docs: Update 10.4.0 release notes

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

VERSION

@@ -1 +1 @@
-10.4.0-rc1
+10.4.0

bin/.cherry-ignore

@@ -0,0 +1,2 @@
+# No whitespace commits in stable.
+a10bf5c10caf27232d4df8da74d5c35c23eb883d

configure.ac

@@ -252,11 +252,29 @@ 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])
+SSE41_CFLAGS="-msse4.1"
+dnl Code compiled by GCC with -msse* assumes a 16 byte aligned
+dnl stack, but on x86-32 such alignment is not guaranteed.
+case "$target_cpu" in
+i?86)
+    SSE41_CFLAGS="$SSE41_CFLAGS -mstackrealign"
+    ;;
+esac
+save_CFLAGS="$CFLAGS"
+CFLAGS="$SSE41_CFLAGS $CFLAGS"
+AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
+#include <smmintrin.h>
+int main () {
+    __m128i a = _mm_set1_epi32 (0), b = _mm_set1_epi32 (0), c;
+    c = _mm_max_epu32(a, b);
+    return 0;
+}]])], SSE41_SUPPORTED=1)
+CFLAGS="$save_CFLAGS"
 if test "x$SSE41_SUPPORTED" = x1; then
     DEFINES="$DEFINES -DUSE_SSE41"
 fi
 AM_CONDITIONAL([SSE41_SUPPORTED], [test x$SSE41_SUPPORTED = x1])
+AC_SUBST([SSE41_CFLAGS], $SSE41_CFLAGS)
 
 dnl Can't have static and shared libraries, default to static if user
 dnl explicitly requested. If both disabled, set to static since shared
@@ -2076,6 +2094,8 @@ AM_CONDITIONAL(HAVE_SPARC_ASM, test "x$asm_arch" = xsparc)
 
 AC_SUBST([NINE_MAJOR], 1)
 AC_SUBST([NINE_MINOR], 0)
+AC_SUBST([NINE_TINY], 0)
+AC_SUBST([NINE_VERSION], "$NINE_MAJOR.$NINE_MINOR.$NINE_TINY")
 
 AC_SUBST([VDPAU_MAJOR], 1)
 AC_SUBST([VDPAU_MINOR], 0)

docs/GL3.txt

@@ -187,7 +187,7 @@ GL 4.4, GLSL 4.40:
 GL 4.5, GLSL 4.50:
 
   GL_ARB_ES3_1_compatibility                           not started
-  GL_ARB_clip_control                                  DONE (llvmpipe, softpipe, r300, r600, radeonsi)
+  GL_ARB_clip_control                                  DONE (nv50, nvc0, r300, r600, radeonsi, llvmpipe, softpipe)
   GL_ARB_conditional_render_inverted                   DONE (i965, nv50, nvc0, llvmpipe, softpipe)
   GL_ARB_cull_distance                                 not started
   GL_ARB_derivative_control                            DONE (i965, nv50, nvc0, r600)

docs/relnotes/10.2.html

@@ -88,6 +88,8 @@ following options during configure, if you would like support for svga driver
 Note: The files are installed in $(libdir)/gallium-pipe/ and the interface
 between them and libxatracker.so is <strong>not</strong> stable.
 </p>
+
+<li>The environment variable GALLIUM_MSAA that forced a multisample GLX visual was removed.</li>
 </ul>
 
 </div>

docs/relnotes/10.3.html

@@ -327,6 +327,7 @@ DRM drivers that don't have a full-fledged GEM (such as qxl or simpledrm)</li>
 <li>Removed support for the GL_ATI_envmap_bumpmap extension</li>
 <li>The hacky --enable-32/64-bit is no longer available in configure. To build
 32/64 bit mesa refer to the default method recommended by your distribution</li>
+</li>The environment variable GALLIUM_MSAA that forced a multisample GLX visual was removed.</li>
 </ul>
 
 </div>

docs/relnotes/10.4.html

@@ -14,7 +14,7 @@
 <iframe src="../contents.html"></iframe>
 <div class="content">
 
-<h1>Mesa 10.4 Release Notes / TBD</h1>
+<h1>Mesa 10.4 Release Notes / December 14, 2014</h1>
 
 <p>
 Mesa 10.4 is a new development release.
@@ -47,18 +47,209 @@ Note: some of the new features are only available with certain drivers.
 <li>GL_ARB_conditional_render_inverted on nv50</li>
 <li>GL_ARB_sample_shading on r600</li>
 <li>GL_ARB_texture_view on nv50, nvc0</li>
-<li>GL_ARB_clip_control on llvmpipe, softpipe, r300, r600, radeonsi</li>
+<li>GL_ARB_clip_control on nv50, nvc0, r300, r600, radeonsi, llvmpipe, softpipe</li>
 <li>GL_KHR_context_flush_control on all drivers</li>
 </ul>
 
 
 <h2>Bug fixes</h2>
 
-TBD.
+<p>This list is likely incomplete.</p>
+
+<ul>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=79963">Bug 79963</a> - [ILK Bisected]some piglit and ogles2conform cases fail </li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=29661">Bug 29661</a> - MSVC built u_format_test fails on Windows</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=38873">Bug 38873</a> - [855gm] gnome-shell misrendered</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=54372">Bug 54372</a> - GLX_INTEL_swap_event crashes driver when swapping window buffers</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=60879">Bug 60879</a> - [radeonsi] X11 can't start with acceleration enabled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=61415">Bug 61415</a> - Clover ignores --with-opencl-libdir path</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=66184">Bug 66184</a> - src/mesa/state_tracker/st_glsl_to_tgsi.cpp:3216:simplify_cmp: Assertion `inst-&gt;dst.index &lt; 4096' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=67672">Bug 67672</a> - [llvmpipe] lp_test_arit fails on old CPUs</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=69200">Bug 69200</a> - [Bisected]Piglit glx/glx-multithread-shader-compile aborted</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=70410">Bug 70410</a> - egl-static/Makefile: linking fails with llvm &gt;= 3.4</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=72819">Bug 72819</a> - [855GM] Incorrect drop shadow color on windows and strange white rectangle when showing/hiding GLX-dock...</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=74563">Bug 74563</a> - Surfaceless contexts are not properly released by DRI drivers</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=75011">Bug 75011</a> - [hyperz] Performance drop since git-01e6371 (disable hyperz by default) with radeonsi</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=75112">Bug 75112</a> - Meta Bug for HyperZ issues on r600g and radeonsi</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=76252">Bug 76252</a> - Dynamic loading/unloading of opengl32.dll results in a deadlock</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=76861">Bug 76861</a> - mid3 generates slow code for constant arguments</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=77957">Bug 77957</a> - Variably-indexed constant arrays result in terrible shader code</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=78468">Bug 78468</a> - Compiling of shader gets stuck in infinite loop</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=78770">Bug 78770</a> - [SNB bisected]Webglc conformance/textures/texture-size-limit.html fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=79155">Bug 79155</a> - [Tesseract Game] Global Illumination: Medium Causes Color Distortion</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=79462">Bug 79462</a> - [NVC0/Codegen] Shader compilation falis in spill logic</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80011">Bug 80011</a> - [softpipe] tgsi/tgsi_exec.c:2023:exec_txf: Assertion `0' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80012">Bug 80012</a> - [softpipe] draw/draw_gs.c:113:tgsi_fetch_gs_outputs: Assertion `!util_is_inf_or_nan(output[slot][0])' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80050">Bug 80050</a> - [855GM] Incorrect drop shadow color under windows in Cinnamon persists with MESA 10.1.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80247">Bug 80247</a> - Khronos conformance test  ES3-CTS.gtf.GL3Tests.transform_feedback.transform_feedback_vertex_id  fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80561">Bug 80561</a> - Incorrect implementation of some VDPAU APIs.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80615">Bug 80615</a> - Files in bellagio directory [omx tracker] don't respect installation folder</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=80848">Bug 80848</a> - [dri3] Building mesa fails with dri3 enabled</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=81680">Bug 81680</a> - [r600g] Firefox crashes with hardware acceleration turned on</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82255">Bug 82255</a> - [VP2] Chroma planes are vertically stretched during VDPAU playback</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82472">Bug 82472</a> - piglit 16385-consecutive-chars regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82537">Bug 82537</a> - Stunt Rally GLSL compiler assertion failure</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82538">Bug 82538</a> - Super Maryo Chronicles fails with st/mesa assertion failure</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82539">Bug 82539</a> - vmw_screen_dri.lo In file included from vmw_screen_dri.c:41: vmwgfx_drm.h:32:17: error: drm.h: No such file or directory</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82796">Bug 82796</a> - [IVB/BYT-M/HSW/BDW Bisected]Synmark2_v6.0_OglTerrainFlyInst/OglTerrainPanInst cannot run as image validation failed</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82804">Bug 82804</a> - unreal engine 4 rendering errors</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82828">Bug 82828</a> - Regression: Crash in 3Dmark2001</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82846">Bug 82846</a> - [BDW Bisected] Gpu hang when running Lightsmark v2008/Warsow v1.0/Xonotic v0.7/unigine-demos</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82881">Bug 82881</a> - test_vec4_register_coalesce regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82882">Bug 82882</a> - [swrast] piglit glsl-fs-uniform-bool-1 regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82921">Bug 82921</a> - layout(location=0) emits error &gt;= MAX_UNIFORM_LOCATIONS due to integer underflow</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82929">Bug 82929</a> - [BDW Bisected]glxgears causes X hang</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=82932">Bug 82932</a> - [SNB+ Bisected]Ogles3conform ES3-CTS.shaders.indexing.vector_subscript.vec3_static_loop_subscript_write_direct_read_vertex fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83079">Bug 83079</a> - [NVC0] Dota 2 (Linux native and Wine) crash with Nouveau Drivers</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83080">Bug 83080</a> - [SNB+ Bisected]ES3-CTS.shaders.loops.do_while_constant_iterations.mixed_break_continue_fragment fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83081">Bug 83081</a> - [BDW Bisected]Piglit spec_ARB_sample_shading_builtin-gl-sample-mask_2 is core dumped</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83127">Bug 83127</a> - [ILK Bisected]Piglit glean_texCombine fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83148">Bug 83148</a> - Unity invisible under Ubuntu 14.04 and 14.10</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83355">Bug 83355</a> - FTBFS: src/mesa/program/program_lexer.l:122:64: error: unknown type name 'YYSTYPE'</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83380">Bug 83380</a> - Linking fails when not writing gl_Position.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83418">Bug 83418</a> - EU IV is incorrectly rendered after git1409011930.d571f2</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83432">Bug 83432</a> - r600_query.c:269:r600_emit_query_end: Assertion `ctx-&gt;num_pipelinestat_queries &gt; 0' failed [Gallium HUD]</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83463">Bug 83463</a> - [swrast] piglit glsl-vs-clamp-1 regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83468">Bug 83468</a> - [UBO] Using bool from UBO as if-statement condition asserts</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83500">Bug 83500</a> - si_dma_copy_tile causes GPU hangs</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83506">Bug 83506</a> - [UBO] row_major layout ignored inside structures</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83533">Bug 83533</a> - [UBO] nested structures don't get appropriate padding</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83573">Bug 83573</a> - [swrast] piglit fs-op-not-bool-using-if regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83574">Bug 83574</a> - [llvmpipe] [softpipe] piglit arb_explicit_uniform_location-use-of-unused-loc regression</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83741">Bug 83741</a> - [UBO] row_major layout partially ignored for arrays of structures</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83777">Bug 83777</a> - [regression] ilo fails to build</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=83934">Bug 83934</a> - Structures must have same name to be considered same type.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84140">Bug 84140</a> - mplayer crashes playing some files using vdpau output</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84145">Bug 84145</a> - UE4: Realistic Rendering Demo render blue</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84178">Bug 84178</a> - Big glamor regression in Xorg server 1.6.99.1 GIT: x11perf 1.5 Test: PutImage XY 500x500 Square</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84355">Bug 84355</a> - texture2DProjLod and textureCubeLod are not supported when using GLES.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84529">Bug 84529</a> - [IVB bisected] glean fragProg1 CMP test failed</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84538">Bug 84538</a> - lp_test_format.c:226:4: error: too few arguments to function ‘gallivm_create’</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84539">Bug 84539</a> - brw_fs_register_coalesce.cpp:183: bool fs_visitor::register_coalesce(): Assertion `src_size &lt;= 11' failed.</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84557">Bug 84557</a> - [HSW] &quot;Emit ELSE/ENDIF JIP with type D on Gen 7&quot; causes Atomic Afterlife and GPU hangs</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84651">Bug 84651</a> - Distorted graphics or black window when running Battle.net app on Intel hardware via wine</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84662">Bug 84662</a> - Long pauses with Unreal demo Elemental on R9270X since : Always flush the HDP cache before submitting a CS to the GPU</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84777">Bug 84777</a> - [BSW]Piglit spec_glsl-1.50_execution_geometry-basic fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=84807">Bug 84807</a> - Build issue starting between bf4aecfb2acc8d0dc815105d2f36eccbc97c284b and a3e9582f09249ad27716ba82c7dfcee685b65d51</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85189">Bug 85189</a> - llvm/invocation.cpp: In function 'void {anonymous}::optimize(llvm::Module*, unsigned int, const std::vector&lt;llvm::Function*&gt;&amp;)': llvm/invocation.cpp:324:18: error: expected type-specifier</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85267">Bug 85267</a> - vlc crashes with vdpau (Radeon 3850HD) [r600]</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85377">Bug 85377</a> - lp_test_format failure with llvm-3.6</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85425">Bug 85425</a> - [bisected] Compiler error in clip control operations in meta</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85429">Bug 85429</a> - indirect.c:296: multiple definition of `__indirect_glNewList'</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85454">Bug 85454</a> - Unigine Sanctuary with Wine crashes on Mesa Git</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85647">Bug 85647</a> - Random radeonsi crashes with mesa 10.3.x</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85683">Bug 85683</a> - [i965 Bisected]Piglit shaders_glsl-vs-raytrace-bug26691 segfault</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=85691">Bug 85691</a> - 'glsl: Drop constant 0.0 components from dot products.' broke piglit shaders/glsl-gnome-shell-dim-window and a few others with Gallium</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86025">Bug 86025</a> - src\glsl\list.h(535) : error C2143: syntax error : missing ';' before 'type'</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86089">Bug 86089</a> - [r600g][mesa 10.4.0-dev] shader failure - r600_sb::bc_finalizer::cf_peephole() when starting Second Life</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86145">Bug 86145</a> - Pipeline statistic counter values for VF always 0</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86618">Bug 86618</a> - [NV96] neg modifiers not working in MIN and MAX operations</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86760">Bug 86760</a> - mesa doesn't build: recipe for target 'r600_llvm.lo' failed</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86764">Bug 86764</a> - [SNB+ Bisected]Piglit glean/pointSprite fails</li>
+
+<li><a href="https://bugs.freedesktop.org/show_bug.cgi?id=86788">Bug 86788</a> - (bisected) 32bit UrbanTerror 4.1 timedemo sse4.1 segfault...</li>
+
+</ul>
 
 <h2>Changes</h2>
 
 <ul>
+<li>The environment variable GALLIUM_MSAA that forced a multisample GLX visual was removed.</li>
 </ul>
 
 </div>

scons/gallium.py

@@ -621,7 +621,7 @@ def generate(env):
     env.Tool('custom')
     createInstallMethods(env)
 
-    env.PkgCheckModules('X11', ['x11', 'xext', 'xdamage', 'xfixes'])
+    env.PkgCheckModules('X11', ['x11', 'xext', 'xdamage', 'xfixes', 'glproto >= 1.4.13'])
     env.PkgCheckModules('XCB', ['x11-xcb', 'xcb-glx >= 1.8.1', 'xcb-dri2 >= 1.8'])
     env.PkgCheckModules('XF86VIDMODE', ['xxf86vm'])
     env.PkgCheckModules('DRM', ['libdrm >= 2.4.38'])

src/gallium/auxiliary/indices/u_primconvert.c

@@ -124,6 +124,9 @@ util_primconvert_draw_vbo(struct primconvert_context *pc,
    new_info.indexed = true;
    new_info.min_index = info->min_index;
    new_info.max_index = info->max_index;
+   new_info.index_bias = info->index_bias;
+   new_info.start_instance = info->start_instance;
+   new_info.instance_count = info->instance_count;
 
    if (info->indexed) {
       u_index_translator(pc->primtypes_mask,
@@ -136,6 +139,7 @@ util_primconvert_draw_vbo(struct primconvert_context *pc,
          src = pipe_buffer_map(pc->pipe, ib->buffer,
                                PIPE_TRANSFER_READ, &src_transfer);
       }
+      src = (const uint8_t *)src + ib->offset;
    }
    else {
       u_index_generator(pc->primtypes_mask,

src/gallium/auxiliary/util/u_simple_shaders.c

@@ -124,6 +124,76 @@ void *util_make_layered_clear_vertex_shader(struct pipe_context *pipe)
    return pipe->create_vs_state(pipe, &state);
 }
 
+/**
+ * Takes position and color, and outputs position, color, and instance id.
+ */
+void *util_make_layered_clear_helper_vertex_shader(struct pipe_context *pipe)
+{
+   static const char text[] =
+         "VERT\n"
+         "DCL IN[0]\n"
+         "DCL IN[1]\n"
+         "DCL SV[0], INSTANCEID\n"
+         "DCL OUT[0], POSITION\n"
+         "DCL OUT[1], GENERIC[0]\n"
+         "DCL OUT[2], GENERIC[1]\n"
+
+         "MOV OUT[0], IN[0]\n"
+         "MOV OUT[1], IN[1]\n"
+         "MOV OUT[2].x, SV[0].xxxx\n"
+         "END\n";
+   struct tgsi_token tokens[1000];
+   struct pipe_shader_state state = {tokens};
+
+   if (!tgsi_text_translate(text, tokens, Elements(tokens))) {
+      assert(0);
+      return NULL;
+   }
+   return pipe->create_vs_state(pipe, &state);
+}
+
+/**
+ * Takes position, color, and target layer, and emits vertices on that target
+ * layer, with the specified color.
+ */
+void *util_make_layered_clear_geometry_shader(struct pipe_context *pipe)
+{
+   static const char text[] =
+      "GEOM\n"
+      "PROPERTY GS_INPUT_PRIMITIVE TRIANGLES\n"
+      "PROPERTY GS_OUTPUT_PRIMITIVE TRIANGLE_STRIP\n"
+      "PROPERTY GS_MAX_OUTPUT_VERTICES 3\n"
+      "PROPERTY GS_INVOCATIONS 1\n"
+      "DCL IN[][0], POSITION\n" /* position */
+      "DCL IN[][1], GENERIC[0]\n" /* color */
+      "DCL IN[][2], GENERIC[1]\n" /* vs invocation */
+      "DCL OUT[0], POSITION\n"
+      "DCL OUT[1], GENERIC[0]\n"
+      "DCL OUT[2], LAYER\n"
+      "IMM[0] INT32 {0, 0, 0, 0}\n"
+
+      "MOV OUT[0], IN[0][0]\n"
+      "MOV OUT[1], IN[0][1]\n"
+      "MOV OUT[2].x, IN[0][2].xxxx\n"
+      "EMIT IMM[0].xxxx\n"
+      "MOV OUT[0], IN[1][0]\n"
+      "MOV OUT[1], IN[1][1]\n"
+      "MOV OUT[2].x, IN[1][2].xxxx\n"
+      "EMIT IMM[0].xxxx\n"
+      "MOV OUT[0], IN[2][0]\n"
+      "MOV OUT[1], IN[2][1]\n"
+      "MOV OUT[2].x, IN[2][2].xxxx\n"
+      "EMIT IMM[0].xxxx\n"
+      "END\n";
+   struct tgsi_token tokens[1000];
+   struct pipe_shader_state state = {tokens};
+
+   if (!tgsi_text_translate(text, tokens, Elements(tokens))) {
+      assert(0);
+      return NULL;
+   }
+   return pipe->create_gs_state(pipe, &state);
+}
 
 /**
  * Make simple fragment texture shader:

src/gallium/auxiliary/util/u_simple_shaders.h

@@ -59,6 +59,12 @@ util_make_vertex_passthrough_shader_with_so(struct pipe_context *pipe,
 extern void *
 util_make_layered_clear_vertex_shader(struct pipe_context *pipe);
 
+extern void *
+util_make_layered_clear_helper_vertex_shader(struct pipe_context *pipe);
+
+extern void *
+util_make_layered_clear_geometry_shader(struct pipe_context *pipe);
+
 extern void *
 util_make_fragment_tex_shader_writemask(struct pipe_context *pipe, 
                                         unsigned tex_target,

src/gallium/auxiliary/util/u_snprintf.c

@@ -176,7 +176,7 @@
 #define HAVE_ASPRINTF 1 /* not needed */
 #define HAVE_STDARG_H 1
 #define HAVE_STDDEF_H 1
-#define HAVE_STDINT_H 0
+#define HAVE_STDINT_H 1
 #define HAVE_STDLIB_H 1
 #define HAVE_INTTYPES_H 0
 #define HAVE_LOCALE_H 0

src/gallium/drivers/freedreno/a3xx/fd3_blend.c

@@ -105,8 +105,7 @@ fd3_blend_state_create(struct pipe_context *pctx,
 				A3XX_RB_MRT_BLEND_CONTROL_RGB_DEST_FACTOR(fd_blend_factor(rt->rgb_dst_factor)) |
 				A3XX_RB_MRT_BLEND_CONTROL_ALPHA_SRC_FACTOR(fd_blend_factor(rt->alpha_src_factor)) |
 				A3XX_RB_MRT_BLEND_CONTROL_ALPHA_BLEND_OPCODE(blend_func(rt->alpha_func)) |
-				A3XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR(fd_blend_factor(rt->alpha_dst_factor)) |
-				A3XX_RB_MRT_BLEND_CONTROL_CLAMP_ENABLE;
+				A3XX_RB_MRT_BLEND_CONTROL_ALPHA_DEST_FACTOR(fd_blend_factor(rt->alpha_dst_factor));
 
 		so->rb_mrt[i].control =
 				A3XX_RB_MRT_CONTROL_ROP_CODE(rop) |

src/gallium/drivers/freedreno/a3xx/fd3_emit.c

@@ -566,16 +566,20 @@ fd3_emit_state(struct fd_context *ctx, struct fd_ringbuffer *ring,
 		}
 	}
 
-	if ((dirty & FD_DIRTY_BLEND) && ctx->blend) {
+	if ((dirty & (FD_DIRTY_BLEND | FD_DIRTY_FRAMEBUFFER)) && ctx->blend) {
 		struct fd3_blend_stateobj *blend = fd3_blend_stateobj(ctx->blend);
 		uint32_t i;
 
 		for (i = 0; i < ARRAY_SIZE(blend->rb_mrt); i++) {
+			bool is_float = util_format_is_float(
+					pipe_surface_format(ctx->framebuffer.cbufs[i]));
+
 			OUT_PKT0(ring, REG_A3XX_RB_MRT_CONTROL(i), 1);
 			OUT_RING(ring, blend->rb_mrt[i].control);
 
 			OUT_PKT0(ring, REG_A3XX_RB_MRT_BLEND_CONTROL(i), 1);
-			OUT_RING(ring, blend->rb_mrt[i].blend_control);
+			OUT_RING(ring, blend->rb_mrt[i].blend_control |
+					COND(!is_float, A3XX_RB_MRT_BLEND_CONTROL_CLAMP_ENABLE));
 		}
 	}
 

src/gallium/drivers/freedreno/ir3/ir3_compiler.c

@@ -2112,7 +2112,7 @@ trans_cov(const struct instr_translater *t,
  *  madsh.m16 tmp1, a, b, tmp0 (mul-add shift high mix, i.e. ah * bl << 16)
  *  madsh.m16 dst, b, a, tmp1 (i.e. al * bh << 16)
  *
- * For UMAD, replace first mull.u with mad.u16.
+ * For UMAD, add in the extra argument after mull.u.
  */
 static void
 trans_umul(const struct instr_translater *t,
@@ -2135,16 +2135,16 @@ trans_umul(const struct instr_translater *t,
 	if (is_rel_or_const(b))
 		b = get_unconst(ctx, b);
 
-	if (t->tgsi_opc == TGSI_OPCODE_UMUL) {
-		/* mull.u tmp0, a, b */
-		instr = instr_create(ctx, 2, OPC_MULL_U);
-		vectorize(ctx, instr, &tmp0_dst, 2, a, 0, b, 0);
-	} else {
+	/* mull.u tmp0, a, b */
+	instr = instr_create(ctx, 2, OPC_MULL_U);
+	vectorize(ctx, instr, &tmp0_dst, 2, a, 0, b, 0);
+
+	if (t->tgsi_opc == TGSI_OPCODE_UMAD) {
 		struct tgsi_src_register *c = &inst->Src[2].Register;
 
