[Mesa-dev] [PATCH 1/5] glsl: print non-zero bindings of variables

2016-10-13 Thread Nicolai Hähnle
From: Nicolai Hähnle --- src/compiler/glsl/ir_print_visitor.cpp | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/src/compiler/glsl/ir_print_visitor.cpp b/src/compiler/glsl/ir_print_visitor.cpp index c238c16..efb728b 100644 ---

Re: [Mesa-dev] Required-for-rendering internal formats

2016-10-13 Thread Nicolai Hähnle
On 13.10.2016 17:15, Ilia Mirkin wrote: Anyone else have opinions on this? Should st/mesa be adding PIPE_BIND_RENDER_TARGET for all textures of the internal formats listed in table 8.12? (Which would, in this case, force RGBA8 to be used since RGBA4 is non-renderable.) Or is it perfectly legal

[Mesa-dev] [Bug 98223] dEQP GLES3.1 program_interface_query failures w/ error "could not find target resource"

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98223 --- Comment #6 from Alejandro Piñeiro (freenode IRC: apinheiro) --- (In reply to Randy from comment #5) > We can use the patch (in > https://lists.freedesktop.org/archives/mesa-dev/2016-August/124910.html) to > workaround

Re: [Mesa-dev] [PATCH mesa] swr: [rasterizer common] fix assert index

2016-10-13 Thread Rowley, Timothy O
Reviewed-by: Tim Rowley > On Oct 12, 2016, at 4:13 PM, Eric Engestrom > wrote: Fixes: b3bd8bb611bb465d2e5e ("swr: [rasterizer core] add support for "RAW" surface format") CovID:

[Mesa-dev] [PATCH 2/2] radv: Use new image load/store intrinsic signatures

2016-10-13 Thread Tom Stellard
These were changed in LLVM r284024. --- src/amd/common/ac_nir_to_llvm.c | 131 1 file changed, 107 insertions(+), 24 deletions(-) diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c index 9c764c7..4fba7d3 100644 ---

[Mesa-dev] [Bug 98169] lm_sensors hud option crashes unigine heaven

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98169 Steven Toth changed: What|Removed |Added Attachment #127222|0 |1 is

[Mesa-dev] [Bug 98223] dEQP GLES3.1 program_interface_query failures w/ error "could not find target resource"

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98223 --- Comment #5 from Randy --- We can use the patch (in https://lists.freedesktop.org/archives/mesa-dev/2016-August/124910.html) to workaround the name issue, while some cases still fail due to wrong property, like ARRAY_SIZE

Re: [Mesa-dev] Required-for-rendering internal formats

2016-10-13 Thread Ilia Mirkin
Anyone else have opinions on this? Should st/mesa be adding PIPE_BIND_RENDER_TARGET for all textures of the internal formats listed in table 8.12? (Which would, in this case, force RGBA8 to be used since RGBA4 is non-renderable.) Or is it perfectly legal for a format to be not renderable when used

[Mesa-dev] [PATCH 1/2] radv: Fix incorrect comment

2016-10-13 Thread Tom Stellard
--- src/amd/common/ac_nir_to_llvm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c index e6ff7c8..9c764c7 100644 --- a/src/amd/common/ac_nir_to_llvm.c +++ b/src/amd/common/ac_nir_to_llvm.c @@ -2373,8

[Mesa-dev] [PATCH 2/5] st/mesa: fix fragment shader output mapping

2016-10-13 Thread Nicolai Hähnle
From: Nicolai Hähnle Properly handle the case where there is a gap in the assigned output locations, e.g. a fragment shader writes to color buffer 2 but not to color buffers 0 & 1. Fixes

[Mesa-dev] [PATCH 4/5] st/glsl_to_tgsi: fix atomic counter addressing

2016-10-13 Thread Nicolai Hähnle
From: Nicolai Hähnle When more than one atomic counter buffer is in use, UniformStorage[n].opaque is set up to contain indices that are contiguous across all used buffers. This appears to be used by i965 via NIR, but for TGSI we do not treat atomic counter buffers as

