mcs(struct brw_context *brw,
     * Note: the clear value for MCS buffers is all 1's, so we memset to 0xff.
     */
    void *data = intel_miptree_map_raw(brw, mt->mcs_mt);
-   memset(data, 0xff, mt->mcs_mt->region->bo->size);
+   memset(data, 0xff, mt->mcs_mt->region->height * mt->mcs_mt->region->pitch);
    intel_miptree_unmap_raw(brw, mt->mcs_mt);
 
    return mt->mcs_mt;

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

@@ -106,13 +106,15 @@ do_blit_readpixels(struct gl_context * ctx,
    /* Mesa flips the dst_stride for pack->Invert, but we want our mt to have a
     * normal dst_stride.
     */
+   struct gl_pixelstore_attrib uninverted_pack = *pack;
    if (pack->Invert) {
       dst_stride = -dst_stride;
       dst_flip = true;
+      uninverted_pack.Invert = false;
    }
 
    dst_offset = (GLintptr)pixels;
-   dst_offset += _mesa_image_offset(2, pack, width, height,
+   dst_offset += _mesa_image_offset(2, &uninverted_pack, width, height,
 				    format, type, 0, 0, 0);
 
    if (!_mesa_clip_copytexsubimage(ctx,

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

@@ -843,7 +843,7 @@ brw_query_renderer_integer(__DRIscreen *psp, int param, unsigned int *value)
          * (uint64_t) system_page_size;
 
       const unsigned system_memory_megabytes =
-         (unsigned) (system_memory_bytes / 1024);
+         (unsigned) (system_memory_bytes / (1024 * 1024));
 
       value[0] = MIN2(system_memory_megabytes, gpu_mappable_megabytes);
       return 0;

src/mesa/drivers/dri/nouveau/nouveau_context.c

@@ -72,7 +72,7 @@ nouveau_context_create(gl_api api,
 		return false;
 	}
 
-	ctx = screen->driver->context_create(screen, visual, share_ctx);
+	ctx = screen->driver->context_create(screen, api, visual, share_ctx);
 	if (!ctx) {
 		*error = __DRI_CTX_ERROR_NO_MEMORY;
 		return GL_FALSE;
@@ -107,7 +107,8 @@ nouveau_context_create(gl_api api,
 }
 
 GLboolean
-nouveau_context_init(struct gl_context *ctx, struct nouveau_screen *screen,
+nouveau_context_init(struct gl_context *ctx, gl_api api,
+		     struct nouveau_screen *screen,
 		     const struct gl_config *visual, struct gl_context *share_ctx)
 {
 	struct nouveau_context *nctx = to_nouveau_context(ctx);
@@ -125,7 +126,7 @@ nouveau_context_init(struct gl_context *ctx, struct nouveau_screen *screen,
 	nouveau_fbo_functions_init(&functions);
 
 	/* Initialize the mesa context. */
-	_mesa_initialize_context(ctx, API_OPENGL_COMPAT, visual,
+	_mesa_initialize_context(ctx, api, visual,
                                  share_ctx, &functions);
 
 	nouveau_state_init(ctx);

src/mesa/drivers/dri/nouveau/nouveau_context.h

@@ -115,7 +115,8 @@ nouveau_context_create(gl_api api,
 		       void *share_ctx);
 
 GLboolean
-nouveau_context_init(struct gl_context *ctx, struct nouveau_screen *screen,
+nouveau_context_init(struct gl_context *ctx, gl_api api,
+		     struct nouveau_screen *screen,
 		     const struct gl_config *visual, struct gl_context *share_ctx);
 
 void

src/mesa/drivers/dri/nouveau/nouveau_driver.h

@@ -48,6 +48,7 @@
 
 struct nouveau_driver {
 	struct gl_context *(*context_create)(struct nouveau_screen *screen,
+				     gl_api api,
 				     const struct gl_config *visual,
 				     struct gl_context *share_ctx);
 	void (*context_destroy)(struct gl_context *ctx);

src/mesa/drivers/dri/nouveau/nouveau_fbo.c

@@ -265,7 +265,8 @@ static void
 nouveau_finish_render_texture(struct gl_context *ctx,
 			      struct gl_renderbuffer *rb)
 {
-	texture_dirty(rb->TexImage->TexObject);
+	if (rb && rb->TexImage)
+		texture_dirty(rb->TexImage->TexObject);
 }
 
 void

src/mesa/drivers/dri/nouveau/nouveau_texture.c

@@ -225,6 +225,7 @@ nouveau_choose_tex_format(struct gl_context *ctx, GLenum target,
 	case GL_INTENSITY12:
 	case GL_INTENSITY16:
 	case GL_INTENSITY8:
+	case GL_COMPRESSED_INTENSITY:
 		return MESA_FORMAT_I8;
 