-		/* mad.u16 tmp0, a, b, c */
-		instr = instr_create(ctx, 3, OPC_MAD_U16);
-		vectorize(ctx, instr, &tmp0_dst, 3, a, 0, b, 0, c, 0);
+		/* add.u tmp0, tmp0, c */
+		instr = instr_create(ctx, 2, OPC_ADD_U);
+		vectorize(ctx, instr, &tmp0_dst, 2, tmp0_src, 0, c, 0);
 	}
 
 	/* madsh.m16 tmp1, a, b, tmp0 */
@@ -2350,6 +2350,9 @@ trans_idiv(const struct instr_translater *t,
 	if (t->tgsi_opc == TGSI_OPCODE_MOD || t->tgsi_opc == TGSI_OPCODE_UMOD) {
 		/* The division result will have ended up in q. */
 
+		if (is_rel_or_const(b))
+			b = get_unconst(ctx, b);
+
 		/* mull.u r, q, b */
 		instr = instr_create(ctx, 2, OPC_MULL_U);
 		vectorize(ctx, instr, &r_dst, 2, q_src, 0, b, 0);

src/gallium/drivers/llvmpipe/lp_test_arit.c

@@ -33,6 +33,7 @@
 #include "util/u_pointer.h"
 #include "util/u_memory.h"
 #include "util/u_math.h"
+#include "util/u_cpu_detect.h"
 
 #include "gallivm/lp_bld.h"
 #include "gallivm/lp_bld_debug.h"
@@ -332,6 +333,38 @@ build_unary_test_func(struct gallivm_state *gallivm,
 }
 
 
+/*
+ * Flush denorms to zero.
+ */
+static float
+flush_denorm_to_zero(float val)
+{
+   /*
+    * If we have a denorm manually set it to (+-)0.
+    * This is because the reference may or may not do the right thing
+    * otherwise because we want the result according to treating all
+    * denormals as zero (FTZ/DAZ). Not using fpclassify because
+    * a) some compilers are stuck at c89 (msvc)
+    * b) not sure it reliably works with non-standard ftz/daz mode
+    * And, right now we only disable denorms with jited code on x86/sse
+    * (albeit this should be classified as a bug) so to get results which
+    * match we must only flush them to zero here in that case too.
+    */
+   union fi fi_val;
+
+   fi_val.f = val;
+
+#if defined(PIPE_ARCH_SSE)
+   if (util_cpu_caps.has_sse) {
+      if ((fi_val.ui & 0x7f800000) == 0) {
+         fi_val.ui &= 0xff800000;
+      }
+   }
+#endif
+
+   return fi_val.f;
+}
+
 /*
  * Test one LLVM unary arithmetic builder function.
  */
@@ -374,10 +407,13 @@ test_unary(unsigned verbose, FILE *fp, const struct unary_test_t *test)
 
       test_func_jit(out, in);
       for (i = 0; i < num_vals; ++i) {
-         float ref = test->ref(in[i]);
+         float testval, ref;
          double error, precision;
          bool pass;
 
+         testval = flush_denorm_to_zero(in[i]);
+         ref = flush_denorm_to_zero(test->ref(testval));
+
          if (util_inf_sign(ref) && util_inf_sign(out[i]) == util_inf_sign(ref)) {
             error = 0;
          } else {

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

@@ -924,7 +924,9 @@ CodeEmitterNV50::emitMINMAX(const Instruction *i)
          break;
       }
       code[1] |= i->src(0).mod.abs() << 20;
+      code[1] |= i->src(0).mod.neg() << 26;
       code[1] |= i->src(1).mod.abs() << 19;
+      code[1] |= i->src(1).mod.neg() << 27;
    }
    emitForm_MAD(i);
 }

src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp

@@ -2343,9 +2343,12 @@ Converter::handleInstruction(const struct tgsi_full_instruction *insn)
          mkMov(dst0[c], val0);
       break;
    case TGSI_OPCODE_ARL:
+   case TGSI_OPCODE_ARR:
       FOR_EACH_DST_ENABLED_CHANNEL(0, c, tgsi) {
+         const RoundMode rnd =
+            tgsi.getOpcode() == TGSI_OPCODE_ARR ? ROUND_N : ROUND_M;
          src0 = fetchSrc(0, c);
-         mkCvt(OP_CVT, TYPE_S32, dst0[c], TYPE_F32, src0)->rnd = ROUND_M;
+         mkCvt(OP_CVT, TYPE_S32, dst0[c], TYPE_F32, src0)->rnd = rnd;
       }
       break;
    case TGSI_OPCODE_UARL:

src/gallium/drivers/nouveau/nv50/nv50_3d.xml.h

@@ -415,7 +415,7 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 #define NV50_3D_VERTEX_BUFFER_COUNT				0x00000d78
 
-#define NV50_3D_UNK0D7C						0x00000d7c
+#define NV50_3D_DEPTH_CLIP_NEGATIVE_Z				0x00000d7c
 
 #define NV50_3D_CLEAR_COLOR(i0)				       (0x00000d80 + 0x4*(i0))
 #define NV50_3D_CLEAR_COLOR__ESIZE				0x00000004

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

@@ -180,7 +180,12 @@ nv50_invalidate_resource_storage(struct nouveau_context *ctx,
       }
    }
 
-   if (res->bind & PIPE_BIND_VERTEX_BUFFER) {
+   if (res->bind & (PIPE_BIND_VERTEX_BUFFER |
+                    PIPE_BIND_INDEX_BUFFER |
+                    PIPE_BIND_CONSTANT_BUFFER |
+                    PIPE_BIND_STREAM_OUTPUT |
+                    PIPE_BIND_SAMPLER_VIEW)) {
+
       assert(nv50->num_vtxbufs <= PIPE_MAX_ATTRIBS);
       for (i = 0; i < nv50->num_vtxbufs; ++i) {
          if (nv50->vtxbuf[i].buffer == res) {
@@ -190,14 +195,11 @@ nv50_invalidate_resource_storage(struct nouveau_context *ctx,
                return ref;
          }
       }
-   }
-   if (res->bind & PIPE_BIND_INDEX_BUFFER) {
+
       if (nv50->idxbuf.buffer == res)
          if (!--ref)
             return ref;
-   }
 
-   if (res->bind & PIPE_BIND_SAMPLER_VIEW) {
       for (s = 0; s < 3; ++s) {
       assert(nv50->num_textures[s] <= PIPE_MAX_SAMPLERS);
       for (i = 0; i < nv50->num_textures[s]; ++i) {
@@ -210,12 +212,11 @@ nv50_invalidate_resource_storage(struct nouveau_context *ctx,
          }
       }
       }
-   }
 
-   if (res->bind & PIPE_BIND_CONSTANT_BUFFER) {
       for (s = 0; s < 3; ++s) {
-      assert(nv50->num_vtxbufs <= NV50_MAX_PIPE_CONSTBUFS);
-      for (i = 0; i < nv50->num_vtxbufs; ++i) {
+      for (i = 0; i < NV50_MAX_PIPE_CONSTBUFS; ++i) {
+         if (!(nv50->constbuf_valid[s] & (1 << i)))
+            continue;
          if (!nv50->constbuf[s][i].user &&
              nv50->constbuf[s][i].u.buf == res) {
             nv50->dirty |= NV50_NEW_CONSTBUF;

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

@@ -173,6 +173,7 @@ nv50_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_TGSI_FS_FINE_DERIVATIVE:
    case PIPE_CAP_SAMPLER_VIEW_TARGET:
    case PIPE_CAP_CONDITIONAL_RENDER_INVERTED:
+   case PIPE_CAP_CLIP_HALFZ:
       return 1;
    case PIPE_CAP_SEAMLESS_CUBE_MAP:
       return 1; /* class_3d >= NVA0_3D_CLASS; */
@@ -204,7 +205,6 @@ nv50_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION:
    case PIPE_CAP_COMPUTE:
    case PIPE_CAP_DRAW_INDIRECT:
-   case PIPE_CAP_CLIP_HALFZ:
       return 0;
 
    case PIPE_CAP_VENDOR_ID:

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

@@ -333,6 +333,9 @@ nv50_rasterizer_state_create(struct pipe_context *pipe,
    SB_BEGIN_3D(so, VIEW_VOLUME_CLIP_CTRL, 1);
    SB_DATA    (so, reg);
 
+   SB_BEGIN_3D(so, DEPTH_CLIP_NEGATIVE_Z, 1);
+   SB_DATA    (so, cso->clip_halfz);
+
    assert(so->size <= (sizeof(so->state) / sizeof(so->state[0])));
    return (void *)so;
 }

src/gallium/drivers/nouveau/nvc0/nvc0_3d.xml.h

@@ -318,6 +318,8 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 #define NVC0_3D_VERTEX_BUFFER_COUNT				0x00000d78
 
+#define NVC0_3D_DEPTH_CLIP_NEGATIVE_Z				0x00000d7c
+
 #define NVC0_3D_CLEAR_COLOR(i0)				       (0x00000d80 + 0x4*(i0))
 #define NVC0_3D_CLEAR_COLOR__ESIZE				0x00000004
 #define NVC0_3D_CLEAR_COLOR__LEN				0x00000004

src/gallium/drivers/nouveau/nvc0/nvc0_context.c

@@ -196,7 +196,12 @@ nvc0_invalidate_resource_storage(struct nouveau_context *ctx,
       }
    }
 
-   if (res->bind & PIPE_BIND_VERTEX_BUFFER) {
+   if (res->bind & (PIPE_BIND_VERTEX_BUFFER |
+                    PIPE_BIND_INDEX_BUFFER |
+                    PIPE_BIND_CONSTANT_BUFFER |
+                    PIPE_BIND_STREAM_OUTPUT |
+                    PIPE_BIND_COMMAND_ARGS_BUFFER |
+                    PIPE_BIND_SAMPLER_VIEW)) {
       for (i = 0; i < nvc0->num_vtxbufs; ++i) {
          if (nvc0->vtxbuf[i].buffer == res) {
             nvc0->dirty |= NVC0_NEW_ARRAYS;
@@ -205,17 +210,14 @@ nvc0_invalidate_resource_storage(struct nouveau_context *ctx,
                return ref;
          }
       }
-   }
-   if (res->bind & PIPE_BIND_INDEX_BUFFER) {
+
       if (nvc0->idxbuf.buffer == res) {
          nvc0->dirty |= NVC0_NEW_IDXBUF;
          nouveau_bufctx_reset(nvc0->bufctx_3d, NVC0_BIND_IDX);
          if (!--ref)
             return ref;
       }
-   }
 
-   if (res->bind & PIPE_BIND_SAMPLER_VIEW) {
       for (s = 0; s < 5; ++s) {
       for (i = 0; i < nvc0->num_textures[s]; ++i) {
          if (nvc0->textures[s][i] &&
@@ -228,11 +230,11 @@ nvc0_invalidate_resource_storage(struct nouveau_context *ctx,
          }
       }
       }
-   }
 
-   if (res->bind & PIPE_BIND_CONSTANT_BUFFER) {
       for (s = 0; s < 5; ++s) {
-      for (i = 0; i < nvc0->num_vtxbufs; ++i) {
+      for (i = 0; i < NVC0_MAX_PIPE_CONSTBUFS; ++i) {
+         if (!(nvc0->constbuf_valid[s] & (1 << i)))
+            continue;
          if (!nvc0->constbuf[s][i].user &&
              nvc0->constbuf[s][i].u.buf == res) {
             nvc0->dirty |= NVC0_NEW_CONSTBUF;

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

@@ -171,6 +171,7 @@ nvc0_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_TGSI_FS_FINE_DERIVATIVE:
    case PIPE_CAP_CONDITIONAL_RENDER_INVERTED:
    case PIPE_CAP_SAMPLER_VIEW_TARGET:
+   case PIPE_CAP_CLIP_HALFZ:
       return 1;
    case PIPE_CAP_SEAMLESS_CUBE_MAP_PER_TEXTURE:
       return (class_3d >= NVE4_3D_CLASS) ? 1 : 0;
@@ -188,7 +189,6 @@ nvc0_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_TGSI_VS_LAYER_VIEWPORT:
    case PIPE_CAP_FAKE_SW_MSAA:
    case PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION:
-   case PIPE_CAP_CLIP_HALFZ:
       return 0;
 
    case PIPE_CAP_VENDOR_ID:

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

@@ -313,6 +313,8 @@ nvc0_rasterizer_state_create(struct pipe_context *pipe,
     SB_BEGIN_3D(so, VIEW_VOLUME_CLIP_CTRL, 1);
     SB_DATA    (so, reg);
 
+    SB_IMMED_3D(so, DEPTH_CLIP_NEGATIVE_Z, cso->clip_halfz);
+
     assert(so->size <= (sizeof(so->state) / sizeof(so->state[0])));
     return (void *)so;
 }

src/gallium/drivers/r600/r600_llvm.c

@@ -23,7 +23,6 @@
 
 #define CONSTANT_BUFFER_0_ADDR_SPACE 8
 #define CONSTANT_BUFFER_1_ADDR_SPACE (CONSTANT_BUFFER_0_ADDR_SPACE + R600_UCP_CONST_BUFFER)
-#define CONSTANT_TXQ_BUFFER (CONSTANT_BUFFER_0_ADDR_SPACE + R600_TXQ_CONST_BUFFER)
 #define LLVM_R600_BUFFER_INFO_CONST_BUFFER \
 	(CONSTANT_BUFFER_0_ADDR_SPACE + R600_BUFFER_INFO_CONST_BUFFER)
 
@@ -690,7 +689,7 @@ static void llvm_emit_tex(
 		if (emit_data->inst->Dst[0].Register.WriteMask & 4) {
 			LLVMValueRef offset = lp_build_const_int32(bld_base->base.gallivm, 0);
 			LLVMValueRef ZLayer = LLVMBuildExtractElement(gallivm->builder,
-				llvm_load_const_buffer(bld_base, offset, CONSTANT_TXQ_BUFFER),
+				llvm_load_const_buffer(bld_base, offset, LLVM_R600_BUFFER_INFO_CONST_BUFFER),
 				lp_build_const_int32(gallivm, 0), "");
 
 			emit_data->output[0] = LLVMBuildInsertElement(gallivm->builder, emit_data->output[0], ZLayer, lp_build_const_int32(gallivm, 2), "");

src/gallium/drivers/r600/r600_pipe.h

@@ -44,20 +44,19 @@
 #define R600_TRACE_CS_DWORDS		7
 
 #define R600_MAX_USER_CONST_BUFFERS 13
-#define R600_MAX_DRIVER_CONST_BUFFERS 4
+#define R600_MAX_DRIVER_CONST_BUFFERS 3
 #define R600_MAX_CONST_BUFFERS (R600_MAX_USER_CONST_BUFFERS + R600_MAX_DRIVER_CONST_BUFFERS)
 
 /* start driver buffers after user buffers */
 #define R600_UCP_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS)
-#define R600_TXQ_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS + 1)
-#define R600_BUFFER_INFO_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS + 2)
-#define R600_GS_RING_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS + 3)
-/* Currently R600_MAX_CONST_BUFFERS is too large, the hardware only has 16 buffers, but the driver is
- * trying to use 17. Avoid accidentally aliasing with user UBOs for SAMPLE_POSITIONS by using an id<16.
+#define R600_BUFFER_INFO_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS + 1)
+#define R600_GS_RING_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS + 2)
+/* Currently R600_MAX_CONST_BUFFERS just fits on the hw, which has a limit
+ * of 16 const buffers.
  * UCP/SAMPLE_POSITIONS are never accessed by same shader stage so they can use the same id.
  *
- * Fixing this properly would require the driver to combine its buffers into a single hardware buffer,
- * which would also allow supporting the d3d 11 mandated minimum of 15 user const buffers.
+ * In order to support d3d 11 mandated minimum of 15 user const buffers
+ * we'd have to squash all use cases into one driver buffer.
  */
 #define R600_SAMPLE_POSITIONS_CONST_BUFFER (R600_MAX_USER_CONST_BUFFERS)
 
@@ -316,7 +315,6 @@ struct r600_samplerview_state {
 	uint32_t			dirty_mask;
 	uint32_t			compressed_depthtex_mask; /* which textures are depth */
 	uint32_t			compressed_colortex_mask;
-	boolean                         dirty_txq_constants;
 	boolean				dirty_buffer_constants;
 };
 

src/gallium/drivers/r600/r600_shader.c

@@ -2728,8 +2728,10 @@ static int cayman_mul_int_instr(struct r600_shader_ctx *ctx)
 	struct tgsi_full_instruction *inst = &ctx->parse.FullToken.FullInstruction;
 	int i, j, k, r;
 	struct r600_bytecode_alu alu;
-	int last_slot = (inst->Dst[0].Register.WriteMask & 0x8) ? 4 : 3;
-	for (k = 0; k < last_slot; k++) {
+	int lasti = tgsi_last_instruction(inst->Dst[0].Register.WriteMask);
+	int t1 = ctx->temp_reg;
+
+	for (k = 0; k <= lasti; k++) {
 		if (!(inst->Dst[0].Register.WriteMask & (1 << k)))
 			continue;
 
@@ -2739,7 +2741,8 @@ static int cayman_mul_int_instr(struct r600_shader_ctx *ctx)
 			for (j = 0; j < inst->Instruction.NumSrcRegs; j++) {
 				r600_bytecode_src(&alu.src[j], &ctx->src[j], k);
 			}
-			tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
+			alu.dst.sel = t1;
+			alu.dst.chan = i;
 			alu.dst.write = (i == k);
 			if (i == 3)
 				alu.last = 1;
@@ -2748,6 +2751,23 @@ static int cayman_mul_int_instr(struct r600_shader_ctx *ctx)
 				return r;
 		}
 	}
+
+	for (i = 0 ; i <= lasti; i++) {
+		if (!(inst->Dst[0].Register.WriteMask & (1 << i)))
+			continue;
+		memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+		alu.op = ALU_OP1_MOV;
+		alu.src[0].sel = t1;
+		alu.src[0].chan = i;
+		tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
+		alu.dst.write = 1;
+		if (i == lasti)
+			alu.last = 1;
+		r = r600_bytecode_add_alu(ctx->bc, &alu);
+		if (r)
+			return r;
+	}
+
 	return 0;
 }
 
@@ -4899,7 +4919,8 @@ static inline boolean tgsi_tex_src_requires_loading(struct r600_shader_ctx *ctx,
 	return 	(inst->Src[index].Register.File != TGSI_FILE_TEMPORARY &&
 		inst->Src[index].Register.File != TGSI_FILE_INPUT &&
 		inst->Src[index].Register.File != TGSI_FILE_OUTPUT) ||
-		ctx->src[index].neg || ctx->src[index].abs;
+		ctx->src[index].neg || ctx->src[index].abs ||
+		(inst->Src[index].Register.File == TGSI_FILE_INPUT && ctx->type == TGSI_PROCESSOR_GEOMETRY);
 }
 
 static inline unsigned tgsi_tex_get_src_gpr(struct r600_shader_ctx *ctx,
@@ -5014,8 +5035,9 @@ static int r600_do_buffer_txq(struct r600_shader_ctx *ctx)
 	alu.op = ALU_OP1_MOV;
 
 	if (ctx->bc->chip_class >= EVERGREEN) {
-		alu.src[0].sel = 512 + (id / 4);
-		alu.src[0].chan = id % 4;
+		/* channel 0 or 2 of each word */
+		alu.src[0].sel = 512 + (id / 2);
+		alu.src[0].chan = (id % 2) * 2;
 	} else {
 		/* r600 we have them at channel 2 of the second dword */
 		alu.src[0].sel = 512 + (id * 2) + 1;
@@ -5074,6 +5096,14 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 	    inst->Instruction.Opcode == TGSI_OPCODE_TG4)
 		sampler_src_reg = 2;
 
+	/* TGSI moves the sampler to src reg 3 for TXD */
+	if (inst->Instruction.Opcode == TGSI_OPCODE_TXD)
+		sampler_src_reg = 3;
+
+	sampler_index_mode = inst->Src[sampler_src_reg].Indirect.Index == 2 ? 2 : 0; // CF_INDEX_1 : CF_INDEX_NONE
+	if (sampler_index_mode)
+		ctx->shader->uses_index_registers = true;
+
 	src_gpr = tgsi_tex_get_src_gpr(ctx, 0);
 
 	if (inst->Texture.Texture == TGSI_TEXTURE_BUFFER) {
@@ -5088,64 +5118,7 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 		}
 	}
 
-	if (inst->Instruction.Opcode == TGSI_OPCODE_TXD) {
-		/* TGSI moves the sampler to src reg 3 for TXD */
-		sampler_src_reg = 3;
-
-		sampler_index_mode = inst->Src[sampler_src_reg].Indirect.Index == 2 ? 2 : 0; // CF_INDEX_1 : CF_INDEX_NONE
-
-		for (i = 1; i < 3; i++) {
-			/* set gradients h/v */
-			memset(&tex, 0, sizeof(struct r600_bytecode_tex));
-			tex.op = (i == 1) ? FETCH_OP_SET_GRADIENTS_H :
-				FETCH_OP_SET_GRADIENTS_V;
-			tex.sampler_id = tgsi_tex_get_src_gpr(ctx, sampler_src_reg);
-			tex.sampler_index_mode = sampler_index_mode;
-			tex.resource_id = tex.sampler_id + R600_MAX_CONST_BUFFERS;
-			tex.resource_index_mode = sampler_index_mode;
-
-			if (tgsi_tex_src_requires_loading(ctx, i)) {
-				tex.src_gpr = r600_get_temp(ctx);
-				tex.src_sel_x = 0;
-				tex.src_sel_y = 1;
-				tex.src_sel_z = 2;
-				tex.src_sel_w = 3;
-
-				for (j = 0; j < 4; j++) {
-					memset(&alu, 0, sizeof(struct r600_bytecode_alu));
-					alu.op = ALU_OP1_MOV;
-                                        r600_bytecode_src(&alu.src[0], &ctx->src[i], j);
-                                        alu.dst.sel = tex.src_gpr;
-                                        alu.dst.chan = j;
-                                        if (j == 3)
-                                                alu.last = 1;
-                                        alu.dst.write = 1;
-                                        r = r600_bytecode_add_alu(ctx->bc, &alu);
-                                        if (r)
-                                                return r;
-				}
-
-			} else {
-				tex.src_gpr = tgsi_tex_get_src_gpr(ctx, i);
-				tex.src_sel_x = ctx->src[i].swizzle[0];
-				tex.src_sel_y = ctx->src[i].swizzle[1];
-				tex.src_sel_z = ctx->src[i].swizzle[2];
-				tex.src_sel_w = ctx->src[i].swizzle[3];
-				tex.src_rel = ctx->src[i].rel;
-			}
-			tex.dst_gpr = ctx->temp_reg; /* just to avoid confusing the asm scheduler */
-			tex.dst_sel_x = tex.dst_sel_y = tex.dst_sel_z = tex.dst_sel_w = 7;
-			if (inst->Texture.Texture != TGSI_TEXTURE_RECT) {
-				tex.coord_type_x = 1;
-				tex.coord_type_y = 1;
-				tex.coord_type_z = 1;
-				tex.coord_type_w = 1;
-			}
-			r = r600_bytecode_add_tex(ctx->bc, &tex);
-			if (r)
-				return r;
-		}
-	} else if (inst->Instruction.Opcode == TGSI_OPCODE_TXP) {
+	if (inst->Instruction.Opcode == TGSI_OPCODE_TXP) {
 		int out_chan;
 		/* Add perspective divide */
 		if (ctx->bc->chip_class == CAYMAN) {
@@ -5209,9 +5182,6 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 		src_gpr = ctx->temp_reg;
 	}
 
-	sampler_index_mode = inst->Src[sampler_src_reg].Indirect.Index == 2 ? 2 : 0; // CF_INDEX_1 : CF_INDEX_NONE
-	if (sampler_index_mode)
-		ctx->shader->uses_index_registers = true;
 
 	if ((inst->Texture.Texture == TGSI_TEXTURE_CUBE ||
 	     inst->Texture.Texture == TGSI_TEXTURE_CUBE_ARRAY ||
@@ -5430,6 +5400,69 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 		src_gpr = ctx->temp_reg;
 	}
 