[Mesa-dev] [PATCH 5/5] st/glsl_to_tgsi: fix [ui]vec[34] conversion to double

2016-10-13 Thread Nicolai Hähnle
From: Nicolai Hähnle The corresponding opcodes for integers need to be treated the same as F2D. Fixes GL45-CTS.gpu_shader_fp64.conversions. --- src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git

[Mesa-dev] [PATCH 3/5] st/glsl_to_tgsi: fix a corner case of std140 layout in uniform buffers

2016-10-13 Thread Nicolai Hähnle
From: Nicolai Hähnle See the comment in the code for an explanation. This fixes GL45-CTS.buffer_storage.map_persistent_draw. --- src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 29 - 1 file changed, 28 insertions(+), 1 deletion(-) diff --git

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #8 from shinji.suz...@gmail.com --- Comment on attachment 127267 --> https://bugs.freedesktop.org/attachment.cgi?id=127267 Lock the shared state mutex and work with a local reference of so->fence Review of attachment 127267:

[Mesa-dev] [Bug 98169] lm_sensors hud option crashes unigine heaven

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98169 --- Comment #31 from Christoph Haag --- Yes, your new patch works. No crashing, all panes are visible and show the correct values. -- You are receiving this mail because: You are the QA Contact for the

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #9 from shinji.suz...@gmail.com --- oops. I'm getting confused. Concurrent programmingis hard. > screen->fence_reference(screen, >fence, NULL); will not be executed in arbitrary order but serially due to the mutex locking. Still the

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #10 from shinji.suz...@gmail.com --- >I'm not confortable about sreen->fence_ference() being called concurrently. I'm not comfortable about sreen->fence_finish() being called concurrently. -- You are receiving this mail because:

Re: [Mesa-dev] [PATCH] gallium: add PIPE_RESOURCE_FLAG_TEXTURING_MORE_LIKELY

2016-10-13 Thread Nicolai Hähnle
Reviewed-by: Nicolai Hähnle On 12.10.2016 15:29, Marek Olšák wrote: From: Marek Olšák For performance tuning in drivers. It filters out window system framebuffers and OpenGL renderbuffers. radeonsi will use this to guess whether a depth buffer

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #11 from shinji.suz...@gmail.com --- Sorry I've made too many mistake in writing. (Should have gone to bed before writing.) I'll rewrite whole post below. I'm afraid execution of st_fence_sync() can still race. Thread-A can run upto

Re: [Mesa-dev] [PATCH] radeonsi: implement TC-compatible HTILE

2016-10-13 Thread Nicolai Hähnle
On 12.10.2016 15:54, Marek Olšák wrote: From: Marek Olšák so that decompress blits aren't needed and depth texturing needs less memory bandwidth. Z16 and Z24 are promoted to Z32_FLOAT by the driver, because TC-compatible HTILE only supports Z32_FLOAT. This doubles memory

Re: [Mesa-dev] [PATCH 00/15] GLSL memory allocation rework for faster compilation

2016-10-13 Thread Tapani Pälli
On 10/13/2016 04:20 PM, Juha-Pekka Heikkila wrote: I forgot to reply here on the list, I've just been talking about this with Tapani face to face. My series rebased and fixed on top of mesa master branch from yesterday is here https://github.com/juhapekka/juha_mesaexperimentals/tree/jenkins

[Mesa-dev] [Bug 98169] lm_sensors hud option crashes unigine heaven

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98169 --- Comment #30 from Steven Toth --- Christoph, based on feedback from the mailing list, I made some changes to use the MESA internal reference counting mechanisms which makes sensor and other cleanup a little more robust.