 	case GL_RGB_S3TC:

src/mesa/drivers/dri/nouveau/nv04_context.c

@@ -138,7 +138,8 @@ nv04_context_destroy(struct gl_context *ctx)
 }
 
 static struct gl_context *
-nv04_context_create(struct nouveau_screen *screen, const struct gl_config *visual,
+nv04_context_create(struct nouveau_screen *screen, gl_api api,
+		    const struct gl_config *visual,
 		    struct gl_context *share_ctx)
 {
 	struct nv04_context *nctx;
@@ -153,7 +154,7 @@ nv04_context_create(struct nouveau_screen *screen, const struct gl_config *visua
 	ctx = &nctx->base.base;
 	hw = &nctx->base.hw;
 
-	if (!nouveau_context_init(ctx, screen, visual, share_ctx))
+	if (!nouveau_context_init(ctx, api, screen, visual, share_ctx))
 		goto fail;
 
 	/* GL constants. */

src/mesa/drivers/dri/nouveau/nv10_context.c

@@ -63,7 +63,7 @@ nv10_use_viewport_zclear(struct gl_context *ctx)
 	struct gl_framebuffer *fb = ctx->DrawBuffer;
 	struct gl_renderbuffer *depthRb = fb->Attachment[BUFFER_DEPTH].Renderbuffer;
 
-	return context_chipset(ctx) < 0x17 &&
+	return context_eng3d(ctx)->oclass < NV17_3D_CLASS &&
 		!nctx->hierz.clear_blocked && depthRb &&
 		(_mesa_get_format_bits(depthRb->Format,
 				       GL_DEPTH_BITS) >= 24);
@@ -184,7 +184,7 @@ nv10_clear(struct gl_context *ctx, GLbitfield buffers)
 	}
 
 	if ((buffers & BUFFER_BIT_DEPTH) && ctx->Depth.Mask) {
-		if (context_chipset(ctx) >= 0x17)
+		if (context_eng3d(ctx)->oclass >= NV17_3D_CLASS)
 			nv17_zclear(ctx, &buffers);
 		else
 			nv10_zclear(ctx, &buffers);
@@ -245,7 +245,7 @@ nv10_hwctx_init(struct gl_context *ctx)
 	BEGIN_NV04(push, NV04_GRAPH(3D, NOP), 1);
 	PUSH_DATA (push, 0);
 
-	if (context_chipset(ctx) >= 0x17) {
+	if (context_eng3d(ctx)->oclass >= NV17_3D_CLASS) {
 		BEGIN_NV04(push, NV17_3D(UNK01AC), 2);
 		PUSH_DATA (push, fifo->vram);
 		PUSH_DATA (push, fifo->vram);
@@ -257,7 +257,7 @@ nv10_hwctx_init(struct gl_context *ctx)
 		PUSH_DATA (push, 1);
 	}
 
-	if (context_chipset(ctx) >= 0x11) {
+	if (context_eng3d(ctx)->oclass >= NV15_3D_CLASS) {
 		BEGIN_NV04(push, SUBC_3D(0x120), 3);
 		PUSH_DATA (push, 0);
 		PUSH_DATA (push, 1);
@@ -427,7 +427,8 @@ nv10_context_destroy(struct gl_context *ctx)
 }
 
 static struct gl_context *
-nv10_context_create(struct nouveau_screen *screen, const struct gl_config *visual,
+nv10_context_create(struct nouveau_screen *screen, gl_api api,
+		    const struct gl_config *visual,
 		    struct gl_context *share_ctx)
 {
 	struct nouveau_context *nctx;
@@ -441,7 +442,7 @@ nv10_context_create(struct nouveau_screen *screen, const struct gl_config *visua
 
 	ctx = &nctx->base;
 
-	if (!nouveau_context_init(ctx, screen, visual, share_ctx))
+	if (!nouveau_context_init(ctx, api, screen, visual, share_ctx))
 		goto fail;
 
 	ctx->Extensions.ARB_texture_env_crossbar = true;

src/mesa/drivers/dri/nouveau/nv10_state_fb.c

@@ -106,7 +106,7 @@ nv10_emit_framebuffer(struct gl_context *ctx, int emit)
 
 	/* At least nv11 seems to get sad if we don't do this before
 	 * swapping RTs.*/
-	if (context_chipset(ctx) < 0x17) {
+	if (context_eng3d(ctx)->oclass < NV17_3D_CLASS) {
 		int i;
 
 		for (i = 0; i < 6; i++) {
@@ -140,7 +140,7 @@ nv10_emit_framebuffer(struct gl_context *ctx, int emit)
 		PUSH_MTHDl(push, NV10_3D(ZETA_OFFSET), BUFCTX_FB,
 				 s->bo, 0, bo_flags);
 