+	if (inst->Instruction.Opcode == TGSI_OPCODE_TXD) {
+		int temp_h = 0, temp_v = 0;
+		int start_val = 0;
+
+		/* if we've already loaded the src (i.e. CUBE don't reload it). */
+		if (src_loaded == TRUE)
+			start_val = 1;
+		else
+			src_loaded = TRUE;
+		for (i = start_val; i < 3; i++) {
+			int treg = r600_get_temp(ctx);
+
+			if (i == 0)
+				src_gpr = treg;
+			else if (i == 1)
+				temp_h = treg;
+			else
+				temp_v = treg;
+
+			for (j = 0; j < 4; j++) {
+				memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+				alu.op = ALU_OP1_MOV;
+                                r600_bytecode_src(&alu.src[0], &ctx->src[i], j);
+                                alu.dst.sel = treg;
+                                alu.dst.chan = j;
+                                if (j == 3)
+                                   alu.last = 1;
+                                alu.dst.write = 1;
+                                r = r600_bytecode_add_alu(ctx->bc, &alu);
+                                if (r)
+                                    return r;
+			}
+		}
+		for (i = 1; i < 3; i++) {
+			/* set gradients h/v */
+			memset(&tex, 0, sizeof(struct r600_bytecode_tex));
+			tex.op = (i == 1) ? FETCH_OP_SET_GRADIENTS_H :
+				FETCH_OP_SET_GRADIENTS_V;
+			tex.sampler_id = tgsi_tex_get_src_gpr(ctx, sampler_src_reg);
+			tex.sampler_index_mode = sampler_index_mode;
+			tex.resource_id = tex.sampler_id + R600_MAX_CONST_BUFFERS;
+			tex.resource_index_mode = sampler_index_mode;
+
+			tex.src_gpr = (i == 1) ? temp_h : temp_v;
+			tex.src_sel_x = 0;
+			tex.src_sel_y = 1;
+			tex.src_sel_z = 2;
+			tex.src_sel_w = 3;
+
+			tex.dst_gpr = r600_get_temp(ctx); /* just to avoid confusing the asm scheduler */
+			tex.dst_sel_x = tex.dst_sel_y = tex.dst_sel_z = tex.dst_sel_w = 7;
+			if (inst->Texture.Texture != TGSI_TEXTURE_RECT) {
+				tex.coord_type_x = 1;
+				tex.coord_type_y = 1;
+				tex.coord_type_z = 1;
+				tex.coord_type_w = 1;
+			}
+			r = r600_bytecode_add_tex(ctx->bc, &tex);
+			if (r)
+				return r;
+		}
+	}
+
 	if (src_requires_loading && !src_loaded) {
 		for (i = 0; i < 4; i++) {
 			memset(&alu, 0, sizeof(struct r600_bytecode_alu));
@@ -5514,9 +5547,24 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 				/* texture offsets do not apply to other texture targets */
 			}
 		} else {
-			offset_x = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleX] << 1;
-			offset_y = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleY] << 1;
-			offset_z = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleZ] << 1;
+			switch (inst->Texture.Texture) {
+			case TGSI_TEXTURE_3D:
+				offset_z = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleZ] << 1;
+				/* fallthrough */
+			case TGSI_TEXTURE_2D:
+			case TGSI_TEXTURE_SHADOW2D:
+			case TGSI_TEXTURE_RECT:
+			case TGSI_TEXTURE_SHADOWRECT:
+			case TGSI_TEXTURE_2D_ARRAY:
+			case TGSI_TEXTURE_SHADOW2D_ARRAY:
+				offset_y = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleY] << 1;
+				/* fallthrough */
+			case TGSI_TEXTURE_1D:
+			case TGSI_TEXTURE_SHADOW1D:
+			case TGSI_TEXTURE_1D_ARRAY:
+			case TGSI_TEXTURE_SHADOW1D_ARRAY:
+				offset_x = ctx->literals[4 * inst->TexOffsets[0].Index + inst->TexOffsets[0].SwizzleX] << 1;
+			}
 		}
 	}
 
@@ -5650,9 +5698,16 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 		memset(&alu, 0, sizeof(struct r600_bytecode_alu));
 		alu.op = ALU_OP1_MOV;
 
-		alu.src[0].sel = 512 + (id / 4);
-		alu.src[0].kc_bank = R600_TXQ_CONST_BUFFER;
-		alu.src[0].chan = id % 4;
+		if (ctx->bc->chip_class >= EVERGREEN) {
+			/* channel 1 or 3 of each word */
+			alu.src[0].sel = 512 + (id / 2);
+			alu.src[0].chan = ((id % 2) * 2) + 1;
+		} else {
+			/* r600 we have them at channel 2 of the second dword */
+			alu.src[0].sel = 512 + (id * 2) + 1;
+			alu.src[0].chan = 2;
+		}
+		alu.src[0].kc_bank = R600_BUFFER_INFO_CONST_BUFFER;
 		tgsi_dst(ctx, &inst->Dst[0], 2, &alu.dst);
 		alu.last = 1;
 		r = r600_bytecode_add_alu(ctx->bc, &alu);
@@ -5743,11 +5798,18 @@ static int tgsi_tex(struct r600_shader_ctx *ctx)
 		int8_t texture_component_select = ctx->literals[4 * inst->Src[1].Register.Index + inst->Src[1].Register.SwizzleX];
 		tex.inst_mod = texture_component_select;
 
+		if (ctx->bc->chip_class == CAYMAN) {
 		/* GATHER4 result order is different from TGSI TG4 */
-		tex.dst_sel_x = (inst->Dst[0].Register.WriteMask & 2) ? 1 : 7;
-		tex.dst_sel_y = (inst->Dst[0].Register.WriteMask & 4) ? 2 : 7;
-		tex.dst_sel_z = (inst->Dst[0].Register.WriteMask & 1) ? 0 : 7;
-		tex.dst_sel_w = (inst->Dst[0].Register.WriteMask & 8) ? 3 : 7;
+			tex.dst_sel_x = (inst->Dst[0].Register.WriteMask & 2) ? 0 : 7;
+			tex.dst_sel_y = (inst->Dst[0].Register.WriteMask & 4) ? 1 : 7;
+			tex.dst_sel_z = (inst->Dst[0].Register.WriteMask & 1) ? 2 : 7;
+			tex.dst_sel_w = (inst->Dst[0].Register.WriteMask & 8) ? 3 : 7;
+		} else {
+			tex.dst_sel_x = (inst->Dst[0].Register.WriteMask & 2) ? 1 : 7;
+			tex.dst_sel_y = (inst->Dst[0].Register.WriteMask & 4) ? 2 : 7;
+			tex.dst_sel_z = (inst->Dst[0].Register.WriteMask & 1) ? 0 : 7;
+			tex.dst_sel_w = (inst->Dst[0].Register.WriteMask & 8) ? 3 : 7;
+		}
 	}
 	else if (inst->Instruction.Opcode == TGSI_OPCODE_LODQ) {
 		tex.dst_sel_x = (inst->Dst[0].Register.WriteMask & 2) ? 1 : 7;

src/gallium/drivers/r600/r600_state.c

@@ -2659,11 +2659,8 @@ void r600_update_gs_state(struct pipe_context *ctx, struct r600_pipe_shader *sha
 	r600_store_context_reg(cb, R_028A6C_VGT_GS_OUT_PRIM_TYPE,
 			       r600_conv_prim_to_gs_out(rshader->gs_output_prim));
 
-	r600_store_context_reg_seq(cb, R_0288C8_SQ_GS_VERT_ITEMSIZE, 4);
-	r600_store_value(cb, cp_shader->ring_item_size >> 2);
-	r600_store_value(cb, 0);
-	r600_store_value(cb, 0);
-	r600_store_value(cb, 0);
+	r600_store_context_reg(cb, R_0288C8_SQ_GS_VERT_ITEMSIZE,
+	                       cp_shader->ring_item_size >> 2);
 
 	r600_store_context_reg(cb, R_0288A8_SQ_ESGS_RING_ITEMSIZE,
 			       (rshader->ring_item_size) >> 2);

src/gallium/drivers/r600/r600_state_common.c

@@ -649,7 +649,6 @@ static void r600_set_sampler_views(struct pipe_context *pipe, unsigned shader,
 	dst->views.dirty_mask |= new_mask;
 	dst->views.compressed_depthtex_mask &= dst->views.enabled_mask;
 	dst->views.compressed_colortex_mask &= dst->views.enabled_mask;
-	dst->views.dirty_txq_constants = TRUE;
 	dst->views.dirty_buffer_constants = TRUE;
 	r600_sampler_views_dirty(rctx, &dst->views);
 
@@ -984,6 +983,7 @@ static void r600_set_sample_mask(struct pipe_context *pipe, unsigned sample_mask
  * then in the shader, we AND the 4 components with 0xffffffff or 0,
  * then OR the alpha with the value given here.
  * We use a 6th constant to store the txq buffer size in
+ * we use 7th slot for number of cube layers in a cube map array.
  */
 static void r600_setup_buffer_constants(struct r600_context *rctx, int shader_type)
 {
@@ -1022,6 +1022,7 @@ static void r600_setup_buffer_constants(struct r600_context *rctx, int shader_ty
 				samplers->buffer_constants[offset + 4] = 0;
 
 			samplers->buffer_constants[offset + 5] = samplers->views.views[i]->base.texture->width0 / util_format_get_blocksize(samplers->views.views[i]->base.format);
+			samplers->buffer_constants[offset + 6] = samplers->views.views[i]->base.texture->array_size / 6;
 		}
 	}
 
@@ -1033,7 +1034,10 @@ static void r600_setup_buffer_constants(struct r600_context *rctx, int shader_ty
 	pipe_resource_reference(&cb.buffer, NULL);
 }
 
-/* On evergreen we only need to store the buffer size for TXQ */
+/* On evergreen we store two values
+ * 1. buffer size for TXQ
+ * 2. number of cube layers in a cube map array.
+ */
 static void eg_setup_buffer_constants(struct r600_context *rctx, int shader_type)
 {
 	struct r600_textures_info *samplers = &rctx->samplers[shader_type];
@@ -1048,12 +1052,16 @@ static void eg_setup_buffer_constants(struct r600_context *rctx, int shader_type
 	samplers->views.dirty_buffer_constants = FALSE;
 
 	bits = util_last_bit(samplers->views.enabled_mask);
-	array_size = bits * sizeof(uint32_t) * 4;
+	array_size = bits * 2 * sizeof(uint32_t) * 4;
 	samplers->buffer_constants = realloc(samplers->buffer_constants, array_size);
 	memset(samplers->buffer_constants, 0, array_size);
-	for (i = 0; i < bits; i++)
-		if (samplers->views.enabled_mask & (1 << i))
-		   samplers->buffer_constants[i] = samplers->views.views[i]->base.texture->width0 / util_format_get_blocksize(samplers->views.views[i]->base.format);
+	for (i = 0; i < bits; i++) {
+		if (samplers->views.enabled_mask & (1 << i)) {
+			uint32_t offset = i * 2;
+			samplers->buffer_constants[offset] = samplers->views.views[i]->base.texture->width0 / util_format_get_blocksize(samplers->views.views[i]->base.format);
+			samplers->buffer_constants[offset + 1] = samplers->views.views[i]->base.texture->array_size / 6;
+		}
+	}
 
 	cb.buffer = NULL;
 	cb.user_buffer = samplers->buffer_constants;
@@ -1063,35 +1071,6 @@ static void eg_setup_buffer_constants(struct r600_context *rctx, int shader_type
 	pipe_resource_reference(&cb.buffer, NULL);
 }
 
-static void r600_setup_txq_cube_array_constants(struct r600_context *rctx, int shader_type)
-{
-	struct r600_textures_info *samplers = &rctx->samplers[shader_type];
-	int bits;
-	uint32_t array_size;
-	struct pipe_constant_buffer cb;
-	int i;
-
-	if (!samplers->views.dirty_txq_constants)
-		return;
-
-	samplers->views.dirty_txq_constants = FALSE;
-
-	bits = util_last_bit(samplers->views.enabled_mask);
-	array_size = bits * sizeof(uint32_t) * 4;
-	samplers->txq_constants = realloc(samplers->txq_constants, array_size);
-	memset(samplers->txq_constants, 0, array_size);
-	for (i = 0; i < bits; i++)
-		if (samplers->views.enabled_mask & (1 << i))
-			samplers->txq_constants[i] = samplers->views.views[i]->base.texture->array_size / 6;
-
-	cb.buffer = NULL;
-	cb.user_buffer = samplers->txq_constants;
-	cb.buffer_offset = 0;
-	cb.buffer_size = array_size;
-	rctx->b.b.set_constant_buffer(&rctx->b.b, shader_type, R600_TXQ_CONST_BUFFER, &cb);
-	pipe_resource_reference(&cb.buffer, NULL);
-}
-
 /* set sample xy locations as array of fragment shader constants */
 void r600_set_sample_locations_constant_buffer(struct r600_context *rctx)
 {
@@ -1175,7 +1154,7 @@ static bool r600_update_derived_state(struct r600_context *rctx)
 	struct pipe_context * ctx = (struct pipe_context*)rctx;
 	bool ps_dirty = false, vs_dirty = false, gs_dirty = false;
 	bool blend_disable;
-
+	bool need_buf_const;
 	if (!rctx->blitter->running) {
 		unsigned i;
 
@@ -1296,29 +1275,35 @@ static bool r600_update_derived_state(struct r600_context *rctx)
 
 	/* on R600 we stuff masks + txq info into one constant buffer */
 	/* on evergreen we only need a txq info one */
-	if (rctx->b.chip_class < EVERGREEN) {
-		if (rctx->ps_shader && rctx->ps_shader->current->shader.uses_tex_buffers)
-			r600_setup_buffer_constants(rctx, PIPE_SHADER_FRAGMENT);
-		if (rctx->vs_shader && rctx->vs_shader->current->shader.uses_tex_buffers)
-			r600_setup_buffer_constants(rctx, PIPE_SHADER_VERTEX);
-		if (rctx->gs_shader && rctx->gs_shader->current->shader.uses_tex_buffers)
-			r600_setup_buffer_constants(rctx, PIPE_SHADER_GEOMETRY);
-	} else {
-		if (rctx->ps_shader && rctx->ps_shader->current->shader.uses_tex_buffers)
-			eg_setup_buffer_constants(rctx, PIPE_SHADER_FRAGMENT);
-		if (rctx->vs_shader && rctx->vs_shader->current->shader.uses_tex_buffers)
-			eg_setup_buffer_constants(rctx, PIPE_SHADER_VERTEX);
-		if (rctx->gs_shader && rctx->gs_shader->current->shader.uses_tex_buffers)
-			eg_setup_buffer_constants(rctx, PIPE_SHADER_GEOMETRY);
+	if (rctx->ps_shader) {
+		need_buf_const = rctx->ps_shader->current->shader.uses_tex_buffers || rctx->ps_shader->current->shader.has_txq_cube_array_z_comp;
+		if (need_buf_const) {
+			if (rctx->b.chip_class < EVERGREEN)
+				r600_setup_buffer_constants(rctx, PIPE_SHADER_FRAGMENT);
+			else
+				eg_setup_buffer_constants(rctx, PIPE_SHADER_FRAGMENT);
+		}
 	}
 
+	if (rctx->vs_shader) {
+		need_buf_const = rctx->vs_shader->current->shader.uses_tex_buffers || rctx->vs_shader->current->shader.has_txq_cube_array_z_comp;
+		if (need_buf_const) {
+			if (rctx->b.chip_class < EVERGREEN)
+				r600_setup_buffer_constants(rctx, PIPE_SHADER_VERTEX);
+			else
+				eg_setup_buffer_constants(rctx, PIPE_SHADER_VERTEX);
+		}
+	}
 
-	if (rctx->ps_shader && rctx->ps_shader->current->shader.has_txq_cube_array_z_comp)
-		r600_setup_txq_cube_array_constants(rctx, PIPE_SHADER_FRAGMENT);
-	if (rctx->vs_shader && rctx->vs_shader->current->shader.has_txq_cube_array_z_comp)
-		r600_setup_txq_cube_array_constants(rctx, PIPE_SHADER_VERTEX);
-	if (rctx->gs_shader && rctx->gs_shader->current->shader.has_txq_cube_array_z_comp)
-		r600_setup_txq_cube_array_constants(rctx, PIPE_SHADER_GEOMETRY);
+	if (rctx->gs_shader) {
+		need_buf_const = rctx->gs_shader->current->shader.uses_tex_buffers || rctx->gs_shader->current->shader.has_txq_cube_array_z_comp;
+		if (need_buf_const) {
+			if (rctx->b.chip_class < EVERGREEN)
+				r600_setup_buffer_constants(rctx, PIPE_SHADER_GEOMETRY);
+			else
+				eg_setup_buffer_constants(rctx, PIPE_SHADER_GEOMETRY);
+		}
+	}
 
 	if (rctx->b.chip_class < EVERGREEN && rctx->ps_shader && rctx->vs_shader) {
 		if (!r600_adjust_gprs(rctx)) {

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

@@ -46,15 +46,22 @@ int bc_finalizer::run() {
 	for (regions_vec::reverse_iterator I = rv.rbegin(), E = rv.rend(); I != E;
 			++I) {
 		region_node *r = *I;
-
+		bool is_if = false;
 		assert(r);
 
-		bool loop = r->is_loop();
+		assert(r->first);
+		if (r->first->is_container()) {
+			container_node *repdep1 = static_cast<container_node*>(r->first);
+			assert(repdep1->is_depart() || repdep1->is_repeat());
+			if_node *n_if = static_cast<if_node*>(repdep1->first);
+			if (n_if && n_if->is_if())
+				is_if = true;
+		}
 
-		if (loop)
-			finalize_loop(r);
-		else
+		if (is_if)
 			finalize_if(r);
+		else
+			finalize_loop(r);
 
 		r->expand();
 	}
@@ -83,14 +90,18 @@ int bc_finalizer::run() {
 		last_cf = c;
 	}
 
-	if (last_cf->bc.op_ptr->flags & CF_ALU) {
+	if (!ctx.is_cayman() && last_cf->bc.op_ptr->flags & CF_ALU) {
 		last_cf = sh.create_cf(CF_OP_NOP);
 		sh.root->push_back(last_cf);
 	}
 
-	if (ctx.is_cayman())
-		last_cf->insert_after(sh.create_cf(CF_OP_CF_END));
-	else
+	if (ctx.is_cayman()) {
+		if (!last_cf) {
+			cf_node *c = sh.create_cf(CF_OP_CF_END);
+			sh.root->push_back(c);
+		} else
+			last_cf->insert_after(sh.create_cf(CF_OP_CF_END));
+	} else
 		last_cf->bc.end_of_program = 1;
 
 	for (unsigned t = EXP_PIXEL; t < EXP_TYPE_COUNT; ++t) {
@@ -108,16 +119,33 @@ void bc_finalizer::finalize_loop(region_node* r) {
 
 	cf_node *loop_start = sh.create_cf(CF_OP_LOOP_START_DX10);
 	cf_node *loop_end = sh.create_cf(CF_OP_LOOP_END);
+	bool has_instr = false;
 
-	loop_start->jump_after(loop_end);
-	loop_end->jump_after(loop_start);
+	if (!r->is_loop()) {
+		for (depart_vec::iterator I = r->departs.begin(), E = r->departs.end();
+		     I != E; ++I) {
+			depart_node *dep = *I;
+			if (!dep->empty()) {
+				has_instr = true;
+				break;
+			}
+		}
+	} else
+		has_instr = true;
+
+	if (has_instr) {
+		loop_start->jump_after(loop_end);
+		loop_end->jump_after(loop_start);
+	}
 
 	for (depart_vec::iterator I = r->departs.begin(), E = r->departs.end();
 			I != E; ++I) {
 		depart_node *dep = *I;
-		cf_node *loop_break = sh.create_cf(CF_OP_LOOP_BREAK);
-		loop_break->jump(loop_end);
-		dep->push_back(loop_break);
+		if (has_instr) {
+			cf_node *loop_break = sh.create_cf(CF_OP_LOOP_BREAK);
+			loop_break->jump(loop_end);
+			dep->push_back(loop_break);
+		}
 		dep->expand();
 	}
 
@@ -133,8 +161,10 @@ void bc_finalizer::finalize_loop(region_node* r) {
 		rep->expand();
 	}
 
-	r->push_front(loop_start);
-	r->push_back(loop_end);
+	if (has_instr) {
+		r->push_front(loop_start);
+		r->push_back(loop_end);
+	}
 }
 
 void bc_finalizer::finalize_if(region_node* r) {

src/gallium/drivers/radeonsi/si_dma.c

@@ -250,6 +250,21 @@ void si_dma_copy(struct pipe_context *ctx,
 		return;
 	}
 
+	/* XXX: Using the asynchronous DMA engine for multi-dimensional
+	 * operations seems to cause random GPU lockups for various people.
+	 * While the root cause for this might need to be fixed in the kernel,
+	 * let's disable it for now.
+	 *
+	 * Before re-enabling this, please make sure you can hit all newly
+	 * enabled paths in your testing, preferably with both piglit and real
+	 * world apps, and get in touch with people on the bug reports below
+	 * for stability testing.
+	 *
+	 * https://bugs.freedesktop.org/show_bug.cgi?id=85647
+	 * https://bugs.freedesktop.org/show_bug.cgi?id=83500
+	 */
+	goto fallback;
+
 	if (src->format != dst->format || src_box->depth > 1 ||
 	    rdst->dirty_level_mask != 0 ||
 	    rdst->cmask.size || rdst->fmask.size ||

src/gallium/drivers/radeonsi/si_state.c

@@ -3081,6 +3081,110 @@ void si_init_state_functions(struct si_context *sctx)
 	sctx->b.b.draw_vbo = si_draw_vbo;
 }
 
+static void
+si_write_harvested_raster_configs(struct si_context *sctx,
+				  struct si_pm4_state *pm4,
+				  unsigned raster_config)
+{
+	unsigned sh_per_se = MAX2(sctx->screen->b.info.max_sh_per_se, 1);
+	unsigned num_se = MAX2(sctx->screen->b.info.max_se, 1);
+	unsigned rb_mask = sctx->screen->b.info.si_backend_enabled_mask;
+	unsigned num_rb = sctx->screen->b.info.r600_num_backends;
+	unsigned rb_per_pkr = num_rb / num_se / sh_per_se;
+	unsigned rb_per_se = num_rb / num_se;
+	unsigned se0_mask = (1 << rb_per_se) - 1;
+	unsigned se1_mask = se0_mask << rb_per_se;
+	unsigned se;
+
+	assert(num_se == 1 || num_se == 2);
+	assert(sh_per_se == 1 || sh_per_se == 2);
+	assert(rb_per_pkr == 1 || rb_per_pkr == 2);
+
+	/* XXX: I can't figure out what the *_XSEL and *_YSEL
+	 * fields are for, so I'm leaving them as their default
+	 * values. */
+
+	se0_mask &= rb_mask;
+	se1_mask &= rb_mask;
+	if (num_se == 2 && (!se0_mask || !se1_mask)) {
+		raster_config &= C_028350_SE_MAP;
+
+		if (!se0_mask) {
+			raster_config |=
+				S_028350_SE_MAP(V_028350_RASTER_CONFIG_SE_MAP_3);
+		} else {
+			raster_config |=
+				S_028350_SE_MAP(V_028350_RASTER_CONFIG_SE_MAP_0);
+		}
+	}
+
+	for (se = 0; se < num_se; se++) {
+		unsigned raster_config_se = raster_config;
+		unsigned pkr0_mask = ((1 << rb_per_pkr) - 1) << (se * rb_per_se);
+		unsigned pkr1_mask = pkr0_mask << rb_per_pkr;
+
+		pkr0_mask &= rb_mask;
+		pkr1_mask &= rb_mask;
+		if (sh_per_se == 2 && (!pkr0_mask || !pkr1_mask)) {
+			raster_config_se &= C_028350_PKR_MAP;
+
+			if (!pkr0_mask) {
+				raster_config_se |=
+					S_028350_PKR_MAP(V_028350_RASTER_CONFIG_PKR_MAP_3);
+			} else {
+				raster_config_se |=
+					S_028350_PKR_MAP(V_028350_RASTER_CONFIG_PKR_MAP_0);
+			}
+		}
+
+		if (rb_per_pkr == 2) {
+			unsigned rb0_mask = 1 << (se * rb_per_se);
+			unsigned rb1_mask = rb0_mask << 1;
+
+			rb0_mask &= rb_mask;
+			rb1_mask &= rb_mask;
+			if (!rb0_mask || !rb1_mask) {
+				raster_config_se &= C_028350_RB_MAP_PKR0;
+
+				if (!rb0_mask) {
+					raster_config_se |=
+						S_028350_RB_MAP_PKR0(V_028350_RASTER_CONFIG_RB_MAP_3);
+				} else {
+					raster_config_se |=
+						S_028350_RB_MAP_PKR0(V_028350_RASTER_CONFIG_RB_MAP_0);
+				}
+			}
+
+			if (sh_per_se == 2) {
+				rb0_mask = 1 << (se * rb_per_se + rb_per_pkr);
+				rb1_mask = rb0_mask << 1;
+				rb0_mask &= rb_mask;
+				rb1_mask &= rb_mask;
+				if (!rb0_mask || !rb1_mask) {
+					raster_config_se &= C_028350_RB_MAP_PKR1;
+
+					if (!rb0_mask) {
+						raster_config_se |=
+							S_028350_RB_MAP_PKR1(V_028350_RASTER_CONFIG_RB_MAP_3);
+					} else {
+						raster_config_se |=
+							S_028350_RB_MAP_PKR1(V_028350_RASTER_CONFIG_RB_MAP_0);
+					}
+				}
+			}
+		}
+
+		si_pm4_set_reg(pm4, GRBM_GFX_INDEX,
+			       SE_INDEX(se) | SH_BROADCAST_WRITES |
+			       INSTANCE_BROADCAST_WRITES);
+		si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, raster_config_se);
+	}
+
+	si_pm4_set_reg(pm4, GRBM_GFX_INDEX,
+		       SE_BROADCAST_WRITES | SH_BROADCAST_WRITES |
+		       INSTANCE_BROADCAST_WRITES);
+}
+
 void si_init_config(struct si_context *sctx)
 {
 	struct si_pm4_state *pm4 = CALLOC_STRUCT(si_pm4_state);
@@ -3152,24 +3256,38 @@ void si_init_config(struct si_context *sctx)
 			break;
 		}
 	} else {
+		unsigned rb_mask = sctx->screen->b.info.si_backend_enabled_mask;
+		unsigned num_rb = sctx->screen->b.info.r600_num_backends;
+		unsigned raster_config;
+
 		switch (sctx->screen->b.family) {
 		case CHIP_TAHITI:
 		case CHIP_PITCAIRN:
-			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, 0x2a00126a);
+			raster_config = 0x2a00126a;
 			break;
 		case CHIP_VERDE:
-			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, 0x0000124a);
+			raster_config = 0x0000124a;
 			break;
 		case CHIP_OLAND:
-			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, 0x00000082);
+			raster_config = 0x00000082;
 			break;
 		case CHIP_HAINAN:
-			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, 0x00000000);
+			raster_config = 0x00000000;
 			break;
 		default:
-			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG, 0x00000000);
+			fprintf(stderr,
+				"radeonsi: Unknown GPU, using 0 for raster_config\n");
+			raster_config = 0x00000000;
 			break;
 		}
+
+		/* Always use the default config when all backends are enabled. */
+		if (rb_mask && util_bitcount(rb_mask) >= num_rb) {
+			si_pm4_set_reg(pm4, R_028350_PA_SC_RASTER_CONFIG,
+				       raster_config);
+		} else {
+			si_write_harvested_raster_configs(sctx, pm4, raster_config);
+		}
 	}
 