Re: [Mesa-dev] Required-for-rendering internal formats

2016-10-13 Thread Erik Faye-Lund
On Thu, Oct 13, 2016 at 5:35 PM, Nicolai Hähnle wrote: > On 13.10.2016 17:15, Ilia Mirkin wrote: >> >> Anyone else have opinions on this? Should st/mesa be adding >> PIPE_BIND_RENDER_TARGET for all textures of the internal formats >> listed in table 8.12? (Which would, in this

Re: [Mesa-dev] [PATCH 2/2] radv: Use new image load/store intrinsic signatures

2016-10-13 Thread Kai Wasserbäch
Disregard this, the mbox file only contained the second patch. Sorry for the noise. Kai Wasserbäch wrote on 13.10.2016 19:20: > Dear Tom, > just FYI: this fails to apply on top of master > (761388a0eb586b1dcaec063ee561056ed132dc1a). git am chokes on the > visit_image_store() hunk for me.

Re: [Mesa-dev] [PATCH 2/2] radeonsi: adjust and clean up Z_ORDER and EXEC_ON_x settings

2016-10-13 Thread Nicolai Hähnle
The series is Reviewed-by: Nicolai Hähnle On 12.10.2016 23:19, Marek Olšák wrote: From: Marek Olšák The table was copied from the Vulkan driver. The comment lines are as long as the table for cosmetic reasons. ---

Re: [Mesa-dev] [PATCH 1/5] st/va: Return more useful config attributes

2016-10-13 Thread Andy Furniss
Mark Thompson wrote: On 13/10/16 08:20, Christian König wrote: Am 13.10.2016 um 00:52 schrieb Mark Thompson: The encoder attributes are needed for a user of the encoder to be able to configure it sensibly without internal knowledge. Reviewed-by: Christian König for

Re: [Mesa-dev] [PATCH 2/2] radv: Use new image load/store intrinsic signatures

2016-10-13 Thread Kai Wasserbäch
Dear Tom, just FYI: this fails to apply on top of master (761388a0eb586b1dcaec063ee561056ed132dc1a). git am chokes on the visit_image_store() hunk for me. Attached is a "refreshed" version, which applies for me. I hope I didn't butcher anything inadvertently. Cheers, Kai Tom Stellard wrote on

Re: [Mesa-dev] [RFC] EGL_MESA_platform_surfaceless

2016-10-13 Thread Chad Versace
On Fri 23 Sep 2016, Chad Versace wrote: > diff --git a/docs/specs/EGL_MESA_platform_surfaceless.txt > b/docs/specs/EGL_MESA_platform_surfaceless.txt > new file mode 100644 > index 000..b700370 > --- /dev/null > +++ b/docs/specs/EGL_MESA_platform_surfaceless.txt > @@ -0,0 +1,108 @@ > +Name

[Mesa-dev] [PATCH] radeonsi: clear DB_RENDER_OVERRIDE

2016-10-13 Thread Marek Olšák
From: Marek Olšák Vulkan doesn't set these fields even though it doesn't use HiS. HiS is disabled by programming DB_SRESULTS_COMPARE_STATEn to 0. --- src/gallium/drivers/radeonsi/si_state.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git

Re: [Mesa-dev] [PATCH] clover: Allow OpenCL version override

2016-10-13 Thread Vedran Miletić
On 10/07/2016 12:05 PM, Serge Martin wrote: > On Thursday 06 October 2016 16:26:21 Vedran Miletić wrote: >> CLOVER_CL_VERSION_OVERRIDE allows overriding default OpenCL version >> supported by Clover, analogous to MESA_GL_VERSION_OVERRIDE for OpenGL. >> CLOVER_CL_C_VERSION_OVERRIDE allows overridng

Re: [Mesa-dev] [RFC] EGL_MESA_platform_surfaceless

2016-10-13 Thread Chad Versace
On Thu 13 Oct 2016, Chad Versace wrote: > On Fri 23 Sep 2016, Chad Versace wrote: > > > diff --git a/docs/specs/EGL_MESA_platform_surfaceless.txt > > b/docs/specs/EGL_MESA_platform_surfaceless.txt > > new file mode 100644 > > index 000..b700370 > > --- /dev/null > > +++