-		if (context_chipset(ctx) >= 0x17) {
+		if (context_eng3d(ctx)->oclass >= NV17_3D_CLASS) {
 			setup_hierz_buffer(ctx);
 			context_dirty(ctx, ZCLEAR);
 		}

src/mesa/drivers/dri/nouveau/nv10_state_raster.c

@@ -28,6 +28,7 @@
 #include "nouveau_context.h"
 #include "nouveau_gldefs.h"
 #include "nouveau_util.h"
+#include "nv_object.xml.h"
 #include "nv10_3d.xml.h"
 #include "nv10_driver.h"
 
@@ -120,7 +121,7 @@ nv10_emit_logic_opcode(struct gl_context *ctx, int emit)
 	struct nouveau_pushbuf *push = context_push(ctx);
 
 	assert(!ctx->Color.ColorLogicOpEnabled
-	       || context_chipset(ctx) >= 0x11);
+	       || context_eng3d(ctx)->oclass >= NV15_3D_CLASS);
 
 	BEGIN_NV04(push, NV11_3D(COLOR_LOGIC_OP_ENABLE), 2);
 	PUSH_DATAb(push, ctx->Color.ColorLogicOpEnabled);

src/mesa/drivers/dri/nouveau/nv20_context.c

@@ -438,7 +438,8 @@ nv20_context_destroy(struct gl_context *ctx)
 }
 
 static struct gl_context *
-nv20_context_create(struct nouveau_screen *screen, const struct gl_config *visual,
+nv20_context_create(struct nouveau_screen *screen, gl_api api,
+		    const struct gl_config *visual,
 		    struct gl_context *share_ctx)
 {
 	struct nouveau_context *nctx;
@@ -452,7 +453,7 @@ nv20_context_create(struct nouveau_screen *screen, const struct gl_config *visua
 
 	ctx = &nctx->base;
 
-	if (!nouveau_context_init(ctx, screen, visual, share_ctx))
+	if (!nouveau_context_init(ctx, api, screen, visual, share_ctx))
 		goto fail;
 
 	ctx->Extensions.ARB_texture_env_crossbar = true;

src/mesa/drivers/dri/r200/r200_context.c

@@ -271,7 +271,7 @@ GLboolean r200CreateContext( gl_api api,
    r200InitShaderFuncs(&functions);
    radeonInitQueryObjFunctions(&functions);
 
-   if (!radeonInitContext(&rmesa->radeon, &functions,
+   if (!radeonInitContext(&rmesa->radeon, api, &functions,
 			  glVisual, driContextPriv,
 			  sharedContextPrivate)) {
      free(rmesa);
@@ -279,12 +279,13 @@ GLboolean r200CreateContext( gl_api api,
      return GL_FALSE;
    }
 
-   driContextSetFlags(ctx, flags);
-
    rmesa->radeon.swtcl.RenderIndex = ~0;
    rmesa->radeon.hw.all_dirty = 1;
 
    ctx = &rmesa->radeon.glCtx;
+
+   driContextSetFlags(ctx, flags);
+
    /* Initialize the software rasterizer and helper modules.
     */
    _swrast_CreateContext( ctx );

src/mesa/drivers/dri/radeon/radeon_common_context.c

@@ -126,6 +126,7 @@ static void radeonInitDriverFuncs(struct dd_function_table *functions)
  * including the Mesa context itself.
  */
 GLboolean radeonInitContext(radeonContextPtr radeon,
+                            gl_api api,
 			    struct dd_function_table* functions,
 			    const struct gl_config * glVisual,
 			    __DRIcontext * driContextPriv,
@@ -147,7 +148,7 @@ GLboolean radeonInitContext(radeonContextPtr radeon,
 	else
 		shareCtx = NULL;
 
-	if (!_mesa_initialize_context(&radeon->glCtx, API_OPENGL_COMPAT,
+	if (!_mesa_initialize_context(&radeon->glCtx, api,
 				      glVisual, shareCtx,
 				      functions))
 		return GL_FALSE;

src/mesa/drivers/dri/radeon/radeon_common_context.h

@@ -515,6 +515,7 @@ static inline __DRIdrawable* radeon_get_readable(radeonContextPtr radeon)
 }
 
 GLboolean radeonInitContext(radeonContextPtr radeon,
+                            gl_api api,
 			    struct dd_function_table* functions,
 			    const struct gl_config * glVisual,
 			    __DRIcontext * driContextPriv,

src/mesa/drivers/dri/radeon/radeon_context.c

@@ -234,7 +234,7 @@ r100CreateContext( gl_api api,
    radeonInitTextureFuncs( &rmesa->radeon, &functions );
    radeonInitQueryObjFunctions(&functions);
 