 	si_pm4_set_reg(pm4, R_028204_PA_SC_WINDOW_SCISSOR_TL, S_028204_WINDOW_OFFSET_DISABLE(1));

src/gallium/drivers/radeonsi/si_state_draw.c

@@ -237,20 +237,21 @@ static void si_shader_ps(struct si_shader *shader)
 	for (i = 0; i < info->num_inputs; i++) {
 		switch (info->input_semantic_name[i]) {
 		case TGSI_SEMANTIC_POSITION:
-			if (info->input_interpolate_loc[i] ==
-			    TGSI_INTERPOLATE_LOC_CENTROID) {
-				/* SPI_BARYC_CNTL.POS_FLOAT_LOCATION
-				 * Possible vaules:
-				 * 0 -> Position = pixel center (default)
-				 * 1 -> Position = pixel centroid
-				 * 2 -> Position = iterated sample number XXX:
-				 *                        What does this mean?
-			 	 */
+			/* SPI_BARYC_CNTL.POS_FLOAT_LOCATION
+			 * Possible vaules:
+			 * 0 -> Position = pixel center (default)
+			 * 1 -> Position = pixel centroid
+			 * 2 -> Position = at sample position
+			 */
+			switch (info->input_interpolate_loc[i]) {
+			case TGSI_INTERPOLATE_LOC_CENTROID:
 				spi_baryc_cntl |= S_0286E0_POS_FLOAT_LOCATION(1);
+				break;
+			case TGSI_INTERPOLATE_LOC_SAMPLE:
+				spi_baryc_cntl |= S_0286E0_POS_FLOAT_LOCATION(2);
+				break;
 			}
-			/* Fall through */
-		case TGSI_SEMANTIC_FACE:
-			continue;
+			break;
 		}
 	}
 

src/gallium/drivers/radeonsi/sid.h

@@ -204,7 +204,13 @@
  * 6. COMMAND [29:22] | BYTE_COUNT [20:0]
  */
 
-
+#define GRBM_GFX_INDEX                                                  0x802C
+#define         INSTANCE_INDEX(x)                                     ((x) << 0)
+#define         SH_INDEX(x)                                           ((x) << 8)
+#define         SE_INDEX(x)                                           ((x) << 16)
+#define         SH_BROADCAST_WRITES                                   (1 << 29)
+#define         INSTANCE_BROADCAST_WRITES                             (1 << 30)
+#define         SE_BROADCAST_WRITES                                   (1 << 31)
 #define R_0084FC_CP_STRMOUT_CNTL		                        0x0084FC
 #define   S_0084FC_OFFSET_UPDATE_DONE(x)		              (((x) & 0x1) << 0)
 #define R_0085F0_CP_COHER_CNTL                                          0x0085F0

src/gallium/state_trackers/nine/adapter9.c

@@ -990,7 +990,7 @@ NineAdapter9_CreateDevice( struct NineAdapter9 *This,
     params.BehaviorFlags = BehaviorFlags;
 
     hr = NineDevice9_new(screen, &params, &caps, pPresentationParameters,
-                         pD3D9, pPresentationGroup, This->ctx,
+                         pD3D9, pPresentationGroup, This->ctx, FALSE, NULL,
                          (struct NineDevice9 **)ppReturnedDeviceInterface);
     if (FAILED(hr)) {
         DBG("Failed to create device.\n");

src/gallium/state_trackers/nine/basetexture9.c

@@ -41,25 +41,31 @@
 HRESULT
 NineBaseTexture9_ctor( struct NineBaseTexture9 *This,
                        struct NineUnknownParams *pParams,
+                       struct pipe_resource *initResource,
                        D3DRESOURCETYPE Type,
-                       D3DPOOL Pool )
+                       D3DFORMAT format,
+                       D3DPOOL Pool,
+                       DWORD Usage)
 {
-    BOOL alloc = (Pool == D3DPOOL_DEFAULT) && !This->base.resource &&
-        (This->format != D3DFMT_NULL);
+    BOOL alloc = (Pool == D3DPOOL_DEFAULT) && !initResource &&
+        (format != D3DFMT_NULL);
     HRESULT hr;
-    DWORD usage = This->base.usage;
 
-    user_assert(!(usage & (D3DUSAGE_RENDERTARGET | D3DUSAGE_DEPTHSTENCIL)) ||
+    DBG("This=%p, pParams=%p initResource=%p Type=%d format=%d Pool=%d Usage=%d\n",
+        This, pParams, initResource, Type, format, Pool, Usage);
+
+    user_assert(!(Usage & (D3DUSAGE_RENDERTARGET | D3DUSAGE_DEPTHSTENCIL)) ||
                 Pool == D3DPOOL_DEFAULT, D3DERR_INVALIDCALL);
-    user_assert(!(usage & D3DUSAGE_DYNAMIC) ||
+    user_assert(!(Usage & D3DUSAGE_DYNAMIC) ||
                 Pool != D3DPOOL_MANAGED, D3DERR_INVALIDCALL);
 
-    hr = NineResource9_ctor(&This->base, pParams, alloc, Type, Pool);
+    hr = NineResource9_ctor(&This->base, pParams, initResource, alloc, Type, Pool, Usage);
     if (FAILED(hr))
         return hr;
 
+    This->format = format;
     This->pipe = pParams->device->pipe;
-    This->mipfilter = (This->base.usage & D3DUSAGE_AUTOGENMIPMAP) ?
+    This->mipfilter = (Usage & D3DUSAGE_AUTOGENMIPMAP) ?
         D3DTEXF_LINEAR : D3DTEXF_NONE;
     This->lod = 0;
     This->lod_resident = -1;
@@ -90,6 +96,8 @@ NineBaseTexture9_SetLOD( struct NineBaseTexture9 *This,
 {
     DWORD old = This->lod;
 
+    DBG("This=%p LODNew=%d\n", This, LODNew);
+
     user_assert(This->base.pool == D3DPOOL_MANAGED, 0);
 
     This->lod = MIN2(LODNew, This->base.info.last_level);
@@ -103,12 +111,16 @@ NineBaseTexture9_SetLOD( struct NineBaseTexture9 *This,
 DWORD WINAPI
 NineBaseTexture9_GetLOD( struct NineBaseTexture9 *This )
 {
+    DBG("This=%p\n", This);
+
     return This->lod;
 }
 
 DWORD WINAPI
 NineBaseTexture9_GetLevelCount( struct NineBaseTexture9 *This )
 {
+    DBG("This=%p\n", This);
+
     if (This->base.usage & D3DUSAGE_AUTOGENMIPMAP)
         return 1;
     return This->base.info.last_level + 1;
@@ -118,6 +130,8 @@ HRESULT WINAPI
 NineBaseTexture9_SetAutoGenFilterType( struct NineBaseTexture9 *This,
                                        D3DTEXTUREFILTERTYPE FilterType )
 {
+    DBG("This=%p FilterType=%d\n", This, FilterType);
+
     if (!(This->base.usage & D3DUSAGE_AUTOGENMIPMAP))
         return D3D_OK;
     user_assert(FilterType != D3DTEXF_NONE, D3DERR_INVALIDCALL);
@@ -130,6 +144,8 @@ NineBaseTexture9_SetAutoGenFilterType( struct NineBaseTexture9 *This,
 D3DTEXTUREFILTERTYPE WINAPI
 NineBaseTexture9_GetAutoGenFilterType( struct NineBaseTexture9 *This )
 {
+    DBG("This=%p\n", This);
+
     return This->mipfilter;
 }
 
@@ -430,6 +446,8 @@ NineBaseTexture9_UpdateSamplerView( struct NineBaseTexture9 *This,
     struct pipe_sampler_view templ;
     uint8_t swizzle[4];
 
+    DBG("This=%p sRGB=%d\n", This, sRGB);
+
     if (unlikely(!resource)) {
 	if (unlikely(This->format == D3DFMT_NULL))
             return D3D_OK;
@@ -482,6 +500,8 @@ NineBaseTexture9_UpdateSamplerView( struct NineBaseTexture9 *This,
 void WINAPI
 NineBaseTexture9_PreLoad( struct NineBaseTexture9 *This )
 {
+    DBG("This=%p\n", This);
+
     if (This->dirty && This->base.pool == D3DPOOL_MANAGED)
         NineBaseTexture9_UploadSelf(This);
 }
@@ -490,9 +510,9 @@ NineBaseTexture9_PreLoad( struct NineBaseTexture9 *This )
 void
 NineBaseTexture9_Dump( struct NineBaseTexture9 *This )
 {
-    DBG("\nNineBaseTexture9(%p->%p/%p): Pool=%s Type=%s Usage=%s\n"
+    DBG("\nNineBaseTexture9(%p->NULL/%p): Pool=%s Type=%s Usage=%s\n"
         "Format=%s Dims=%ux%ux%u/%u LastLevel=%u Lod=%u(%u)\n", This,
-        This->base.resource, This->base.data,
+        This->base.resource,
         nine_D3DPOOL_to_str(This->base.pool),
         nine_D3DRTYPE_to_str(This->base.type),
         nine_D3DUSAGE_to_str(This->base.usage),

src/gallium/state_trackers/nine/basetexture9.h

@@ -59,8 +59,11 @@ NineBaseTexture9( void *data )
 HRESULT
 NineBaseTexture9_ctor( struct NineBaseTexture9 *This,
                        struct NineUnknownParams *pParams,
+                       struct pipe_resource *initResource,
                        D3DRESOURCETYPE Type,
-                       D3DPOOL Pool );
+                       D3DFORMAT format,
+                       D3DPOOL Pool,
+                       DWORD Usage);
 
 void
 NineBaseTexture9_dtor( struct NineBaseTexture9 *This );

src/gallium/state_trackers/nine/cubetexture9.c

@@ -42,6 +42,11 @@ NineCubeTexture9_ctor( struct NineCubeTexture9 *This,
     D3DSURFACE_DESC sfdesc;
     HRESULT hr;
 
+    DBG("This=%p pParams=%p EdgeLength=%u Levels=%u Usage=%d "
+        "Format=%d Pool=%d pSharedHandle=%p\n",
+        This, pParams, EdgeLength, Levels, Usage,
+        Format, Pool, pSharedHandle);
+
     user_assert(!(Usage & D3DUSAGE_AUTOGENMIPMAP) ||
                 (Pool != D3DPOOL_SYSTEMMEM && Levels <= 1), D3DERR_INVALIDCALL);
 
@@ -50,9 +55,6 @@ NineCubeTexture9_ctor( struct NineCubeTexture9 *This,
     if (Usage & D3DUSAGE_AUTOGENMIPMAP)
         Levels = 0;
 
-    This->base.format = Format;
-    This->base.base.usage = Usage;
-
     info->screen = pParams->device->screen;
     info->target = PIPE_TEXTURE_CUBE;
     info->format = d3d9_to_pipe_format(Format);
@@ -85,8 +87,8 @@ NineCubeTexture9_ctor( struct NineCubeTexture9 *This,
     if (!This->surfaces)
         return E_OUTOFMEMORY;
 
-    hr = NineBaseTexture9_ctor(&This->base, pParams, D3DRTYPE_CUBETEXTURE,
-                               Pool);
+    hr = NineBaseTexture9_ctor(&This->base, pParams, NULL, D3DRTYPE_CUBETEXTURE,
+                               Format, Pool, Usage);
     if (FAILED(hr))
         return hr;
     This->base.pstype = 2;
@@ -105,7 +107,7 @@ NineCubeTexture9_ctor( struct NineCubeTexture9 *This,
         sfdesc.Width = sfdesc.Height = u_minify(EdgeLength, i / 6);
 
         hr = NineSurface9_new(This->base.base.base.device, NineUnknown(This),
-                              This->base.base.resource, D3DRTYPE_CUBETEXTURE,
+                              This->base.base.resource, NULL, D3DRTYPE_CUBETEXTURE,
                               i / 6, i % 6,
                               &sfdesc, &This->surfaces[i]);
         if (FAILED(hr))
@@ -138,6 +140,8 @@ NineCubeTexture9_GetLevelDesc( struct NineCubeTexture9 *This,
                                UINT Level,
                                D3DSURFACE_DESC *pDesc )
 {
+    DBG("This=%p Level=%u pDesc=%p\n", This, Level, pDesc);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(Level == 0 || !(This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP),
                 D3DERR_INVALIDCALL);
@@ -155,6 +159,9 @@ NineCubeTexture9_GetCubeMapSurface( struct NineCubeTexture9 *This,
 {
     const unsigned s = Level * 6 + FaceType;
 
+    DBG("This=%p FaceType=%d Level=%u ppCubeMapSurface=%p\n",
+        This, FaceType, Level, ppCubeMapSurface);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(Level == 0 || !(This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP),
                 D3DERR_INVALIDCALL);
@@ -176,6 +183,9 @@ NineCubeTexture9_LockRect( struct NineCubeTexture9 *This,
 {
     const unsigned s = Level * 6 + FaceType;
 
+    DBG("This=%p FaceType=%d Level=%u pLockedRect=%p pRect=%p Flags=%d\n",
+        This, FaceType, Level, pLockedRect, pRect, Flags);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(Level == 0 || !(This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP),
                 D3DERR_INVALIDCALL);
@@ -191,6 +201,8 @@ NineCubeTexture9_UnlockRect( struct NineCubeTexture9 *This,
 {
     const unsigned s = Level * 6 + FaceType;
 
+    DBG("This=%p FaceType=%d Level=%u\n", This, FaceType, Level);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(FaceType < 6, D3DERR_INVALIDCALL);
 
@@ -202,6 +214,8 @@ NineCubeTexture9_AddDirtyRect( struct NineCubeTexture9 *This,
                                D3DCUBEMAP_FACES FaceType,
                                const RECT *pDirtyRect )
 {
+    DBG("This=%p FaceType=%d pDirtyRect=%p\n", This, FaceType, pDirtyRect);
+
     user_assert(FaceType < 6, D3DERR_INVALIDCALL);
 
     if (This->base.base.pool != D3DPOOL_MANAGED) {

src/gallium/state_trackers/nine/device9.c

@@ -58,6 +58,8 @@ NineDevice9_SetDefaultState( struct NineDevice9 *This, boolean is_reset )
 {
     struct NineSurface9 *refSurf = NULL;
 
+    DBG("This=%p is_reset=%d\n", This, (int) is_reset);
+
     assert(!This->is_recording);
 
     nine_state_set_defaults(&This->state, &This->caps, is_reset);
@@ -94,6 +96,8 @@ NineDevice9_RestoreNonCSOState( struct NineDevice9 *This, unsigned mask )
 {
     struct pipe_context *pipe = This->pipe;
 
+    DBG("This=%p mask=%u\n", This, mask);
+
     if (mask & 0x1) {
         struct pipe_constant_buffer cb;
         cb.buffer_offset = 0;
@@ -139,10 +143,18 @@ NineDevice9_ctor( struct NineDevice9 *This,
                   D3DPRESENT_PARAMETERS *pPresentationParameters,
                   IDirect3D9 *pD3D9,
                   ID3DPresentGroup *pPresentationGroup,
-                  struct d3dadapter9_context *pCTX )
+                  struct d3dadapter9_context *pCTX,
+                  boolean ex,
+                  D3DDISPLAYMODEEX *pFullscreenDisplayMode )
 {
     unsigned i;
     HRESULT hr = NineUnknown_ctor(&This->base, pParams);
+
+    DBG("This=%p pParams=%p pScreen=%p pCreationParameters=%p pCaps=%p pPresentationParameters=%p "
+        "pD3D9=%p pPresentationGroup=%p pCTX=%p ex=%d pFullscreenDisplayMode=%p\n",
+        This, pParams, pScreen, pCreationParameters, pCaps, pPresentationParameters, pD3D9,
+        pPresentationGroup, pCTX, (int) ex, pFullscreenDisplayMode);
+
     if (FAILED(hr)) { return hr; }
 
     list_inithead(&This->update_textures);
@@ -151,6 +163,7 @@ NineDevice9_ctor( struct NineDevice9 *This,
     This->caps = *pCaps;
     This->d3d9 = pD3D9;
     This->params = *pCreationParameters;
+    This->ex = ex;
     This->present = pPresentationGroup;
     IDirect3D9_AddRef(This->d3d9);
     ID3DPresentGroup_AddRef(This->present);
@@ -177,12 +190,12 @@ NineDevice9_ctor( struct NineDevice9 *This,
         if (FAILED(hr))
             return hr;
 
-        if (This->ex) {
+        if (ex) {
             D3DDISPLAYMODEEX *mode = NULL;
             struct NineSwapChain9Ex **ret =
                 (struct NineSwapChain9Ex **)&This->swapchains[i];
 
-            if (This->pFullscreenDisplayMode) mode = &(This->pFullscreenDisplayMode[i]);
+            if (pFullscreenDisplayMode) mode = &(pFullscreenDisplayMode[i]);
             /* when this is a Device9Ex, it should create SwapChain9Exs */
             hr = NineSwapChain9Ex_new(This, TRUE, present,
                                       &pPresentationParameters[i], pCTX,
@@ -310,9 +323,6 @@ NineDevice9_ctor( struct NineDevice9 *This,
     This->update = &This->state;
     nine_update_state(This, ~0);
 
-    /* Is just used to pass the parameter from NineDevice9Ex_ctor */
-    This->pFullscreenDisplayMode = NULL;
-
     ID3DPresentGroup_Release(This->present);
 
     return D3D_OK;
@@ -413,7 +423,11 @@ NineDevice9_TestCooperativeLevel( struct NineDevice9 *This )
 UINT WINAPI
 NineDevice9_GetAvailableTextureMem( struct NineDevice9 *This )
 {
-    return This->screen->get_param(This->screen, PIPE_CAP_VIDEO_MEMORY);
+   const unsigned mem = This->screen->get_param(This->screen, PIPE_CAP_VIDEO_MEMORY);
+   if (mem < 4096)
+      return mem << 20;
+   else
+      return UINT_MAX;
 }
 
 HRESULT WINAPI
@@ -536,6 +550,8 @@ NineDevice9_SetCursorPosition( struct NineDevice9 *This,
 {
     struct NineSwapChain9 *swap = This->swapchains[0];
 
+    DBG("This=%p X=%d Y=%d Flags=%d\n", This, X, Y, Flags);
+
     This->cursor.pos.x = X;
     This->cursor.pos.y = Y;
 
@@ -548,6 +564,9 @@ NineDevice9_ShowCursor( struct NineDevice9 *This,
                         BOOL bShow )
 {
     BOOL old = This->cursor.visible;
+
+    DBG("This=%p bShow=%d\n", This, (int) bShow);
+
     This->cursor.visible = bShow && (This->cursor.hotspot.x != -1);
     if (!This->cursor.software)
         ID3DPresent_SetCursor(This->swapchains[0]->present, NULL, NULL, bShow);
@@ -564,6 +583,9 @@ NineDevice9_CreateAdditionalSwapChain( struct NineDevice9 *This,
     ID3DPresent *present;
     HRESULT hr;
 
+    DBG("This=%p pPresentationParameters=%p pSwapChain=%p\n",
+        This, pPresentationParameters, pSwapChain);
+
     user_assert(pPresentationParameters, D3DERR_INVALIDCALL);
 
     hr = ID3DPresentGroup_CreateAdditionalPresent(This->present, pPresentationParameters, &present);
@@ -641,6 +663,9 @@ NineDevice9_Present( struct NineDevice9 *This,
     unsigned i;
     HRESULT hr;
 
+    DBG("This=%p pSourceRect=%p pDestRect=%p hDestWindowOverride=%p pDirtyRegion=%p\n",
+        This, pSourceRect, pDestRect, hDestWindowOverride, pDirtyRegion);
+
     /* XXX is this right? */
     for (i = 0; i < This->nswapchains; ++i) {
         hr = NineSwapChain9_Present(This->swapchains[i], pSourceRect, pDestRect,
@@ -974,7 +999,7 @@ create_zs_or_rt_surface(struct NineDevice9 *This,
     } else {
         resource = NULL;
     }
-    hr = NineSurface9_new(This, NULL, resource, 0, 0, 0, &desc, &surface);
+    hr = NineSurface9_new(This, NULL, resource, NULL, 0, 0, 0, &desc, &surface);
     pipe_resource_reference(&resource, NULL);
 
     if (SUCCEEDED(hr))
@@ -1710,6 +1735,9 @@ NineDevice9_SetTransform( struct NineDevice9 *This,
 {
     struct nine_state *state = This->update;
     D3DMATRIX *M = nine_state_access_transform(state, State, TRUE);
+
+    DBG("This=%p State=%d pMatrix=%p\n", This, State, pMatrix);
+
     user_assert(M, D3DERR_INVALIDCALL);
 
     *M = *pMatrix;
@@ -1738,6 +1766,9 @@ NineDevice9_MultiplyTransform( struct NineDevice9 *This,
     struct nine_state *state = This->update;
     D3DMATRIX T;
     D3DMATRIX *M = nine_state_access_transform(state, State, TRUE);
+
+    DBG("This=%p State=%d pMatrix=%p\n", This, State, pMatrix);
+
     user_assert(M, D3DERR_INVALIDCALL);
 
     nine_d3d_matrix_matrix_mul(&T, pMatrix, M);
@@ -2172,6 +2203,7 @@ NineDevice9_SetTexture( struct NineDevice9 *This,
                         IDirect3DBaseTexture9 *pTexture )
 {
     struct nine_state *state = This->update;
+    struct NineBaseTexture9 *tex = NineBaseTexture9(pTexture);
 
     DBG("This=%p Stage=%u pTexture=%p\n", This, Stage, pTexture);
 
@@ -2179,12 +2211,19 @@ NineDevice9_SetTexture( struct NineDevice9 *This,
                 Stage == D3DDMAPSAMPLER ||
                 (Stage >= D3DVERTEXTEXTURESAMPLER0 &&
                  Stage <= D3DVERTEXTEXTURESAMPLER3), D3DERR_INVALIDCALL);
+    user_assert(!tex || tex->base.pool != D3DPOOL_SCRATCH, D3DERR_INVALIDCALL);
+
+    if (unlikely(tex && tex->base.pool == D3DPOOL_SYSTEMMEM)) {
+        /* TODO: Currently not implemented. Better return error
+         * with message telling what's wrong */
+        ERR("This=%p D3DPOOL_SYSTEMMEM not implemented for SetTexture\n", This);
+        user_assert(tex->base.pool != D3DPOOL_SYSTEMMEM, D3DERR_INVALIDCALL);
+    }
 
     if (Stage >= D3DDMAPSAMPLER)
         Stage = Stage - D3DDMAPSAMPLER + NINE_MAX_SAMPLERS_PS;
 
     if (!This->is_recording) {
-        struct NineBaseTexture9 *tex = NineBaseTexture9(pTexture);
         struct NineBaseTexture9 *old = state->texture[Stage];
         if (old == tex)
             return D3D_OK;
@@ -2648,6 +2687,11 @@ NineDevice9_ProcessVertices( struct NineDevice9 *This,
     HRESULT hr;
     unsigned buffer_offset, buffer_size;
 
+    DBG("This=%p SrcStartIndex=%u DestIndex=%u VertexCount=%u "
+        "pDestBuffer=%p pVertexDecl=%p Flags=%d\n",
+        This, SrcStartIndex, DestIndex, VertexCount, pDestBuffer,
+        pVertexDecl, Flags);
+
     if (!screen->get_param(screen, PIPE_CAP_MAX_STREAM_OUTPUT_BUFFERS))
         STUB(D3DERR_INVALIDCALL);
 
@@ -2725,6 +2769,9 @@ NineDevice9_CreateVertexDeclaration( struct NineDevice9 *This,
 {
     struct NineVertexDeclaration9 *vdecl;
 
+    DBG("This=%p pVertexElements=%p ppDecl=%p\n",
+        This, pVertexElements, ppDecl);
+
     HRESULT hr = NineVertexDeclaration9_new(This, pVertexElements, &vdecl);
     if (SUCCEEDED(hr))
         *ppDecl = (IDirect3DVertexDeclaration9 *)vdecl;
@@ -2801,6 +2848,8 @@ NineDevice9_CreateVertexShader( struct NineDevice9 *This,
     struct NineVertexShader9 *vs;
     HRESULT hr;
 
+    DBG("This=%p pFunction=%p ppShader=%p\n", This, pFunction, ppShader);
+
     hr = NineVertexShader9_new(This, &vs, pFunction, NULL);
     if (FAILED(hr))
         return hr;
@@ -2981,6 +3030,9 @@ NineDevice9_SetStreamSource( struct NineDevice9 *This,
     struct NineVertexBuffer9 *pVBuf9 = NineVertexBuffer9(pStreamData);
     const unsigned i = StreamNumber;
 
+    DBG("This=%p StreamNumber=%u pStreamData=%p OffsetInBytes=%u Stride=%u\n",
+        This, StreamNumber, pStreamData, OffsetInBytes, Stride);
+
     user_assert(StreamNumber < This->caps.MaxStreams, D3DERR_INVALIDCALL);
     user_assert(Stride <= This->caps.MaxStreamStride, D3DERR_INVALIDCALL);
 
@@ -3068,6 +3120,8 @@ NineDevice9_SetIndices( struct NineDevice9 *This,
 {
     struct nine_state *state = This->update;
 
+    DBG("This=%p pIndexData=%p\n", This, pIndexData);
+
     if (likely(!This->is_recording))
         if (state->idxbuf == NineIndexBuffer9(pIndexData))
             return D3D_OK;
@@ -3099,6 +3153,8 @@ NineDevice9_CreatePixelShader( struct NineDevice9 *This,
     struct NinePixelShader9 *ps;
     HRESULT hr;
 
+    DBG("This=%p pFunction=%p ppShader=%p\n", This, pFunction, ppShader);
+
     hr = NinePixelShader9_new(This, &ps, pFunction, NULL);
     if (FAILED(hr))
         return hr;
@@ -3301,6 +3357,8 @@ NineDevice9_CreateQuery( struct NineDevice9 *This,
     struct NineQuery9 *query;
     HRESULT hr;
 
+    DBG("This=%p Type=%d ppQuery=%p\n", This, Type, ppQuery);
+
     if (!ppQuery)
         return nine_is_query_supported(Type);
 
@@ -3447,6 +3505,8 @@ NineDevice9_new( struct pipe_screen *pScreen,
                  IDirect3D9 *pD3D9,
                  ID3DPresentGroup *pPresentationGroup,
                  struct d3dadapter9_context *pCTX,
+                 boolean ex,
+                 D3DDISPLAYMODEEX *pFullscreenDisplayMode,
                  struct NineDevice9 **ppOut )
 {
     BOOL lock;
@@ -3454,5 +3514,6 @@ NineDevice9_new( struct pipe_screen *pScreen,
 
     NINE_NEW(Device9, ppOut, lock, /* args */
              pScreen, pCreationParameters, pCaps,
-             pPresentationParameters, pD3D9, pPresentationGroup, pCTX);
+             pPresentationParameters, pD3D9, pPresentationGroup, pCTX,
+             ex, pFullscreenDisplayMode);
 }

src/gallium/state_trackers/nine/device9.h

@@ -48,7 +48,6 @@ struct NineDevice9
 {
     struct NineUnknown base;
     boolean ex;
-    D3DDISPLAYMODEEX *pFullscreenDisplayMode;
 
     /* G3D context */
     struct pipe_screen *screen;
@@ -134,6 +133,8 @@ NineDevice9_new( struct pipe_screen *pScreen,
                  IDirect3D9 *pD3D9,
                  ID3DPresentGroup *pPresentationGroup,
                  struct d3dadapter9_context *pCTX,
+                 boolean ex,
+                 D3DDISPLAYMODEEX *pFullscreenDisplayMode,
                  struct NineDevice9 **ppOut );
 
 HRESULT
@@ -145,7 +146,9 @@ NineDevice9_ctor( struct NineDevice9 *This,
                   D3DPRESENT_PARAMETERS *pPresentationParameters,
                   IDirect3D9 *pD3D9,
                   ID3DPresentGroup *pPresentationGroup,
-                  struct d3dadapter9_context *pCTX );
+                  struct d3dadapter9_context *pCTX,
+                  boolean ex,
+                  D3DDISPLAYMODEEX *pFullscreenDisplayMode );
 
 void
 NineDevice9_dtor( struct NineDevice9 *This );

src/gallium/state_trackers/nine/device9ex.c

@@ -39,13 +39,18 @@ NineDevice9Ex_ctor( struct NineDevice9Ex *This,
                     ID3DPresentGroup *pPresentationGroup,
                     struct d3dadapter9_context *pCTX )
 {
-    This->base.ex = TRUE;
-    This->base.pFullscreenDisplayMode = pFullscreenDisplayMode;
+    DBG("This=%p pParams=%p pScreen=%p pCreationParameters=%p pCaps=%p "
+        "pPresentationParameters=%p pFullscreenDisplayMode=%p "
+        "pD3D9Ex=%p pPresentationGroup=%p pCTX=%p\n",
+        This, pParams, pScreen, pCreationParameters, pCaps,
+        pPresentationParameters, pFullscreenDisplayMode,
+        pD3D9Ex, pPresentationGroup, pCTX);
 
     return NineDevice9_ctor(&This->base, pParams,
                             pScreen, pCreationParameters, pCaps,
                             pPresentationParameters,
-                            (IDirect3D9 *)pD3D9Ex, pPresentationGroup, pCTX);
+                            (IDirect3D9 *)pD3D9Ex, pPresentationGroup, pCTX,
+                            TRUE, pFullscreenDisplayMode);
 }
 
 static void
@@ -89,6 +94,11 @@ NineDevice9Ex_PresentEx( struct NineDevice9Ex *This,
     unsigned i;
     HRESULT hr;
 
+    DBG("This=%p pSourceRect=%p pDestRect=%p hDestWindowOverride=%p "
+        "pDirtyRegion=%p dwFlags=%d\n",
+        This, pSourceRect, pDestRect, hDestWindowOverride,
+        pDirtyRegion, dwFlags);
+
     for (i = 0; i < This->base.nswapchains; i++) {
         hr = NineSwapChain9_Present(This->base.swapchains[i], pSourceRect, pDestRect,
                                     hDestWindowOverride, pDirtyRegion, dwFlags);
@@ -229,6 +239,9 @@ NineDevice9Ex_GetDisplayModeEx( struct NineDevice9Ex *This,
 {
     struct NineSwapChain9Ex *swapchain;
 
+    DBG("This=%p iSwapChain=%u pMode=%p pRotation=%p\n",
+        This, iSwapChain, pMode, pRotation);
+
     user_assert(iSwapChain < This->base.nswapchains, D3DERR_INVALIDCALL);
 
     swapchain = NineSwapChain9Ex(This->base.swapchains[iSwapChain]);

src/gallium/state_trackers/nine/indexbuffer9.c

@@ -76,8 +76,8 @@ NineIndexBuffer9_ctor( struct NineIndexBuffer9 *This,
     info->last_level = 0;
     info->nr_samples = 0;
 
-    hr = NineResource9_ctor(&This->base, pParams, TRUE, D3DRTYPE_INDEXBUFFER,
-                            pDesc->Pool);
+    hr = NineResource9_ctor(&This->base, pParams, NULL, TRUE, D3DRTYPE_INDEXBUFFER,
+                            pDesc->Pool, pDesc->Usage);
     if (FAILED(hr))
         return hr;
 

src/gallium/state_trackers/nine/iunknown.c

@@ -58,6 +58,8 @@ NineUnknown_QueryInterface( struct NineUnknown *This,
 {
     unsigned i = 0;
 
+    DBG("This=%p riid=%p ppvObject=%p\n", This, riid, ppvObject);
+
     if (!ppvObject) return E_POINTER;
 
     do {

src/gallium/state_trackers/nine/nine_defines.h

@@ -30,25 +30,27 @@
 #define NINE_RESOURCE_FLAG_DUMMY    (PIPE_RESOURCE_FLAG_ST_PRIV << 2)
 
 /* vertexdeclaration9.c */
-unsigned nine_d3d9_to_nine_declusage(unsigned usage, unsigned index);
+uint16_t nine_d3d9_to_nine_declusage(unsigned usage, unsigned index);
 
-#define NINE_DECLUSAGE_POSITION(i)     ( 0 + (i))
-#define NINE_DECLUSAGE_BLENDWEIGHT(i)  ( 5 + (i))
-#define NINE_DECLUSAGE_BLENDINDICES(i) ( 9 + (i))
-#define NINE_DECLUSAGE_NORMAL(i)       (13 + (i))
-#define NINE_DECLUSAGE_PSIZE            15
-#define NINE_DECLUSAGE_TEXCOORD(i)     (16 + (i))
-#define NINE_DECLUSAGE_TANGENT(i)      (32 + (i))
-#define NINE_DECLUSAGE_BINORMAL(i)     (34 + (i))
-#define NINE_DECLUSAGE_TESSFACTOR       36
-#define NINE_DECLUSAGE_POSITIONT        37
-#define NINE_DECLUSAGE_COLOR(i)        (38 + (i))
-#define NINE_DECLUSAGE_DEPTH            43
-#define NINE_DECLUSAGE_FOG              44
-#define NINE_DECLUSAGE_SAMPLE           45
-#define NINE_DECLUSAGE_NONE             46
-#define NINE_DECLUSAGE_LAST             NINE_DECLUSAGE_NONE
-#define NINE_DECLUSAGE_COUNT           (NINE_DECLUSAGE_LAST + 1)
+#define NINE_DECLUSAGE_POSITION         0
+#define NINE_DECLUSAGE_BLENDWEIGHT      1
+#define NINE_DECLUSAGE_BLENDINDICES     2
+#define NINE_DECLUSAGE_NORMAL           3
+#define NINE_DECLUSAGE_TEXCOORD         4
+#define NINE_DECLUSAGE_TANGENT          5
+#define NINE_DECLUSAGE_BINORMAL         6
+#define NINE_DECLUSAGE_COLOR            7
+#define NINE_DECLUSAGE_POSITIONT        8
+
+#define NINE_DECLUSAGE_PSIZE            9
+#define NINE_DECLUSAGE_TESSFACTOR       10
+#define NINE_DECLUSAGE_DEPTH            11
+#define NINE_DECLUSAGE_FOG              12
+#define NINE_DECLUSAGE_SAMPLE           13
+#define NINE_DECLUSAGE_NONE             14
+#define NINE_DECLUSAGE_COUNT            (NINE_DECLUSAGE_NONE + 1)
+
+#define NINE_DECLUSAGE_i(declusage, n) NINE_DECLUSAGE_##declusage + n * NINE_DECLUSAGE_COUNT
 
 #define NINED3DCLEAR_DEPTHSTENCIL   (D3DCLEAR_ZBUFFER | D3DCLEAR_STENCIL)
 

src/gallium/state_trackers/nine/nine_ff.c

@@ -189,13 +189,10 @@ static void nine_ureg_tgsi_dump(struct ureg_program *ureg, boolean override)
 
 /* AL should contain base address of lights table. */
 #define LIGHT_CONST(i)                                                \
-    ureg_src_indirect(ureg_src_register(TGSI_FILE_CONSTANT, (i)), _X(AL))
+    ureg_src_indirect(ureg_DECL_constant(ureg, i), _X(AL))
 
 #define MATERIAL_CONST(i) \
-    ureg_src_register(TGSI_FILE_CONSTANT, 19 + (i))
-
-#define MISC_CONST(i) \
-    ureg_src_register(TGSI_FILE_CONSTANT, (i))
+    ureg_DECL_constant(ureg, 19 + (i))
 
 #define _CONST(n) ureg_DECL_constant(ureg, n)
 
@@ -278,7 +275,7 @@ struct vs_build_ctx
     struct ureg_program *ureg;
     const struct nine_ff_vs_key *key;
 
-    unsigned input[PIPE_MAX_ATTRIBS];
+    uint16_t input[PIPE_MAX_ATTRIBS];
     unsigned num_inputs;
 
     struct ureg_src aVtx;
@@ -307,7 +304,7 @@ get_texcoord_sn(struct pipe_screen *screen)
 }
 
 static INLINE struct ureg_src
-build_vs_add_input(struct vs_build_ctx *vs, unsigned ndecl)
+build_vs_add_input(struct vs_build_ctx *vs, uint16_t ndecl)
 {
     const unsigned i = vs->num_inputs++;
     assert(i < PIPE_MAX_ATTRIBS);
@@ -373,10 +370,10 @@ nine_ff_build_vs(struct NineDevice9 *device, struct vs_build_ctx *vs)
      * (texture coordinates handled later)
      */
     vs->aVtx = build_vs_add_input(vs,
-        key->position_t ? NINE_DECLUSAGE_POSITIONT : NINE_DECLUSAGE_POSITION(0));
+        key->position_t ? NINE_DECLUSAGE_POSITIONT : NINE_DECLUSAGE_POSITION);
 
     if (need_rNrm)
-        vs->aNrm = build_vs_add_input(vs, NINE_DECLUSAGE_NORMAL(0));
+        vs->aNrm = build_vs_add_input(vs, NINE_DECLUSAGE_NORMAL);
 
     vs->aCol[0] = ureg_imm1f(ureg, 1.0f);
     vs->aCol[1] = ureg_imm1f(ureg, 1.0f);
@@ -385,9 +382,9 @@ nine_ff_build_vs(struct NineDevice9 *device, struct vs_build_ctx *vs)
         const unsigned mask = key->mtl_diffuse | key->mtl_specular |
                               key->mtl_ambient | key->mtl_emissive;
         if ((mask & 0x1) && !key->color0in_one)
-            vs->aCol[0] = build_vs_add_input(vs, NINE_DECLUSAGE_COLOR(0));
+            vs->aCol[0] = build_vs_add_input(vs, NINE_DECLUSAGE_i(COLOR, 0));
         if ((mask & 0x2) && !key->color1in_one)
-            vs->aCol[1] = build_vs_add_input(vs, NINE_DECLUSAGE_COLOR(1));
+            vs->aCol[1] = build_vs_add_input(vs, NINE_DECLUSAGE_i(COLOR, 1));
 
         vs->mtlD = MATERIAL_CONST(1);
         vs->mtlA = MATERIAL_CONST(2);
@@ -402,20 +399,20 @@ nine_ff_build_vs(struct NineDevice9 *device, struct vs_build_ctx *vs)
         if (key->mtl_emissive == 1) vs->mtlE = vs->aCol[0]; else
         if (key->mtl_emissive == 2) vs->mtlE = vs->aCol[1];
     } else {
-        if (!key->color0in_one) vs->aCol[0] = build_vs_add_input(vs, NINE_DECLUSAGE_COLOR(0));
-        if (!key->color1in_one) vs->aCol[1] = build_vs_add_input(vs, NINE_DECLUSAGE_COLOR(1));
+        if (!key->color0in_one) vs->aCol[0] = build_vs_add_input(vs, NINE_DECLUSAGE_i(COLOR, 0));
+        if (!key->color1in_one) vs->aCol[1] = build_vs_add_input(vs, NINE_DECLUSAGE_i(COLOR, 1));
     }
 
     if (key->vertexpointsize)
         vs->aPsz = build_vs_add_input(vs, NINE_DECLUSAGE_PSIZE);
 
     if (key->vertexblend_indexed)
-        vs->aInd = build_vs_add_input(vs, NINE_DECLUSAGE_BLENDINDICES(0));
+        vs->aInd = build_vs_add_input(vs, NINE_DECLUSAGE_BLENDINDICES);
     if (key->vertexblend)
-        vs->aWgt = build_vs_add_input(vs, NINE_DECLUSAGE_BLENDWEIGHT(0));
+        vs->aWgt = build_vs_add_input(vs, NINE_DECLUSAGE_BLENDWEIGHT);
     if (key->vertextween) {
-        vs->aVtx1 = build_vs_add_input(vs, NINE_DECLUSAGE_POSITION(1));
-        vs->aNrm1 = build_vs_add_input(vs, NINE_DECLUSAGE_NORMAL(1));
+        vs->aVtx1 = build_vs_add_input(vs, NINE_DECLUSAGE_i(POSITION,1));
+        vs->aNrm1 = build_vs_add_input(vs, NINE_DECLUSAGE_i(NORMAL,1));
     }
 
     /* Declare outputs:
@@ -599,7 +596,7 @@ nine_ff_build_vs(struct NineDevice9 *device, struct vs_build_ctx *vs)
         oTex[i] = ureg_DECL_output(ureg, texcoord_sn, i);
 
         if (tci == NINED3DTSS_TCI_PASSTHRU)
-            vs->aTex[idx] = build_vs_add_input(vs, NINE_DECLUSAGE_TEXCOORD(idx));
+            vs->aTex[idx] = build_vs_add_input(vs, NINE_DECLUSAGE_i(TEXCOORD,idx));
 
         if (!dim) {
             dst[c = 4] = oTex[i];
@@ -1377,7 +1374,7 @@ nine_ff_get_vs(struct NineDevice9 *device)
     enum pipe_error err;
     struct vs_build_ctx bld;
     struct nine_ff_vs_key key;
-    unsigned s;
+    unsigned s, i;
 
     assert(sizeof(key) <= sizeof(key.value32));
 
@@ -1388,14 +1385,19 @@ nine_ff_get_vs(struct NineDevice9 *device)
 
     /* FIXME: this shouldn't be NULL, but it is on init */
     if (state->vdecl) {
-        if (state->vdecl->usage_map[NINE_DECLUSAGE_POSITIONT] != 0xff)
-            key.position_t = 1;
-        if (state->vdecl->usage_map[NINE_DECLUSAGE_COLOR(0)] == 0xff)
-            key.color0in_one = 1;
-        if (state->vdecl->usage_map[NINE_DECLUSAGE_COLOR(1)] == 0xff)
-            key.color1in_one = 1;
-        if (state->vdecl->usage_map[NINE_DECLUSAGE_PSIZE] != 0xff)
-            key.vertexpointsize = 1;
+        key.color0in_one = 1;
+        key.color1in_one = 1;
+        for (i = 0; i < state->vdecl->nelems; i++) {
+            uint16_t usage = state->vdecl->usage_map[i];
+            if (usage == NINE_DECLUSAGE_POSITIONT)
+                key.position_t = 1;
+            else if (usage == NINE_DECLUSAGE_i(COLOR, 0))
+                key.color0in_one = 0;
+            else if (usage == NINE_DECLUSAGE_i(COLOR, 1))
+                key.color1in_one = 0;
+            else if (usage == NINE_DECLUSAGE_PSIZE)
+                key.vertexpointsize = 1;
+        }
     }
     if (!key.vertexpointsize)
         key.pointscale = !!state->rs[D3DRS_POINTSCALEENABLE];

src/gallium/state_trackers/nine/nine_shader.c

@@ -239,7 +239,7 @@ struct sm1_dst_param
     BYTE file;
     BYTE mask;
     BYTE mod;
-    BYTE shift; /* sint4 */
+    int8_t shift; /* sint4 */
     BYTE type;
 };
 
@@ -1685,101 +1685,93 @@ sm1_declusage_to_tgsi(struct tgsi_declaration_semantic *sem,
                       boolean tc,
                       struct sm1_semantic *dcl)
 {
-    const unsigned generic_base = tc ? 0 : 8; /* TEXCOORD[0..7] */
+    BYTE index = dcl->usage_idx;
 
-    sem->Name = TGSI_SEMANTIC_GENERIC;
-    sem->Index = 0;
-
-    /* TGSI_SEMANTIC_GENERIC assignments (+8 if !PIPE_CAP_TGSI_TEXCOORD):
-     * Try to put frequently used semantics at low GENERIC indices.
+    /* For everything that is not matching to a TGSI_SEMANTIC_****,
+     * we match to a TGSI_SEMANTIC_GENERIC with index.
      *
-     * POSITION[1..4]: 17, 27, 28, 29
-     * COLOR[2..4]: 14, 15, 26
-     * TEXCOORD[8..15]: 10, 11, 18, 19, 20, 21, 22, 23
-     * BLENDWEIGHT[0..3]: 0, 4, 8, 12
-     * BLENDINDICES[0..3]: 1, 5, 9, 13
-     * NORMAL[0..1]: 2, 6
-     * TANGENT[0]: 3, 24
-     * BINORMAL[0]: 7, 25
-     * TESSFACTOR[0]: 16
+     * The index can be anything UINT16 and usage_idx is BYTE,
+     * so we can fit everything. It doesn't matter if indices
+     * are close together or low.
+     *
+     *
+     * POSITION >= 1: 10 * index + 6
+     * COLOR >= 2: 10 * (index-1) + 7
+     * TEXCOORD[0..15]: index
+     * BLENDWEIGHT: 10 * index + 18
+     * BLENDINDICES: 10 * index + 19
+     * NORMAL: 10 * index + 20
+     * TANGENT: 10 * index + 21
+     * BINORMAL: 10 * index + 22
+     * TESSFACTOR: 10 * index + 23
      */
 
     switch (dcl->usage) {
     case D3DDECLUSAGE_POSITION:
     case D3DDECLUSAGE_POSITIONT:
     case D3DDECLUSAGE_DEPTH:
-        sem->Name = TGSI_SEMANTIC_POSITION;
-        assert(dcl->usage_idx <= 4);
-        if (dcl->usage_idx == 1) {
+        if (index == 0) {
+            sem->Name = TGSI_SEMANTIC_POSITION;
+            sem->Index = 0;
+        } else {
             sem->Name = TGSI_SEMANTIC_GENERIC;
-            sem->Index = generic_base + 17;
-        } else
-        if (dcl->usage_idx >= 2) {
-            sem->Name = TGSI_SEMANTIC_GENERIC;
-            sem->Index = generic_base + 27 + (dcl->usage_idx - 2);
+            sem->Index = 10 * index + 6;
         }
         break;
     case D3DDECLUSAGE_COLOR:
-        assert(dcl->usage_idx <= 4);
-        if (dcl->usage_idx < 2) {
+        if (index < 2) {
             sem->Name = TGSI_SEMANTIC_COLOR;
-            sem->Index = dcl->usage_idx;
-        } else
-        if (dcl->usage_idx < 4) {
-            sem->Index = generic_base + 14 + (dcl->usage_idx - 2);
+            sem->Index = index;
         } else {
-            sem->Index = generic_base + 26;
+            sem->Name = TGSI_SEMANTIC_GENERIC;
+            sem->Index = 10 * (index-1) + 7;
         }
         break;
     case D3DDECLUSAGE_FOG:
+        assert(index == 0);
         sem->Name = TGSI_SEMANTIC_FOG;
-        assert(dcl->usage_idx == 0);
+        sem->Index = 0;
         break;
     case D3DDECLUSAGE_PSIZE:
+        assert(index == 0);
         sem->Name = TGSI_SEMANTIC_PSIZE;
-        assert(dcl->usage_idx == 0);
+        sem->Index = 0;
         break;
     case D3DDECLUSAGE_TEXCOORD:
-        assert(dcl->usage_idx < 16);
-        if (dcl->usage_idx < 8) {
-            if (tc)
-                sem->Name = TGSI_SEMANTIC_TEXCOORD;
-            sem->Index = dcl->usage_idx;
-        } else
-        if (dcl->usage_idx < 10) {
-            sem->Index = generic_base + 10 + (dcl->usage_idx - 8);
-        } else {
-            sem->Index = generic_base + 18 + (dcl->usage_idx - 10);
-        }
+        assert(index < 16);
+        if (index < 8 && tc)
+            sem->Name = TGSI_SEMANTIC_TEXCOORD;
+        else
+            sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = index;
         break;
-    case D3DDECLUSAGE_BLENDWEIGHT: /* 0, 4, 8, 12 */
-        assert(dcl->usage_idx < 4);
-        sem->Index = generic_base + dcl->usage_idx * 4;
+    case D3DDECLUSAGE_BLENDWEIGHT:
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 18;
         break;
-    case D3DDECLUSAGE_BLENDINDICES: /* 1, 5, 9, 13 */
-        assert(dcl->usage_idx < 4);
-        sem->Index = generic_base + dcl->usage_idx * 4 + 1;
+    case D3DDECLUSAGE_BLENDINDICES:
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 19;
         break;
-    case D3DDECLUSAGE_NORMAL: /* 2, 3 */
-        assert(dcl->usage_idx < 2);
-        sem->Index = generic_base + 2 + dcl->usage_idx * 4;
+    case D3DDECLUSAGE_NORMAL:
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 20;
         break;
     case D3DDECLUSAGE_TANGENT:
-        /* Yes these are weird, but we try to fit the more frequently used
-         * into lower slots. */
-        assert(dcl->usage_idx <= 1);
-        sem->Index = generic_base + (dcl->usage_idx ? 24 : 3);
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 21;
         break;
     case D3DDECLUSAGE_BINORMAL:
-        assert(dcl->usage_idx <= 1);
-        sem->Index = generic_base + (dcl->usage_idx ? 25 : 7);
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 22;
         break;
     case D3DDECLUSAGE_TESSFACTOR:
-        assert(dcl->usage_idx == 0);
-        sem->Index = generic_base + 16;
+        sem->Name = TGSI_SEMANTIC_GENERIC;
+        sem->Index = 10 * index + 23;
         break;
     case D3DDECLUSAGE_SAMPLE:
         sem->Name = TGSI_SEMANTIC_COUNT;
+        sem->Index = 0;
         break;
     default:
         assert(!"Invalid DECLUSAGE.");
@@ -2535,6 +2527,7 @@ sm1_parse_get_param(struct shader_translator *tx, DWORD *reg, DWORD *rel)
 static void
 sm1_parse_dst_param(struct sm1_dst_param *dst, DWORD tok)
 {
+    uint8_t shift;
     dst->file =
         (tok & D3DSP_REGTYPE_MASK)  >> D3DSP_REGTYPE_SHIFT |
         (tok & D3DSP_REGTYPE_MASK2) >> D3DSP_REGTYPE_SHIFT2;
@@ -2543,7 +2536,8 @@ sm1_parse_dst_param(struct sm1_dst_param *dst, DWORD tok)
     dst->rel = NULL;
     dst->mask = (tok & NINED3DSP_WRITEMASK_MASK) >> NINED3DSP_WRITEMASK_SHIFT;
     dst->mod = (tok & D3DSP_DSTMOD_MASK) >> D3DSP_DSTMOD_SHIFT;
-    dst->shift = (tok & D3DSP_DSTSHIFT_MASK) >> D3DSP_DSTSHIFT_SHIFT;
+    shift = (tok & D3DSP_DSTSHIFT_MASK) >> D3DSP_DSTSHIFT_SHIFT;
+    dst->shift = (shift & 0x8) ? -(shift & 0x7) : shift & 0x7;
 }
 