-   if (!radeonInitContext(&rmesa->radeon, &functions,
+   if (!radeonInitContext(&rmesa->radeon, api, &functions,
 			  glVisual, driContextPriv,
 			  sharedContextPrivate)) {
      free(rmesa);
@@ -242,12 +242,13 @@ r100CreateContext( gl_api api,
      return GL_FALSE;
    }
 
-   driContextSetFlags(ctx, flags);
-
    rmesa->radeon.swtcl.RenderIndex = ~0;
    rmesa->radeon.hw.all_dirty = GL_TRUE;
 
    ctx = &rmesa->radeon.glCtx;
+
+   driContextSetFlags(ctx, flags);
+
    /* Initialize the software rasterizer and helper modules.
     */
    _swrast_CreateContext( ctx );

src/mesa/main/attrib.c

@@ -1377,6 +1377,7 @@ copy_array_object(struct gl_context *ctx,
 
    /* _Enabled must be the same than on push */
    dest->_Enabled = src->_Enabled;
+   dest->NewArrays = src->NewArrays;
    dest->_MaxElement = src->_MaxElement;
 }
 

src/mesa/main/buffers.c

@@ -360,16 +360,18 @@ _mesa_DrawBuffers(GLsizei n, const GLenum *buffers)
             return;
          }         
 
-         /* From the OpenGL 3.0 specification, page 259:
+         /* From the OpenGL 4.0 specification, page 256:
           * "For both the default framebuffer and framebuffer objects, the
           *  constants FRONT, BACK, LEFT, RIGHT, and FRONT_AND_BACK are not
           *  valid in the bufs array passed to DrawBuffers, and will result in
-          *  the error INVALID_OPERATION.  This restriction is because these
+          *  the error INVALID_ENUM. This restriction is because these
           *  constants may themselves refer to multiple buffers, as shown in
           *  table 4.4."
+          *  Previous versions of the OpenGL specification say INVALID_OPERATION,
+          *  but the Khronos conformance tests expect INVALID_ENUM.
           */
          if (_mesa_bitcount(destMask[output]) > 1) {
-            _mesa_error(ctx, GL_INVALID_OPERATION, "glDrawBuffersARB(buffer)");
+            _mesa_error(ctx, GL_INVALID_ENUM, "glDrawBuffersARB(buffer)");
             return;
          }
 

src/mesa/main/clear.c

@@ -274,7 +274,7 @@ make_color_buffer_mask(struct gl_context *ctx, GLint drawbuffer)
       break;
    default:
       {
-         GLuint buf = ctx->DrawBuffer->_ColorDrawBufferIndexes[drawbuffer];
+         GLint buf = ctx->DrawBuffer->_ColorDrawBufferIndexes[drawbuffer];
 
          if (buf >= 0 && att[buf].Renderbuffer) {
             mask |= 1 << buf;

src/mesa/main/enable.c

@@ -762,7 +762,6 @@ _mesa_set_enable(struct gl_context *ctx, GLenum cap, GLboolean state)
       case GL_COLOR_SUM_EXT:
          if (ctx->API != API_OPENGL_COMPAT)
             goto invalid_enum_error;
-         CHECK_EXTENSION(ARB_vertex_program, cap);
          if (ctx->Fog.ColorSumEnabled == state)
             return;
          FLUSH_VERTICES(ctx, _NEW_FOG);
@@ -1462,7 +1461,6 @@ _mesa_IsEnabled( GLenum cap )
       case GL_COLOR_SUM_EXT:
          if (ctx->API != API_OPENGL_COMPAT)
             goto invalid_enum_error;
-         CHECK_EXTENSION(ARB_vertex_program);
          return ctx->Fog.ColorSumEnabled;
 
       /* GL_ARB_multisample */

src/mesa/main/format_pack.c

@@ -1039,8 +1039,8 @@ pack_ubyte_SRGB8(const GLubyte src[4], void *dst)
 {
    GLubyte *d = ((GLubyte *) dst);
    d[2] = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
-   d[1] = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
-   d[0] = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
+   d[1] = linear_ubyte_to_srgb_ubyte(src[GCOMP]);
+   d[0] = linear_ubyte_to_srgb_ubyte(src[BCOMP]);
 }
 
 static void
@@ -1085,8 +1085,8 @@ pack_ubyte_SARGB8(const GLubyte src[4], void *dst)
 {
    GLuint *d = ((GLuint *) dst);
    GLubyte r = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
-   GLubyte g = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
-   GLubyte b = linear_ubyte_to_srgb_ubyte(src[RCOMP]);
+   GLubyte g = linear_ubyte_to_srgb_ubyte(src[GCOMP]);
+   GLubyte b = linear_ubyte_to_srgb_ubyte(src[BCOMP]);
    *d = PACK_COLOR_8888(src[ACOMP], r, g, b);
 }