 static void

src/gallium/state_trackers/nine/nine_shader.h

@@ -48,7 +48,7 @@ struct nine_shader_info
 
     void *cso; /* out, pipe cso for bind_vs,fs_state */
 
-    uint8_t input_map[PIPE_MAX_ATTRIBS]; /* VS input -> NINE_DECLUSAGE_x */
+    uint16_t input_map[PIPE_MAX_ATTRIBS]; /* VS input -> NINE_DECLUSAGE_x */
     uint8_t num_inputs; /* there may be unused inputs (NINE_DECLUSAGE_NONE) */
 
     boolean position_t; /* out, true if VP writes pre-transformed position */

src/gallium/state_trackers/nine/nine_state.c

@@ -141,11 +141,9 @@ update_viewport(struct NineDevice9 *device)
     pvport.scale[0] = (float)vport->Width * 0.5f;
     pvport.scale[1] = (float)vport->Height * -0.5f;
     pvport.scale[2] = vport->MaxZ - vport->MinZ;
-    pvport.scale[3] = 1.0f;
     pvport.translate[0] = (float)vport->Width * 0.5f + (float)vport->X;
     pvport.translate[1] = (float)vport->Height * 0.5f + (float)vport->Y;
     pvport.translate[2] = vport->MinZ;
-    pvport.translate[3] = 0.0f;
 
     pipe->set_viewport_states(pipe, 0, 1, &pvport);
 }
@@ -186,7 +184,8 @@ update_vertex_elements(struct NineDevice9 *device)
     struct nine_state *state = &device->state;
     const struct NineVertexDeclaration9 *vdecl = device->state.vdecl;
     const struct NineVertexShader9 *vs;
-    unsigned n, l, b;
+    unsigned n, b, i;
+    int index;
     struct pipe_vertex_element ve[PIPE_MAX_ATTRIBS];
 
     state->stream_usage_mask = 0;
@@ -199,11 +198,16 @@ update_vertex_elements(struct NineDevice9 *device)
         DBG("looking up input %u (usage %u) from vdecl(%p)\n",
             n, vs->input_map[n].ndecl, vdecl);
 
-        assert(vs->input_map[n].ndecl < Elements(vdecl->usage_map));
-        l = vdecl->usage_map[vs->input_map[n].ndecl];
+        index = -1;
+        for (i = 0; i < vdecl->nelems; i++) {
+            if (vdecl->usage_map[i] == vs->input_map[n].ndecl) {
+                index = i;
+                break;
+            }
+        }
 
-        if (likely(l < vdecl->nelems)) {
-            ve[n] = vdecl->elems[l];
+        if (index >= 0) {
+            ve[n] = vdecl->elems[index];
             b = ve[n].vertex_buffer_index;
             state->stream_usage_mask |= 1 << b;
             /* XXX wine just uses 1 here: */
@@ -400,7 +404,7 @@ update_constants(struct NineDevice9 *device, unsigned shader_type)
         const_i = &device->state.ps_const_i[0][0];
 
         dirty_b = device->state.changed.ps_const_b;
-        device->state.changed.vs_const_b = 0;
+        device->state.changed.ps_const_b = 0;
         const_b = device->state.ps_const_b;
         b_true = device->ps_bool_true;
 

src/gallium/state_trackers/nine/pixelshader9.c

@@ -39,6 +39,8 @@ NinePixelShader9_ctor( struct NinePixelShader9 *This,
     struct nine_shader_info info;
     HRESULT hr;
 
+    DBG("This=%p pParams=%p pFunction=%p cso=%p\n", This, pParams, pFunction, cso);
+
     hr = NineUnknown_ctor(&This->base, pParams);
     if (FAILED(hr))
         return hr;
@@ -110,6 +112,8 @@ NinePixelShader9_GetFunction( struct NinePixelShader9 *This,
                               void *pData,
                               UINT *pSizeOfData )
 {
+    DBG("This=%p pData=%p pSizeOfData=%p\n", This, pData, pSizeOfData);
+
     user_assert(pSizeOfData, D3DERR_INVALIDCALL);
 
     if (!pData) {

src/gallium/state_trackers/nine/query9.c

@@ -107,6 +107,8 @@ NineQuery9_ctor( struct NineQuery9 *This,
     const unsigned ptype = d3dquerytype_to_pipe_query(Type);
     HRESULT hr;
 
+    DBG("This=%p pParams=%p Type=%d\n", This, pParams, Type);
+
     hr = NineUnknown_ctor(&This->base, pParams);
     if (FAILED(hr))
         return hr;
@@ -170,6 +172,8 @@ NineQuery9_Issue( struct NineQuery9 *This,
 {
     struct pipe_context *pipe = This->base.device->pipe;
 
+    DBG("This=%p dwIssueFlags=%d\n", This, dwIssueFlags);
+
     user_assert((dwIssueFlags == D3DISSUE_BEGIN && !This->instant) ||
                 (dwIssueFlags == 0) ||
                 (dwIssueFlags == D3DISSUE_END), D3DERR_INVALIDCALL);
@@ -186,10 +190,12 @@ NineQuery9_Issue( struct NineQuery9 *This,
         pipe->begin_query(pipe, This->pq);
         This->state = NINE_QUERY_STATE_RUNNING;
     } else {
-        if (This->state == NINE_QUERY_STATE_RUNNING) {
-            pipe->end_query(pipe, This->pq);
-            This->state = NINE_QUERY_STATE_ENDED;
-	}
+        if (This->state != NINE_QUERY_STATE_RUNNING &&
+            This->type != D3DQUERYTYPE_EVENT &&
+            This->type != D3DQUERYTYPE_TIMESTAMP)
+            pipe->begin_query(pipe, This->pq);
+        pipe->end_query(pipe, This->pq);
+        This->state = NINE_QUERY_STATE_ENDED;
     }
     return D3D_OK;
 }
@@ -221,6 +227,9 @@ NineQuery9_GetData( struct NineQuery9 *This,
     union pipe_query_result presult;
     union nine_query_result nresult;
 
+    DBG("This=%p pData=%p dwSize=%d dwGetDataFlags=%d\n",
+        This, pData, dwSize, dwGetDataFlags);
+
     user_assert(This->state != NINE_QUERY_STATE_RUNNING, D3DERR_INVALIDCALL);
     user_assert(dwSize == 0 || pData, D3DERR_INVALIDCALL);
     user_assert(dwGetDataFlags == 0 ||

src/gallium/state_trackers/nine/resource9.c

@@ -38,29 +38,42 @@
 HRESULT
 NineResource9_ctor( struct NineResource9 *This,
                     struct NineUnknownParams *pParams,
+                    struct pipe_resource *initResource,
                     BOOL Allocate,
                     D3DRESOURCETYPE Type,
-                    D3DPOOL Pool )
+                    D3DPOOL Pool,
+                    DWORD Usage)
 {
     struct pipe_screen *screen;
     HRESULT hr;
 
+    DBG("This=%p pParams=%p initResource=%p Allocate=%d "
+        "Type=%d Pool=%d Usage=%d\n",
+        This, pParams, initResource, (int) Allocate,
+        Type, Pool, Usage);
+
+    DBG("This=%p pParams=%p initResource=%p Allocate=%d Type=%d Pool=%d usage=%d\n",
+        This, pParams, initResource, Allocate, Type, Pool, Usage);
+
     hr = NineUnknown_ctor(&This->base, pParams);
     if (FAILED(hr))
         return hr;
 
     This->info.screen = screen = This->base.device->screen;
+    if (initResource)
+        pipe_resource_reference(&This->resource, initResource);
 
     if (Allocate) {
+        assert(!initResource);
         DBG("(%p) Creating pipe_resource.\n", This);
         This->resource = screen->resource_create(screen, &This->info);
         if (!This->resource)
             return D3DERR_OUTOFVIDEOMEMORY;
     }
 
-    This->data = NULL; /* FIXME remove, rather set it to null in surface9.c*/
     This->type = Type;
     This->pool = Pool;
+    This->usage = Usage;
     This->priority = 0;
 
     This->pdata = util_hash_table_create(ht_guid_hash, ht_guid_compare);
@@ -82,10 +95,6 @@ NineResource9_dtor( struct NineResource9 *This )
      * still hold a reference. */
     pipe_resource_reference(&This->resource, NULL);
 
-    /* release allocated system memory for non-D3DPOOL_DEFAULT resources */
-    if (This->data)
-        FREE(This->data);
-
     NineUnknown_dtor(&This->base);
 }
 
@@ -201,6 +210,9 @@ NineResource9_SetPriority( struct NineResource9 *This,
                            DWORD PriorityNew )
 {
     DWORD prev = This->priority;
+
+    DBG("This=%p, PriorityNew=%d\n", This, PriorityNew);
+
     This->priority = PriorityNew;
     return prev;
 }

src/gallium/state_trackers/nine/resource9.h

@@ -36,8 +36,6 @@ struct NineResource9
 
     struct pipe_resource *resource; /* device resource */
 
-    uint8_t *data; /* system memory backing */
-
     D3DRESOURCETYPE type;
     D3DPOOL pool;
     DWORD priority;
@@ -57,9 +55,11 @@ NineResource9( void *data )
 HRESULT
 NineResource9_ctor( struct NineResource9 *This,
                     struct NineUnknownParams *pParams,
+                    struct pipe_resource *initResource,
                     BOOL Allocate,
                     D3DRESOURCETYPE Type,
-                    D3DPOOL Pool );
+                    D3DPOOL Pool,
+                    DWORD Usage);
 
 void
 NineResource9_dtor( struct NineResource9 *This );

src/gallium/state_trackers/nine/stateblock9.c

@@ -35,6 +35,9 @@ NineStateBlock9_ctor( struct NineStateBlock9 *This,
                       enum nine_stateblock_type type )
 {
     HRESULT hr = NineUnknown_ctor(&This->base, pParams);
+
+    DBG("This=%p pParams=%p type=%d\n", This, pParams, type);
+
     if (FAILED(hr))
         return hr;
 

src/gallium/state_trackers/nine/surface9.c

@@ -43,6 +43,7 @@ NineSurface9_ctor( struct NineSurface9 *This,
                    struct NineUnknownParams *pParams,
                    struct NineUnknown *pContainer,
                    struct pipe_resource *pResource,
+                   void *user_buffer,
                    uint8_t TextureType,
                    unsigned Level,
                    unsigned Layer,
@@ -62,6 +63,8 @@ NineSurface9_ctor( struct NineSurface9 *This,
     assert(pResource ||
            pDesc->Pool != D3DPOOL_DEFAULT || pDesc->Format == D3DFMT_NULL);
 
+    assert(!pResource || !user_buffer);
+
     This->base.info.screen = pParams->device->screen;
     This->base.info.target = PIPE_TEXTURE_2D;
     This->base.info.format = d3d9_to_pipe_format(pDesc->Format);
@@ -82,20 +85,17 @@ NineSurface9_ctor( struct NineSurface9 *This,
 
     if (pDesc->Pool == D3DPOOL_SYSTEMMEM) {
         This->base.info.usage = PIPE_USAGE_STAGING;
-        if (pResource)
-            This->base.data = (uint8_t *)pResource; /* this is *pSharedHandle */
-        pResource = NULL;
+        This->data = (uint8_t *)user_buffer; /* this is *pSharedHandle */
+        assert(!pResource);
     } else {
         if (pResource && (pDesc->Usage & D3DUSAGE_DYNAMIC))
             pResource->flags |= NINE_RESOURCE_FLAG_LOCKABLE;
-        pipe_resource_reference(&This->base.resource, pResource);
     }
 
-    hr = NineResource9_ctor(&This->base, pParams, FALSE, D3DRTYPE_SURFACE,
-                            pDesc->Pool);
+    hr = NineResource9_ctor(&This->base, pParams, pResource, FALSE, D3DRTYPE_SURFACE,
+                            pDesc->Pool, pDesc->Usage);
     if (FAILED(hr))
         return hr;
-    This->base.usage = pDesc->Usage;
 
     This->pipe = This->base.base.device->pipe;
     This->transfer = NULL;
@@ -109,13 +109,20 @@ NineSurface9_ctor( struct NineSurface9 *This,
     This->stride = util_format_get_stride(This->base.info.format, pDesc->Width);
     This->stride = align(This->stride, 4);
 
-    if (!pResource && !This->base.data) {
-        hr = NineSurface9_AllocateData(This);
-        if (FAILED(hr))
-            return hr;
+    if (!pResource && !This->data) {
+        const unsigned size = This->stride *
+            util_format_get_nblocksy(This->base.info.format, This->desc.Height);
+
+        DBG("(%p(This=%p),level=%u) Allocating 0x%x bytes of system memory.\n",
+            This->base.base.container, This, This->level, size);
+
+        This->data = (uint8_t *)MALLOC(size);
+        if (!This->data)
+            return E_OUTOFMEMORY;
+        This->manage_data = TRUE;
     } else {
         if (pResource && NineSurface9_IsOffscreenPlain(This))
-            pResource->flags |= NINE_RESOURCE_FLAG_LOCKABLE;
+            pResource->flags |= NINE_RESOURCE_FLAG_LOCKABLE; /* why setting this flag there ? too late ? should be before NineResource9_ctor call perhaps ? */
     }
 
     NineSurface9_Dump(This);
@@ -133,6 +140,9 @@ NineSurface9_dtor( struct NineSurface9 *This )
     pipe_surface_reference(&This->surface[0], NULL);
     pipe_surface_reference(&This->surface[1], NULL);
 
+    /* release allocated system memory for non-D3DPOOL_DEFAULT resources */
+    if (This->manage_data && This->data)
+        FREE(This->data);
     NineResource9_dtor(&This->base);
 }
 
@@ -167,7 +177,7 @@ NineSurface9_Dump( struct NineSurface9 *This )
 
     DBG("\nNineSurface9(%p->%p/%p): Pool=%s Type=%s Usage=%s\n"
         "Dims=%ux%u Format=%s Stride=%u Lockable=%i\n"
-        "Level=%u(%u), Layer=%u\n", This, This->base.resource, This->base.data,
+        "Level=%u(%u), Layer=%u\n", This, This->base.resource, This->data,
         nine_D3DPOOL_to_str(This->desc.Pool),
         nine_D3DRTYPE_to_str(This->desc.Type),
         nine_D3DUSAGE_to_str(This->desc.Usage),
@@ -240,6 +250,8 @@ NineSurface9_AddDirtyRect( struct NineSurface9 *This,
     float area[2];
     struct u_rect rect, cover_a, cover_b;
 
+    DBG("This=%p box=%p\n", This, box);
+
     if (!box) {
         This->dirty_rects[0].x0 = 0;
         This->dirty_rects[0].y0 = 0;
@@ -286,8 +298,8 @@ NineSurface9_GetSystemMemPointer(struct NineSurface9 *This, int x, int y)
 
     y = util_format_get_nblocksy(This->base.info.format, y);
 
-    assert(This->base.data);
-    return This->base.data + (y * This->stride + x_offset);
+    assert(This->data);
+    return This->data + (y * This->stride + x_offset);
 }
 
 HRESULT WINAPI
@@ -360,7 +372,7 @@ NineSurface9_LockRect( struct NineSurface9 *This,
 
     user_warn(This->desc.Format == D3DFMT_NULL);
 
-    if (This->base.data) {
+    if (This->data) {
         DBG("returning system memory\n");
 
         pLockedRect->Pitch = This->stride;
@@ -419,73 +431,6 @@ NineSurface9_ReleaseDC( struct NineSurface9 *This,
     STUB(D3DERR_INVALIDCALL);
 }
 
-/* nine private */
-
-HRESULT
-NineSurface9_AllocateData( struct NineSurface9 *This )
-{
-#if 0
-    struct pipe_screen *screen = This->base.info.screen;
-    /* XXX: Can't use staging resource because apparently apps expect
-     * memory offsets to be the same across locks.
-     * NV50 doesn't support direct mapping yet so only enable this if
-     * everything else works.
-     */
-    if (This->base.pool == D3DPOOL_SYSTEMMEM) {
-        /* Allocate a staging resource to save a copy:
-         * user -> staging resource
-         * staging resource -> (blit) -> video memory
-         *
-         * Instead of:
-         * user -> system memory
-         * system memory -> transfer staging area
-         * transfer -> video memory
-         *
-         * Does this work if we "lose" the device ?
-         */
-        struct pipe_resource *resource;
-        struct pipe_resource templ;
-
-        templ.target = PIPE_TEXTURE_2D;
-        templ.format = This->base.info.format;
-        templ.width0 = This->desc.Width;
-        templ.height0 = This->desc.Height;
-        templ.depth0 = 1;
-        templ.array_size = 1;
-        templ.last_level = 0;
-        templ.nr_samples = 0;
-        templ.usage = PIPE_USAGE_STAGING;
-        templ.bind =
-            PIPE_BIND_SAMPLER_VIEW |
-            PIPE_BIND_TRANSFER_WRITE |
-            PIPE_BIND_TRANSFER_READ;
-        templ.flags = 0;
-
-        DBG("(%p(This=%p),level=%u) Allocating staging resource.\n",
-            This->base.base.container, This, This->level);
-
-        resource = screen->resource_create(screen, &templ);
-        if (!resource)
-            DBG("Failed to allocate staging resource.\n");
-
-        /* Also deallocate old staging resource. */
-        pipe_resource_reference(&This->base.resource, resource);
-    }
-#endif
-    if (!This->base.resource) {
-        const unsigned size = This->stride *
-            util_format_get_nblocksy(This->base.info.format, This->desc.Height);
-
-        DBG("(%p(This=%p),level=%u) Allocating 0x%x bytes of system memory.\n",
-            This->base.base.container, This, This->level, size);
-
-        This->base.data = (uint8_t *)MALLOC(size);
-        if (!This->base.data)
-            return E_OUTOFMEMORY;
-    }
-    return D3D_OK;
-}
-
 IDirect3DSurface9Vtbl NineSurface9_vtable = {
     (void *)NineUnknown_QueryInterface,
     (void *)NineUnknown_AddRef,
@@ -528,6 +473,9 @@ NineSurface9_CopySurface( struct NineSurface9 *This,
     uint8_t *p_dst;
     const uint8_t *p_src;
 
+    DBG("This=%p From=%p pDestPoint=%p pSourceRect=%p\n",
+        This, From, pDestPoint, pSourceRect);
+
     user_assert(This->desc.Format == From->desc.Format, D3DERR_INVALIDCALL);
 
     dst_box.x = pDestPoint ? pDestPoint->x : 0;
@@ -643,6 +591,8 @@ NineSurface9_UploadSelf( struct NineSurface9 *This )
     uint8_t *ptr;
     unsigned i;
 
+    DBG("This=%p\n", This);
+
     assert(This->base.pool == D3DPOOL_MANAGED);
 
     if (!NineSurface9_IsDirty(This))
@@ -699,6 +649,7 @@ HRESULT
 NineSurface9_new( struct NineDevice9 *pDevice,
                   struct NineUnknown *pContainer,
                   struct pipe_resource *pResource,
+                  void *user_buffer,
                   uint8_t TextureType,
                   unsigned Level,
                   unsigned Layer,
@@ -706,6 +657,6 @@ NineSurface9_new( struct NineDevice9 *pDevice,
                   struct NineSurface9 **ppOut )
 {
     NINE_DEVICE_CHILD_NEW(Surface9, ppOut, pDevice, /* args */
-                          pContainer, pResource,
+                          pContainer, pResource, user_buffer,
                           TextureType, Level, Layer, pDesc);
 }

src/gallium/state_trackers/nine/surface9.h

@@ -47,6 +47,8 @@ struct NineSurface9
     unsigned layer;
     D3DSURFACE_DESC desc;
 
+    uint8_t *data; /* system memory backing */
+    boolean manage_data;
     unsigned stride; /* for system memory backing */
 
     /* wine doesn't even use these, 2 will be enough */
@@ -62,6 +64,7 @@ HRESULT
 NineSurface9_new( struct NineDevice9 *pDevice,
                   struct NineUnknown *pContainer,
                   struct pipe_resource *pResource,
+                  void *user_buffer,
                   uint8_t TextureType, /* 0 if pContainer isn't BaseTexure9 */
                   unsigned Level,
                   unsigned Layer,
@@ -73,6 +76,7 @@ NineSurface9_ctor( struct NineSurface9 *This,
                    struct NineUnknownParams *pParams,
                    struct NineUnknown *pContainer,
                    struct pipe_resource *pResource,
+                   void *user_buffer,
                    uint8_t TextureType,
                    unsigned Level,
                    unsigned Layer,
@@ -124,9 +128,6 @@ NineSurface9_ClearDirtyRects( struct NineSurface9 *This )
     memset(&This->dirty_rects, 0, sizeof(This->dirty_rects));
 }
 
-HRESULT
-NineSurface9_AllocateData( struct NineSurface9 *This );
-
 HRESULT
 NineSurface9_UploadSelf( struct NineSurface9 *This );
 

src/gallium/state_trackers/nine/swapchain9.c

@@ -304,7 +304,7 @@ NineSwapChain9_Resize( struct NineSwapChain9 *This,
         } else {
             desc.Format = pParams->BackBufferFormat;
             desc.Usage = D3DUSAGE_RENDERTARGET;
-            hr = NineSurface9_new(pDevice, NineUnknown(This), resource, 0,
+            hr = NineSurface9_new(pDevice, NineUnknown(This), resource, NULL, 0,
                                   0, 0, &desc, &This->buffers[i]);
             if (has_present_buffers)
                 pipe_resource_reference(&resource, NULL);
@@ -347,7 +347,7 @@ NineSwapChain9_Resize( struct NineSwapChain9 *This,
             /* XXX wine thinks the container of this should be the device */
             desc.Format = pParams->AutoDepthStencilFormat;
             desc.Usage = D3DUSAGE_DEPTHSTENCIL;
-            hr = NineSurface9_new(pDevice, NineUnknown(pDevice), resource, 0,
+            hr = NineSurface9_new(pDevice, NineUnknown(pDevice), resource, NULL, 0,
                                   0, 0, &desc, &This->zsbuf);
             pipe_resource_reference(&resource, NULL);
             if (FAILED(hr)) {
@@ -737,6 +737,11 @@ NineSwapChain9_Present( struct NineSwapChain9 *This,
     HRESULT hr = present(This, pSourceRect, pDestRect,
                          hDestWindowOverride, pDirtyRegion, dwFlags);
 
+    DBG("This=%p pSourceRect=%p pDestRect=%p hDestWindowOverride=%p "
+        "pDirtyRegion=%p dwFlags=%d\n",
+        This, pSourceRect, pDestRect, hDestWindowOverride,
+        pDirtyRegion,dwFlags);
+
     if (hr == D3DERR_WASSTILLDRAWING)
         return hr;
 
@@ -832,7 +837,7 @@ NineSwapChain9_GetFrontBufferData( struct NineSwapChain9 *This,
     /* NineSurface9_CopySurface needs same format. */
     desc.Format = dest_surface->desc.Format;
     desc.Usage = D3DUSAGE_RENDERTARGET;
-    hr = NineSurface9_new(pDevice, NineUnknown(This), temp_resource, 0,
+    hr = NineSurface9_new(pDevice, NineUnknown(This), temp_resource, NULL, 0,
                           0, 0, &desc, &temp_surface);
     pipe_resource_reference(&temp_resource, NULL);
     if (FAILED(hr)) {

src/gallium/state_trackers/nine/swapchain9ex.c

@@ -37,6 +37,10 @@ NineSwapChain9Ex_ctor( struct NineSwapChain9Ex *This,
                        HWND hFocusWindow,
                        D3DDISPLAYMODEEX *mode )
 {
+    DBG("This=%p pParams=%p implicit=%d pPresent=%p pPresentationParameters=%p "
+        "pCTX=%p hFocusWindow=%p mode=%p",
+        This, pParams, (int) implicit, pPresent, pPresentationParameters, pCTX, hFocusWindow, mode);
+
     return NineSwapChain9_ctor(&This->base, pParams, implicit, pPresent,
                                pPresentationParameters, pCTX, hFocusWindow, mode);
 }

src/gallium/state_trackers/nine/texture9.c

@@ -46,10 +46,11 @@ NineTexture9_ctor( struct NineTexture9 *This,
 {
     struct pipe_screen *screen = pParams->device->screen;
     struct pipe_resource *info = &This->base.base.info;
+    struct pipe_resource *resource;
     unsigned l;
     D3DSURFACE_DESC sfdesc;
     HRESULT hr;
-    const boolean shared_create = pSharedHandle && !*pSharedHandle;
+    void *user_buffer = NULL;
 
     DBG("(%p) Width=%u Height=%u Levels=%u Usage=%s Format=%s Pool=%s "
         "pSharedHandle=%p\n", This, Width, Height, Levels,
@@ -77,10 +78,7 @@ NineTexture9_ctor( struct NineTexture9 *This,
                 Pool == D3DPOOL_DEFAULT, D3DERR_INVALIDCALL);
 
     if (pSharedHandle && Pool == D3DPOOL_DEFAULT) {
-        /* Note: Below there is some implementation to support buffer sharing in
-         * this case, but it won't work for cross-process. Thus just ignore
-         * that code. */
-        if (shared_create) {
+        if (!*pSharedHandle) {
             DBG("Creating Texture with invalid handle. Importing will fail\n.");
             *pSharedHandle = (HANDLE)1; /* Wine would keep it NULL */
             pSharedHandle = NULL;
@@ -94,9 +92,6 @@ NineTexture9_ctor( struct NineTexture9 *This,
     if (Usage & D3DUSAGE_AUTOGENMIPMAP)
         Levels = 0;
 
-    This->base.format = Format;
-    This->base.base.usage = Usage;
-
     info->screen = screen;
     info->target = PIPE_TEXTURE_2D;
     info->format = d3d9_to_pipe_format(Format);
@@ -130,25 +125,15 @@ NineTexture9_ctor( struct NineTexture9 *This,
     if (Pool == D3DPOOL_SYSTEMMEM)
         info->usage = PIPE_USAGE_STAGING;
 
-    if (pSharedHandle && !shared_create) {
-        if (Pool == D3DPOOL_SYSTEMMEM) {
-            /* Hack for surface creation. */
-            This->base.base.resource = (struct pipe_resource *)*pSharedHandle;
-        } else {
-            struct pipe_resource *res;
-            res = screen->resource_from_handle(screen, info,
-                                      (struct winsys_handle *)pSharedHandle);
-            if (!res)
-                return D3DERR_NOTFOUND;
-            This->base.base.resource = res;
-        }
+    if (pSharedHandle && *pSharedHandle) { /* Pool == D3DPOOL_SYSTEMMEM */
+        user_buffer = (void *)*pSharedHandle;
     }
 
     This->surfaces = CALLOC(info->last_level + 1, sizeof(*This->surfaces));
     if (!This->surfaces)
         return E_OUTOFMEMORY;
 
-    hr = NineBaseTexture9_ctor(&This->base, pParams, D3DRTYPE_TEXTURE, Pool);
+    hr = NineBaseTexture9_ctor(&This->base, pParams, NULL, D3DRTYPE_TEXTURE, Format, Pool, Usage);
     if (FAILED(hr))
         return hr;
     This->base.pstype = (Height == 1) ? 1 : 0;
@@ -163,12 +148,19 @@ NineTexture9_ctor( struct NineTexture9 *This,
     sfdesc.Pool = Pool;
     sfdesc.MultiSampleType = D3DMULTISAMPLE_NONE;
     sfdesc.MultiSampleQuality = 0;
+
+    if (Pool == D3DPOOL_SYSTEMMEM)
+        resource = NULL;
+    else
+        resource = This->base.base.resource;
+
     for (l = 0; l <= info->last_level; ++l) {
         sfdesc.Width = u_minify(Width, l);
         sfdesc.Height = u_minify(Height, l);
 
         hr = NineSurface9_new(This->base.base.base.device, NineUnknown(This),
-                              This->base.base.resource, D3DRTYPE_TEXTURE, l, 0,
+                              resource, user_buffer,
+                              D3DRTYPE_TEXTURE, l, 0,
                               &sfdesc, &This->surfaces[l]);
         if (FAILED(hr))
             return hr;
@@ -176,19 +168,8 @@ NineTexture9_ctor( struct NineTexture9 *This,
 
     This->dirty_rect.depth = 1; /* widht == 0 means empty, depth stays 1 */
 
-    if (pSharedHandle) {
-        if (Pool == D3DPOOL_SYSTEMMEM) {
-            This->base.base.resource = NULL;
-            if (shared_create)
-                *pSharedHandle = This->surfaces[0]->base.data;
-        } else
-        if (shared_create) {
-            boolean ok;
-            ok = screen->resource_get_handle(screen, This->base.base.resource,
-                                         (struct winsys_handle *)pSharedHandle);
-            if (!ok)
-                return D3DERR_DRIVERINTERNALERROR;
-        }
+    if (pSharedHandle && !*pSharedHandle) {/* Pool == D3DPOOL_SYSTEMMEM */
+        *pSharedHandle = This->surfaces[0]->data;
     }
 
     return D3D_OK;
@@ -245,6 +226,9 @@ NineTexture9_LockRect( struct NineTexture9 *This,
                        const RECT *pRect,
                        DWORD Flags )
 {
+    DBG("This=%p Level=%u pLockedRect=%p pRect=%p Flags=%d\n",
+        This, Level, pLockedRect, pRect, Flags);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(Level == 0 || !(This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP),
                 D3DERR_INVALIDCALL);
@@ -257,6 +241,8 @@ HRESULT WINAPI
 NineTexture9_UnlockRect( struct NineTexture9 *This,
                          UINT Level )
 {
+    DBG("This=%p Level=%u\n", This, Level);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
 
     return NineSurface9_UnlockRect(This->surfaces[Level]);

src/gallium/state_trackers/nine/vertexbuffer9.c

@@ -85,8 +85,8 @@ NineVertexBuffer9_ctor( struct NineVertexBuffer9 *This,
     info->last_level = 0;
     info->nr_samples = 0;
 
-    hr = NineResource9_ctor(&This->base, pParams, TRUE, D3DRTYPE_VERTEXBUFFER,
-                            pDesc->Pool);
+    hr = NineResource9_ctor(&This->base, pParams, NULL, TRUE,
+                            D3DRTYPE_VERTEXBUFFER, pDesc->Pool, pDesc->Usage);
     if (FAILED(hr))
         return hr;
 

src/gallium/state_trackers/nine/vertexdeclaration9.c

@@ -95,21 +95,20 @@ nine_d3ddeclusage_check(unsigned usage, unsigned usage_idx)
 {
     switch (usage) {
     case D3DDECLUSAGE_POSITIONT:
-    case D3DDECLUSAGE_PSIZE:
     case D3DDECLUSAGE_TESSFACTOR:
     case D3DDECLUSAGE_DEPTH:
-    case D3DDECLUSAGE_FOG:
-    case D3DDECLUSAGE_SAMPLE:
-        return usage_idx <= 0;
     case D3DDECLUSAGE_NORMAL:
     case D3DDECLUSAGE_TANGENT:
     case D3DDECLUSAGE_BINORMAL:
-        return usage_idx <= 1;
     case D3DDECLUSAGE_POSITION:
     case D3DDECLUSAGE_BLENDWEIGHT:
     case D3DDECLUSAGE_BLENDINDICES:
     case D3DDECLUSAGE_COLOR:
-        return usage_idx <= 4;
+        return TRUE;
+    case D3DDECLUSAGE_PSIZE:
+    case D3DDECLUSAGE_FOG:
+    case D3DDECLUSAGE_SAMPLE:
+        return usage_idx <= 0;
     case D3DDECLUSAGE_TEXCOORD:
         return usage_idx <= 15;
     default:
@@ -118,8 +117,8 @@ nine_d3ddeclusage_check(unsigned usage, unsigned usage_idx)
 }
 
 #define NINE_DECLUSAGE_CASE0(n) case D3DDECLUSAGE_##n: return NINE_DECLUSAGE_##n
-#define NINE_DECLUSAGE_CASEi(n) case D3DDECLUSAGE_##n: return NINE_DECLUSAGE_##n(usage_idx)
-INLINE unsigned
+#define NINE_DECLUSAGE_CASEi(n) case D3DDECLUSAGE_##n: return NINE_DECLUSAGE_i(n, usage_idx)
+uint16_t
 nine_d3d9_to_nine_declusage(unsigned usage, unsigned usage_idx)
 {
     if (!nine_d3ddeclusage_check(usage, usage_idx))
@@ -135,7 +134,7 @@ nine_d3d9_to_nine_declusage(unsigned usage, unsigned usage_idx)
     NINE_DECLUSAGE_CASEi(TANGENT);
     NINE_DECLUSAGE_CASEi(BINORMAL);
     NINE_DECLUSAGE_CASE0(TESSFACTOR);
-    NINE_DECLUSAGE_CASE0(POSITIONT);
+    NINE_DECLUSAGE_CASEi(POSITIONT);
     NINE_DECLUSAGE_CASEi(COLOR);
     NINE_DECLUSAGE_CASE0(DEPTH);
     NINE_DECLUSAGE_CASE0(FOG);
@@ -148,58 +147,25 @@ nine_d3d9_to_nine_declusage(unsigned usage, unsigned usage_idx)
 
 static const char *nine_declusage_names[] =
 {
-    [NINE_DECLUSAGE_POSITION(0)]     = "POSITION",
-    [NINE_DECLUSAGE_POSITION(1)]     = "POSITION1",
-    [NINE_DECLUSAGE_POSITION(2)]     = "POSITION2",
-    [NINE_DECLUSAGE_POSITION(3)]     = "POSITION3",
-    [NINE_DECLUSAGE_POSITION(4)]     = "POSITION4",
-    [NINE_DECLUSAGE_BLENDWEIGHT(0)]  = "BLENDWEIGHT",
-    [NINE_DECLUSAGE_BLENDWEIGHT(1)]  = "BLENDWEIGHT1",
-    [NINE_DECLUSAGE_BLENDWEIGHT(2)]  = "BLENDWEIGHT2",
-    [NINE_DECLUSAGE_BLENDWEIGHT(3)]  = "BLENDWEIGHT3",
-    [NINE_DECLUSAGE_BLENDINDICES(0)] = "BLENDINDICES",
-    [NINE_DECLUSAGE_BLENDINDICES(1)] = "BLENDINDICES1",
-    [NINE_DECLUSAGE_BLENDINDICES(2)] = "BLENDINDICES2",
-    [NINE_DECLUSAGE_BLENDINDICES(3)] = "BLENDINDICES3",
-    [NINE_DECLUSAGE_NORMAL(0)]       = "NORMAL",
-    [NINE_DECLUSAGE_NORMAL(1)]       = "NORMAL1",
+    [NINE_DECLUSAGE_POSITION]        = "POSITION",
+    [NINE_DECLUSAGE_BLENDWEIGHT]     = "BLENDWEIGHT",
+    [NINE_DECLUSAGE_BLENDINDICES]    = "BLENDINDICES",
+    [NINE_DECLUSAGE_NORMAL]          = "NORMAL",
     [NINE_DECLUSAGE_PSIZE]           = "PSIZE",
-    [NINE_DECLUSAGE_TEXCOORD(0)]     = "TEXCOORD0",
-    [NINE_DECLUSAGE_TEXCOORD(1)]     = "TEXCOORD1",
-    [NINE_DECLUSAGE_TEXCOORD(2)]     = "TEXCOORD2",
-    [NINE_DECLUSAGE_TEXCOORD(3)]     = "TEXCOORD3",
-    [NINE_DECLUSAGE_TEXCOORD(4)]     = "TEXCOORD4",
-    [NINE_DECLUSAGE_TEXCOORD(5)]     = "TEXCOORD5",
-    [NINE_DECLUSAGE_TEXCOORD(6)]     = "TEXCOORD6",
-    [NINE_DECLUSAGE_TEXCOORD(7)]     = "TEXCOORD7",
-    [NINE_DECLUSAGE_TEXCOORD(8)]     = "TEXCOORD8",
-    [NINE_DECLUSAGE_TEXCOORD(9)]     = "TEXCOORD9",
-    [NINE_DECLUSAGE_TEXCOORD(10)]    = "TEXCOORD10",
-    [NINE_DECLUSAGE_TEXCOORD(11)]    = "TEXCOORD11",
-    [NINE_DECLUSAGE_TEXCOORD(12)]    = "TEXCOORD12",
-    [NINE_DECLUSAGE_TEXCOORD(13)]    = "TEXCOORD13",
-    [NINE_DECLUSAGE_TEXCOORD(14)]    = "TEXCOORD14",
-    [NINE_DECLUSAGE_TEXCOORD(15)]    = "TEXCOORD15",
-    [NINE_DECLUSAGE_TANGENT(0)]      = "TANGENT",
-    [NINE_DECLUSAGE_TANGENT(1)]      = "TANGENT1",
-    [NINE_DECLUSAGE_BINORMAL(0)]     = "BINORMAL",
-    [NINE_DECLUSAGE_BINORMAL(1)]     = "BINORMAL1",
+    [NINE_DECLUSAGE_TEXCOORD]        = "TEXCOORD",
+    [NINE_DECLUSAGE_TANGENT]         = "TANGENT",
+    [NINE_DECLUSAGE_BINORMAL]        = "BINORMAL",
     [NINE_DECLUSAGE_TESSFACTOR]      = "TESSFACTOR",
     [NINE_DECLUSAGE_POSITIONT]       = "POSITIONT",
-    [NINE_DECLUSAGE_COLOR(0)]        = "DIFFUSE",
-    [NINE_DECLUSAGE_COLOR(1)]        = "SPECULAR",
-    [NINE_DECLUSAGE_COLOR(2)]        = "COLOR2",
-    [NINE_DECLUSAGE_COLOR(3)]        = "COLOR3",
-    [NINE_DECLUSAGE_COLOR(4)]        = "COLOR4",
+    [NINE_DECLUSAGE_COLOR]           = "DIFFUSE",
     [NINE_DECLUSAGE_DEPTH]           = "DEPTH",
     [NINE_DECLUSAGE_FOG]             = "FOG",
     [NINE_DECLUSAGE_NONE]            = "(NONE)",
-    [NINE_DECLUSAGE_COUNT]           = "(OOB)"
 };
 static INLINE const char *
 nine_declusage_name(unsigned ndcl)
 {
-    return nine_declusage_names[MIN2(ndcl, Elements(nine_declusage_names) - 1)];
+    return nine_declusage_names[ndcl % NINE_DECLUSAGE_COUNT];
 }
 
 HRESULT
@@ -210,6 +176,8 @@ NineVertexDeclaration9_ctor( struct NineVertexDeclaration9 *This,
     const D3DCAPS9 *caps;
     unsigned i;
 
+    DBG("This=%p pParams=%p pElements=%p\n", This, pParams, pElements);
+
     HRESULT hr = NineUnknown_ctor(&This->base, pParams);
     if (FAILED(hr)) { return hr; }
 
@@ -223,15 +191,14 @@ NineVertexDeclaration9_ctor( struct NineVertexDeclaration9 *This,
 
     This->decls = CALLOC(This->nelems+1, sizeof(D3DVERTEXELEMENT9));
     This->elems = CALLOC(This->nelems, sizeof(struct pipe_vertex_element));
+    This->usage_map = CALLOC(This->nelems, sizeof(uint16_t));
     if (!This->decls || !This->elems) { return E_OUTOFMEMORY; }
     memcpy(This->decls, pElements, sizeof(D3DVERTEXELEMENT9)*(This->nelems+1));
 
-    memset(This->usage_map, 0xff, sizeof(This->usage_map));
-
     for (i = 0; i < This->nelems; ++i) {
-        uint8_t usage = nine_d3d9_to_nine_declusage(This->decls[i].Usage,
-                                                    This->decls[i].UsageIndex);
-        This->usage_map[usage] = i;
+        uint16_t usage = nine_d3d9_to_nine_declusage(This->decls[i].Usage,
+                                                     This->decls[i].UsageIndex);
+        This->usage_map[i] = usage;
 
         This->elems[i].src_offset = This->decls[i].Offset;
         This->elems[i].instance_divisor = 0;
@@ -239,11 +206,12 @@ NineVertexDeclaration9_ctor( struct NineVertexDeclaration9 *This,
         This->elems[i].src_format = decltype_format(This->decls[i].Type);
         /* XXX Remember Method (tesselation), Usage, UsageIndex */
 
-        DBG("VERTEXELEMENT[%u]: Stream=%u Offset=%u Type=%s DeclUsage=%s\n", i,
+        DBG("VERTEXELEMENT[%u]: Stream=%u Offset=%u Type=%s DeclUsage=%s%d\n", i,
             This->decls[i].Stream,
             This->decls[i].Offset,
             util_format_name(This->elems[i].src_format),
-            nine_declusage_name(usage));
+            nine_declusage_name(usage),
+            usage / NINE_DECLUSAGE_COUNT);
     }
 
     return D3D_OK;
@@ -256,6 +224,8 @@ NineVertexDeclaration9_dtor( struct NineVertexDeclaration9 *This )
         FREE(This->decls);
     if (This->elems)
         FREE(This->elems);
+    if (This->usage_map)
+        FREE(This->usage_map);
 
     NineUnknown_dtor(&This->base);
 }
@@ -460,6 +430,9 @@ NineVertexDeclaration9_ConvertStreamOutput(
     void *src_map;
     void *dst_map;
 
+    DBG("This=%p pDstBuf=%p DestIndex=%u VertexCount=%u pSrcBuf=%p so=%p\n",
+        This, pDstBuf, DestIndex, VertexCount, pSrcBuf, so);
+
     transkey.output_stride = 0;
     for (i = 0; i < This->nelems; ++i) {
         enum pipe_format format;

src/gallium/state_trackers/nine/vertexdeclaration9.h

@@ -40,9 +40,9 @@ struct NineVertexDeclaration9
     struct pipe_vertex_element *elems;
     unsigned nelems;
 
-    /* DECLUSAGE -> element index, for selecting the vertex element
+    /* index -> DECLUSAGE, for selecting the vertex element
      * for each VS input */
-    uint8_t usage_map[NINE_DECLUSAGE_COUNT];
+    uint16_t *usage_map;
 
     D3DVERTEXELEMENT9 *decls;
     DWORD fvf;

src/gallium/state_trackers/nine/vertexshader9.c

@@ -40,6 +40,9 @@ NineVertexShader9_ctor( struct NineVertexShader9 *This,
     HRESULT hr;
     unsigned i;
 
+    DBG("This=%p pParams=%p pFunction=%p cso=%p\n",
+        This, pParams, pFunction, cso);
+
     hr = NineUnknown_ctor(&This->base, pParams);
     if (FAILED(hr))
         return hr;

src/gallium/state_trackers/nine/vertexshader9.h

@@ -32,7 +32,7 @@ struct NineVertexShader9
     struct nine_shader_variant variant;
 
     struct {
-        uint8_t ndecl; /* NINE_DECLUSAGE_x */
+        uint16_t ndecl; /* NINE_DECLUSAGE_x */
     } input_map[PIPE_MAX_ATTRIBS];
     unsigned num_inputs;
 

src/gallium/state_trackers/nine/volume9.c

@@ -342,6 +342,9 @@ NineVolume9_CopyVolume( struct NineVolume9 *This,
     uint8_t *p_dst;
     const uint8_t *p_src;
 
+    DBG("This=%p From=%p dstx=%u dsty=%u dstz=%u pSrcBox=%p\n",
+        This, From, dstx, dsty, dstz, pSrcBox);
+
     user_assert(This->desc.Format == From->desc.Format, D3DERR_INVALIDCALL);
 
     dst_box.x = dstx;
@@ -527,6 +530,9 @@ NineVolume9_SetPrivateData( struct NineVolume9 *This,
     struct pheader *header;
     const void *user_data = pData;
 
+    DBG("This=%p refguid=%p pData=%p SizeOfData=%d Flags=%d\n",
+        This, refguid, pData, SizeOfData, Flags);
+
     if (Flags & D3DSPD_IUNKNOWN)
         user_assert(SizeOfData == sizeof(IUnknown *), D3DERR_INVALIDCALL);
 
@@ -593,6 +599,8 @@ NineVolume9_FreePrivateData( struct NineVolume9 *This,
 {
     struct pheader *header;
 
+    DBG("This=%p refguid=%p\n", This, refguid);
+
     header = util_hash_table_get(This->pdata, refguid);
     if (!header) { return D3DERR_NOTFOUND; }
 

src/gallium/state_trackers/nine/volumetexture9.c

@@ -41,6 +41,11 @@ NineVolumeTexture9_ctor( struct NineVolumeTexture9 *This,
     D3DVOLUME_DESC voldesc;
     HRESULT hr;
 
+    DBG("This=%p pParams=%p Width=%u Height=%u Depth=%u Levels=%u "
+        "Usage=%d Format=%d Pool=%d pSharedHandle=%p\n",
+        This, pParams, Width, Height, Depth, Levels,
+        Usage, Format, Pool, pSharedHandle);
+
     /* An IDirect3DVolume9 cannot be bound as a render target can it ? */
     user_assert(!(Usage & (D3DUSAGE_RENDERTARGET | D3DUSAGE_DEPTHSTENCIL)),
                 D3DERR_INVALIDCALL);
@@ -52,9 +57,6 @@ NineVolumeTexture9_ctor( struct NineVolumeTexture9 *This,
     if (Usage & D3DUSAGE_AUTOGENMIPMAP)
         Levels = 0;
 
-    This->base.format = Format;
-    This->base.base.usage = Usage;
-
     info->screen = pParams->device->screen;
     info->target = PIPE_TEXTURE_3D;
     info->format = d3d9_to_pipe_format(Format);
@@ -83,8 +85,8 @@ NineVolumeTexture9_ctor( struct NineVolumeTexture9 *This,
         return E_OUTOFMEMORY;
     This->base.pstype = 3;
 
-    hr = NineBaseTexture9_ctor(&This->base, pParams,
-                               D3DRTYPE_VOLUMETEXTURE, Pool);
+    hr = NineBaseTexture9_ctor(&This->base, pParams, NULL,
+                               D3DRTYPE_VOLUMETEXTURE, Format, Pool, Usage);
     if (FAILED(hr))
         return hr;
 
@@ -159,6 +161,9 @@ NineVolumeTexture9_LockBox( struct NineVolumeTexture9 *This,
                             const D3DBOX *pBox,
                             DWORD Flags )
 {
+    DBG("This=%p Level=%u pLockedVolume=%p pBox=%p Flags=%d\n",
+        This, Level, pLockedVolume, pBox, Flags);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
     user_assert(Level == 0 || !(This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP),
                 D3DERR_INVALIDCALL);
@@ -171,6 +176,8 @@ HRESULT WINAPI
 NineVolumeTexture9_UnlockBox( struct NineVolumeTexture9 *This,
                               UINT Level )
 {
+    DBG("This=%p Level=%u\n", This, Level);
+
     user_assert(Level <= This->base.base.info.last_level, D3DERR_INVALIDCALL);
 
     return NineVolume9_UnlockBox(This->volumes[Level]);
@@ -180,6 +187,8 @@ HRESULT WINAPI
 NineVolumeTexture9_AddDirtyBox( struct NineVolumeTexture9 *This,
                                 const D3DBOX *pDirtyBox )
 {
+    DBG("This=%p pDirtybox=%p\n", This, pDirtyBox);
+
     if (This->base.base.pool != D3DPOOL_MANAGED) {
         if (This->base.base.usage & D3DUSAGE_AUTOGENMIPMAP)
             This->base.dirty_mip = TRUE;

src/gallium/state_trackers/wgl/opengl32.def

@@ -367,7 +367,6 @@ EXPORTS
 	wglUseFontBitmapsW
 	wglUseFontOutlinesA
 	wglUseFontOutlinesW
-	wglGetExtensionsStringARB
 	DrvCopyContext
 	DrvCreateContext
 	DrvCreateLayerContext

src/gallium/targets/d3dadapter9/Makefile.am

@@ -62,7 +62,7 @@ d3dadapter9_la_LDFLAGS = \
 	-shrext .so \
 	-module \
 	-no-undefined \
-	-version-number $(NINE_MAJOR):$(NINE_MINOR) \
+	-version-number $(NINE_MAJOR):$(NINE_MINOR):$(NINE_TINY) \
 	$(GC_SECTIONS) \
 	$(LD_NO_UNDEFINED)
 

src/gallium/targets/d3dadapter9/d3d.pc.in

@@ -6,6 +6,6 @@ moduledir=@D3D_DRIVER_INSTALL_DIR@
 
 Name: d3d
 Description: Native D3D driver modules
-Version: @VERSION@
+Version: @NINE_VERSION@
 Requires.private: @DRI_PC_REQ_PRIV@
 Cflags: -I${includedir}

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

@@ -325,6 +325,9 @@ static boolean do_winsys_init(struct radeon_drm_winsys *ws)
                          &ws->info.max_sclk);
     ws->info.max_sclk /= 1000;
 
+    radeon_get_drm_value(ws->fd, RADEON_INFO_SI_BACKEND_ENABLED_MASK, NULL,
+                         &ws->info.si_backend_enabled_mask);
+
     ws->num_cpus = sysconf(_SC_NPROCESSORS_ONLN);
 
     /* Generation-specific queries. */

src/gallium/winsys/radeon/drm/radeon_winsys.h

@@ -231,6 +231,7 @@ struct radeon_info {
 
     boolean                     si_tile_mode_array_valid;
     uint32_t                    si_tile_mode_array[32];
+    uint32_t                    si_backend_enabled_mask;
 
     boolean                     cik_macrotile_mode_array_valid;
     uint32_t                    cik_macrotile_mode_array[16];

src/glsl/lower_const_arrays_to_uniforms.cpp

@@ -49,6 +49,7 @@ public:
    {
       instructions = insts;
       progress = false;
+      index = 0;
    }
 
    bool run()
@@ -62,6 +63,7 @@ public:
 private:
    exec_list *instructions;
    bool progress;
+   unsigned index;
 };
 
 void
@@ -70,14 +72,20 @@ lower_const_array_visitor::handle_rvalue(ir_rvalue **rvalue)
    if (!*rvalue)
       return;
 
-   ir_constant *con = (*rvalue)->as_constant();
+   ir_dereference_array *dra = (*rvalue)->as_dereference_array();
+   if (!dra)
+      return;
+
+   ir_constant *con = dra->array->as_constant();
    if (!con || !con->type->is_array())
       return;
 
    void *mem_ctx = ralloc_parent(con);
 
+   char *uniform_name = ralloc_asprintf(mem_ctx, "constarray__%d", index++);
+
    ir_variable *uni =
-      new(mem_ctx) ir_variable(con->type, "constarray", ir_var_uniform);
+      new(mem_ctx) ir_variable(con->type, uniform_name, ir_var_uniform);
    uni->constant_initializer = con;
    uni->constant_value = con;
    uni->data.has_initializer = true;
@@ -87,7 +95,8 @@ lower_const_array_visitor::handle_rvalue(ir_rvalue **rvalue)
    uni->data.max_array_access = uni->type->length - 1;
    instructions->push_head(uni);
 
-   *rvalue = new(mem_ctx) ir_dereference_variable(uni);
+   ir_dereference_variable *varref = new(mem_ctx) ir_dereference_variable(uni);
+   *rvalue = new(mem_ctx) ir_dereference_array(varref, dra->array_index);
 
    progress = true;
 }

src/mapi/glapi/SConscript

@@ -16,6 +16,7 @@ if env['platform'] == 'windows':
     env.Append(CPPDEFINES = [
         '_GDI32_', # prevent gl* being declared __declspec(dllimport) in MS headers
         'BUILD_GL32', # declare gl* as __declspec(dllexport) in Mesa headers
+        'KHRONOS_DLL_EXPORTS', # declare gl* as __declspec(dllexport) in Khronos headers
     ])
     if env['gles']:
         env.Append(CPPDEFINES = ['_GLAPI_DLL_EXPORTS'])

src/mesa/Makefile.am

@@ -153,7 +153,7 @@ libmesagallium_la_LIBADD = \
 libmesa_sse41_la_SOURCES = \
 	main/streaming-load-memcpy.c \
 	main/sse_minmax.c
-libmesa_sse41_la_CFLAGS = $(AM_CFLAGS) -msse4.1
+libmesa_sse41_la_CFLAGS = $(AM_CFLAGS) $(SSE41_CFLAGS)
 
 pkgconfigdir = $(libdir)/pkgconfig
 pkgconfig_DATA = gl.pc

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

@@ -28,7 +28,6 @@
 static const struct brw_device_info brw_device_info_i965 = {
    .gen = 4,
    .has_negative_rhw_bug = true,
-   .needs_unlit_centroid_workaround = true,
    .max_vs_threads = 16,
    .max_gs_threads = 2,
    .max_wm_threads = 8 * 4,
@@ -42,7 +41,6 @@ static const struct brw_device_info brw_device_info_g4x = {
    .has_pln = true,
    .has_compr4 = true,
    .has_surface_tile_offset = true,
-   .needs_unlit_centroid_workaround = true,
    .is_g4x = true,
    .max_vs_threads = 32,
    .max_gs_threads = 2,
@@ -57,7 +55,6 @@ static const struct brw_device_info brw_device_info_ilk = {
    .has_pln = true,
    .has_compr4 = true,
    .has_surface_tile_offset = true,
-   .needs_unlit_centroid_workaround = true,
    .max_vs_threads = 72,
    .max_gs_threads = 32,
    .max_wm_threads = 12 * 6,

src/mesa/drivers/dri/i965/brw_vec4_generator.cpp

@@ -526,13 +526,17 @@ vec4_generator::generate_gs_set_write_offset(struct brw_reg dst,
     *
     * We can do this with the following EU instruction:
     *
-    *     mul(2) dst.3<1>UD src0<8;2,4>UD src1   { Align1 WE_all }
+    *     mul(2) dst.3<1>UD src0<8;2,4>UD src1<...>UW   { Align1 WE_all }
     */
    brw_push_insn_state(p);
    brw_set_default_access_mode(p, BRW_ALIGN_1);
    brw_set_default_mask_control(p, BRW_MASK_DISABLE);
+   assert(brw->gen >= 7 &&
+          src1.file == BRW_IMMEDIATE_VALUE &&
+          src1.type == BRW_REGISTER_TYPE_UD &&
+          src1.dw1.ud <= USHRT_MAX);
    brw_MUL(p, suboffset(stride(dst, 2, 2, 1), 3), stride(src0, 8, 2, 4),
-           src1);
+           retype(src1, BRW_REGISTER_TYPE_UW));
    brw_set_default_access_mode(p, BRW_ALIGN_16);
    brw_pop_insn_state(p);
 }

src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp

@@ -3354,6 +3354,7 @@ vec4_visitor::move_uniform_array_access_to_pull_constants()
 {
    int pull_constant_loc[this->uniforms];
    memset(pull_constant_loc, -1, sizeof(pull_constant_loc));
+   bool nested_reladdr;
 
    /* Walk through and find array access of uniforms.  Put a copy of that
     * uniform in the pull constant buffer.
@@ -3361,44 +3362,51 @@ vec4_visitor::move_uniform_array_access_to_pull_constants()
     * Note that we don't move constant-indexed accesses to arrays.  No
     * testing has been done of the performance impact of this choice.
     */
-   foreach_block_and_inst_safe(block, vec4_instruction, inst, cfg) {
-      for (int i = 0 ; i < 3; i++) {
-	 if (inst->src[i].file != UNIFORM || !inst->src[i].reladdr)
-	    continue;
+   do {
+      nested_reladdr = false;
 
-	 int uniform = inst->src[i].reg;
+      foreach_block_and_inst_safe(block, vec4_instruction, inst, cfg) {
+         for (int i = 0 ; i < 3; i++) {
+            if (inst->src[i].file != UNIFORM || !inst->src[i].reladdr)
+               continue;
 
-	 /* If this array isn't already present in the pull constant buffer,
-	  * add it.
-	  */
-	 if (pull_constant_loc[uniform] == -1) {
-	    const gl_constant_value **values =
-               &stage_prog_data->param[uniform * 4];
+            int uniform = inst->src[i].reg;
 
-	    pull_constant_loc[uniform] = stage_prog_data->nr_pull_params / 4;
+            if (inst->src[i].reladdr->reladdr)
+               nested_reladdr = true;  /* will need another pass */
 
-	    assert(uniform < uniform_array_size);
-	    for (int j = 0; j < uniform_size[uniform] * 4; j++) {
-	       stage_prog_data->pull_param[stage_prog_data->nr_pull_params++]
-                  = values[j];
-	    }
-	 }
+            /* If this array isn't already present in the pull constant buffer,
+             * add it.
+             */
+            if (pull_constant_loc[uniform] == -1) {
+               const gl_constant_value **values =
+                  &stage_prog_data->param[uniform * 4];
 
-	 /* Set up the annotation tracking for new generated instructions. */
-	 base_ir = inst->ir;
-	 current_annotation = inst->annotation;
+               pull_constant_loc[uniform] = stage_prog_data->nr_pull_params / 4;
 
-	 dst_reg temp = dst_reg(this, glsl_type::vec4_type);
+               assert(uniform < uniform_array_size);
+               for (int j = 0; j < uniform_size[uniform] * 4; j++) {
+                  stage_prog_data->pull_param[stage_prog_data->nr_pull_params++]
+                     = values[j];
+               }
+            }
 
-	 emit_pull_constant_load(block, inst, temp, inst->src[i],
-				 pull_constant_loc[uniform]);
+            /* Set up the annotation tracking for new generated instructions. */
+            base_ir = inst->ir;
+            current_annotation = inst->annotation;
 
-	 inst->src[i].file = temp.file;
-	 inst->src[i].reg = temp.reg;
-	 inst->src[i].reg_offset = temp.reg_offset;
-	 inst->src[i].reladdr = NULL;
+            dst_reg temp = dst_reg(this, glsl_type::vec4_type);
+
+            emit_pull_constant_load(block, inst, temp, inst->src[i],
+                                    pull_constant_loc[uniform]);
+
+            inst->src[i].file = temp.file;
+            inst->src[i].reg = temp.reg;
+            inst->src[i].reg_offset = temp.reg_offset;
+            inst->src[i].reladdr = NULL;
+         }
       }
-   }
+   } while (nested_reladdr);
 
    /* Now there are no accesses of the UNIFORM file with a reladdr, so
     * no need to track them as larger-than-vec4 objects.  This will be

src/mesa/drivers/dri/i965/gen6_sf_state.c

@@ -129,6 +129,26 @@ get_attr_override(const struct brw_vue_map *vue_map, int urb_entry_read_offset,
 }
 
 
+static bool
+is_drawing_points(const struct brw_context *brw)
+{
+   /* Determine if the primitives *reaching the SF* are points */
+   /* _NEW_POLYGON */
+   if (brw->ctx.Polygon.FrontMode == GL_POINT ||
+       brw->ctx.Polygon.BackMode == GL_POINT) {
+      return true;
+   }
+
+   if (brw->geometry_program) {
+      /* BRW_NEW_GEOMETRY_PROGRAM */
+      return brw->geometry_program->OutputType == GL_POINTS;
+   } else {
+      /* BRW_NEW_PRIMITIVE */
+      return brw->primitive == _3DPRIM_POINTLIST;
+   }
+}
+
+
 /**
  * Create the mapping from the FS inputs we produce to the previous pipeline
  * stage (GS or VS) outputs they source from.
@@ -149,6 +169,23 @@ calculate_attr_overrides(const struct brw_context *brw,
    /* _NEW_LIGHT */
    bool shade_model_flat = brw->ctx.Light.ShadeModel == GL_FLAT;
 
+   /* From the Ivybridge PRM, Vol 2 Part 1, 3DSTATE_SBE,
+    * description of dw10 Point Sprite Texture Coordinate Enable:
+    *
+    * "This field must be programmed to zero when non-point primitives
+    * are rendered."
+    *
+    * The SandyBridge PRM doesn't explicitly say that point sprite enables
+    * must be programmed to zero when rendering non-point primitives, but
+    * the IvyBridge PRM does, and if we don't, we get garbage.
+    *
+    * This is not required on Haswell, as the hardware ignores this state
+    * when drawing non-points -- although we do still need to be careful to
+    * correctly set the attr overrides.
+    */
+   /* BRW_NEW_PRIMITIVE | BRW_NEW_GEOMETRY_PROGRAM */
+   bool drawing_points = is_drawing_points(brw);
+
    /* Initialize all the attr_overrides to 0.  In the loop below we'll modify
     * just the ones that correspond to inputs used by the fs.
     */
@@ -167,18 +204,20 @@ calculate_attr_overrides(const struct brw_context *brw,
 
       /* _NEW_POINT */
       bool point_sprite = false;
-      if (brw->ctx.Point.PointSprite &&
-	  (attr >= VARYING_SLOT_TEX0 && attr <= VARYING_SLOT_TEX7) &&
-	  brw->ctx.Point.CoordReplace[attr - VARYING_SLOT_TEX0]) {
-         point_sprite = true;
+      if (drawing_points) {
+         if (brw->ctx.Point.PointSprite &&
+             (attr >= VARYING_SLOT_TEX0 && attr <= VARYING_SLOT_TEX7) &&
+             brw->ctx.Point.CoordReplace[attr - VARYING_SLOT_TEX0]) {
+            point_sprite = true;
+         }
+
+         if (attr == VARYING_SLOT_PNTC)
+            point_sprite = true;
+
+         if (point_sprite)
+            *point_sprite_enables |= (1 << input_index);
       }
 
-      if (attr == VARYING_SLOT_PNTC)
-         point_sprite = true;
-
-      if (point_sprite)
-	 *point_sprite_enables |= (1 << input_index);
-
       /* flat shading */
       if (interp_qualifier == INTERP_QUALIFIER_FLAT ||
           (shade_model_flat && is_gl_Color &&
@@ -410,7 +449,9 @@ const struct brw_tracked_state gen6_sf_state = {
 		_NEW_POINT |
                 _NEW_MULTISAMPLE),
       .brw   = (BRW_NEW_CONTEXT |
-		BRW_NEW_FRAGMENT_PROGRAM |
+                BRW_NEW_FRAGMENT_PROGRAM |
+                BRW_NEW_GEOMETRY_PROGRAM |
+                BRW_NEW_PRIMITIVE |
                 BRW_NEW_VUE_MAP_GEOM_OUT),
       .cache = CACHE_NEW_WM_PROG
    },

src/mesa/drivers/dri/i965/gen7_sf_state.c

@@ -92,7 +92,9 @@ const struct brw_tracked_state gen7_sbe_state = {
 		_NEW_POINT |
 		_NEW_PROGRAM),
       .brw   = (BRW_NEW_CONTEXT |
-		BRW_NEW_FRAGMENT_PROGRAM |
+                BRW_NEW_FRAGMENT_PROGRAM |
+                BRW_NEW_GEOMETRY_PROGRAM |
+                BRW_NEW_PRIMITIVE |
                 BRW_NEW_VUE_MAP_GEOM_OUT),
       .cache = CACHE_NEW_WM_PROG
    },

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

@@ -1745,8 +1745,8 @@ intel_miptree_map_gtt(struct brw_context *brw,
    unsigned int bw, bh;
    void *base;
    unsigned int image_x, image_y;
-   int x = map->x;
-   int y = map->y;
+   intptr_t x = map->x;
+   intptr_t y = map->y;
 
    /* For compressed formats, the stride is the number of bytes per
     * row of blocks.  intel_miptree_get_image_offset() already does
@@ -1772,7 +1772,8 @@ intel_miptree_map_gtt(struct brw_context *brw,
       map->ptr = base + y * map->stride + x * mt->cpp;
    }
 
-   DBG("%s: %d,%d %dx%d from mt %p (%s) %d,%d = %p/%d\n", __FUNCTION__,
+   DBG("%s: %d,%d %dx%d from mt %p (%s) "
+       "%"PRIiPTR",%"PRIiPTR" = %p/%d\n", __FUNCTION__,
        map->x, map->y, map->w, map->h,
        mt, _mesa_get_format_name(mt->format),
        x, y, map->ptr, map->stride);

src/mesa/drivers/haiku/swrast/SConscript

@@ -13,6 +13,7 @@ env.Append(CPPPATH = [
 ])
 
 env.Prepend(LIBS = [
+    mesautil,
     glsl,
     mesa,
 ])

src/mesa/main/get_hash_params.py

@@ -627,7 +627,7 @@ descriptor=[
 # == GL_CURRENT_MATRIX_NV
   [ "CURRENT_MATRIX_ARB", "LOC_CUSTOM, TYPE_MATRIX, 0, extra_ARB_vertex_program_ARB_fragment_program" ],
 # == GL_CURRENT_MATRIX_NV
-  [ "TRANSPOSE_CURRENT_MATRIX_ARB", "LOC_CUSTOM, TYPE_MATRIX, 0, extra_ARB_vertex_program_ARB_fragment_program" ],
+  [ "TRANSPOSE_CURRENT_MATRIX_ARB", "LOC_CUSTOM, TYPE_MATRIX_T, 0, extra_ARB_vertex_program_ARB_fragment_program" ],
 # == GL_PROGRAM_ERROR_POSITION_NV
   [ "PROGRAM_ERROR_POSITION_ARB", "CONTEXT_INT(Program.ErrorPos), extra_ARB_vertex_program_ARB_fragment_program" ],
 

src/mesa/main/teximage.c

@@ -1542,7 +1542,7 @@ _mesa_legal_texture_dimensions(struct gl_context *ctx, GLenum target,
       maxSize >>= level;
       if (width < 2 * border || width > 2 * border + maxSize)
          return GL_FALSE;
-      if (height < 1 || height > ctx->Const.MaxArrayTextureLayers)
+      if (height < 0 || height > ctx->Const.MaxArrayTextureLayers)
          return GL_FALSE;
       if (!ctx->Extensions.ARB_texture_non_power_of_two) {
          if (width > 0 && !_mesa_is_pow_two(width - 2 * border))

src/mesa/main/texstore.c

@@ -2229,7 +2229,7 @@ _mesa_compute_compressed_pixelstore(GLuint dims, mesa_format texFormat,
 
       if (packing->RowLength) {
          store->TotalBytesPerRow = packing->CompressedBlockSize *
-            (packing->RowLength + bw - 1) / bw;
+            ((packing->RowLength + bw - 1) / bw);
       }
 
       store->SkipBytes += packing->SkipPixels * packing->CompressedBlockSize / bw;

src/mesa/program/ir_to_mesa.cpp

@@ -2946,9 +2946,7 @@ _mesa_ir_link_shader(struct gl_context *ctx, struct gl_shader_program *prog)
 	 GLenum target = _mesa_shader_stage_to_program(prog->_LinkedShaders[i]->Stage);
 	 lower_instructions(ir, (MOD_TO_FRACT | DIV_TO_MUL_RCP | EXP_TO_EXP2
 				 | LOG_TO_LOG2 | INT_DIV_TO_MUL_RCP
-				 | ((options->EmitNoPow) ? POW_TO_EXP2 : 0)
-				 | ((target == GL_VERTEX_PROGRAM_ARB) ? SAT_TO_CLAMP
-                                    : 0)));
+				 | ((options->EmitNoPow) ? POW_TO_EXP2 : 0)));
 
 	 progress = do_lower_jumps(ir, true, true, options->EmitNoMainReturn, options->EmitNoCont, options->EmitNoLoops) || progress;
 

src/mesa/state_tracker/st_cb_clear.c

@@ -88,6 +88,14 @@ st_destroy_clear(struct st_context *st)
       cso_delete_vertex_shader(st->cso_context, st->clear.vs);
       st->clear.vs = NULL;
    }
+   if (st->clear.vs_layered) {
+      cso_delete_vertex_shader(st->cso_context, st->clear.vs_layered);
+      st->clear.vs_layered = NULL;
+   }
+   if (st->clear.gs_layered) {
+      cso_delete_geometry_shader(st->cso_context, st->clear.gs_layered);
+      st->clear.gs_layered = NULL;
+   }
 }
 
 
@@ -127,6 +135,7 @@ set_vertex_shader(struct st_context *st)
    }
 
    cso_set_vertex_shader_handle(st->cso_context, st->clear.vs);
+   cso_set_geometry_shader_handle(st->cso_context, NULL);
 }
 
 
@@ -135,18 +144,25 @@ set_vertex_shader_layered(struct st_context *st)
 {
    struct pipe_context *pipe = st->pipe;
 
-   if (!pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_INSTANCEID) ||
-       !pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_VS_LAYER_VIEWPORT)) {
-      assert(!"Got layered clear, but the VS layer output is unsupported");
+   if (!pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_INSTANCEID)) {
+      assert(!"Got layered clear, but VS instancing is unsupported");
       set_vertex_shader(st);
       return;
    }
 
    if (!st->clear.vs_layered) {
-      st->clear.vs_layered = util_make_layered_clear_vertex_shader(pipe);
+      bool vs_layer =
+         pipe->screen->get_param(pipe->screen, PIPE_CAP_TGSI_VS_LAYER_VIEWPORT);
+      if (vs_layer) {
+         st->clear.vs_layered = util_make_layered_clear_vertex_shader(pipe);
+      } else {
+         st->clear.vs_layered = util_make_layered_clear_helper_vertex_shader(pipe);
+         st->clear.gs_layered = util_make_layered_clear_geometry_shader(pipe);
+      }
    }
 
    cso_set_vertex_shader_handle(st->cso_context, st->clear.vs_layered);
+   cso_set_geometry_shader_handle(st->cso_context, st->clear.gs_layered);
 }
 
 
@@ -332,7 +348,6 @@ clear_with_quad(struct gl_context *ctx, unsigned clear_buffers)
    }
 
    set_fragment_shader(st);
-   cso_set_geometry_shader_handle(st->cso_context, NULL);
 
    if (num_layers > 1)
       set_vertex_shader_layered(st);

src/mesa/state_tracker/st_context.h

@@ -181,6 +181,7 @@ struct st_context
       void *vs;
       void *fs;
       void *vs_layered;
+      void *gs_layered;
    } clear;
 
    /** used for anything using util_draw_vertex_buffer */

src/mesa/state_tracker/st_glsl_to_tgsi.cpp

@@ -4022,6 +4022,7 @@ get_pixel_transfer_visitor(struct st_fragment_program *fp,
 
       newinst = v->emit(NULL, inst->op, inst->dst, src_regs[0], src_regs[1], src_regs[2]);
       newinst->tex_target = inst->tex_target;
+      newinst->sampler_array_size = inst->sampler_array_size;
    }
 
    /* Make modifications to fragment program info. */
@@ -4101,6 +4102,7 @@ get_bitmap_visitor(struct st_fragment_program *fp,
 
       newinst = v->emit(NULL, inst->op, inst->dst, src_regs[0], src_regs[1], src_regs[2]);
       newinst->tex_target = inst->tex_target;
+      newinst->sampler_array_size = inst->sampler_array_size;
    }
 
    /* Make modifications to fragment program info. */
@@ -4524,8 +4526,10 @@ compile_tgsi_instruction(struct st_translate *t,
                              inst->saturate,
                              clamp_dst_color_output);
 
-   for (i = 0; i < num_src; i++) 
+   for (i = 0; i < num_src; i++) {
+      assert(inst->src[i].file != PROGRAM_UNDEFINED);
       src[i] = translate_src(t, &inst->src[i]);
+   }
 
    switch(inst->op) {
    case TGSI_OPCODE_BGNLOOP:
@@ -4555,6 +4559,7 @@ compile_tgsi_instruction(struct st_translate *t,
    case TGSI_OPCODE_TG4:
    case TGSI_OPCODE_LODQ:
       src[num_src] = t->samplers[inst->sampler.index];
+      assert(src[num_src].File != TGSI_FILE_NULL);
       if (inst->sampler.reladdr)
          src[num_src] =
             ureg_src_indirect(src[num_src], ureg_src(t->address[2]));