[Mesa-dev] [PATCH] radeonsi: shorten "shader->selector" to "sel" in si_shader_create

2016-10-13 Thread Marek Olšák
From: Marek Olšák --- src/gallium/drivers/radeonsi/si_shader.c | 15 --- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c index b2d7699..e6edd90 100644 ---

[Mesa-dev] [PATCH 1/2] radv: Fix incorrect comment

2016-10-13 Thread Tom Stellard
--- src/amd/common/ac_nir_to_llvm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c index e6ff7c8..9c764c7 100644 --- a/src/amd/common/ac_nir_to_llvm.c +++ b/src/amd/common/ac_nir_to_llvm.c @@ -2373,8

[Mesa-dev] [PATCH 2/2] radv: Use new image load/store intrinsic signatures v2

2016-10-13 Thread Tom Stellard
These were changed in LLVM r284024. v2: - Only use float types for vdata of llvm.amdgcn.image.store. LLVM doesn't support integer types for this intrinsic. --- src/amd/common/ac_nir_to_llvm.c | 133 1 file changed, 108 insertions(+), 25

Re: [Mesa-dev] [PATCH] clover: Allow OpenCL version override

2016-10-13 Thread Vedran Miletić
On 10/13/2016 07:45 PM, Vedran Miletić wrote: > On 10/07/2016 12:05 PM, Serge Martin wrote: >> On Thursday 06 October 2016 16:26:21 Vedran Miletić wrote: >>> PACKAGE_VERSION #ifdef MESA_GIT_SHA1 >>> " (" MESA_GIT_SHA1 ")" >>> #endif >>> diff --git

Re: [Mesa-dev] [PATCH 2/2] radeonsi: unify the constant load paths

2016-10-13 Thread Marek Olšák
For the series: Reviewed-by: Marek Olšák Marek On Thu, Oct 13, 2016 at 3:44 PM, Nicolai Hähnle wrote: > From: Nicolai Hähnle > > Remove the split between direct and indirect. > --- > src/gallium/drivers/radeonsi/si_shader.c |

[Mesa-dev] [PATCH] gallium/hud: Sensor extension is segfaulting.

2016-10-13 Thread Steven Toth
Round two of the patchset, incorporating feedback from nhaeh...@gmail.com The fixes in this set address bugfix #68169, HUD crashing when testing with unigine (heaven). The bug also manifested itself as a lack of data in HUD charts when multiple instanced were created and destroyed in a specific

Re: [Mesa-dev] [PATCH 2/2] radv: Use new image load/store intrinsic signatures

2016-10-13 Thread Tom Stellard
On Thu, Oct 13, 2016 at 07:20:30PM +0200, Kai Wasserbäch wrote: > Dear Tom, > just FYI: this fails to apply on top of master > (761388a0eb586b1dcaec063ee561056ed132dc1a). git am chokes on the > visit_image_store() hunk for me. Attached is a "refreshed" version, which > applies for me. I hope I

Re: [Mesa-dev] [PATCH 05/16] loader: reimplement loader_get_user_preferred_fd via libdrm

2016-10-13 Thread Emil Velikov
On Thursday, 13 October 2016, Emil Velikov wrote: > On 12 October 2016 at 18:59, Axel Davy > > wrote: > > On 12/10/2016 19:15, Emil Velikov wrote: > >> > >> On 11 October 2016 at 19:54, Axel Davy > > wrote:

[Mesa-dev] [PATCH v3 25/25] configure.ac: Add required LLVM versions to the top

2016-10-13 Thread Emil Velikov
On Thursday, 13 October 2016, Mike Lothian > wrote: > If you recompile llvm and mesa each time, are there any downsides of using > the shared libs? > > How many devs rebuild package X when one of its dependencies has a bug

Re: [Mesa-dev] [PATCH 5/5] st/glsl_to_tgsi: fix [ui]vec[34] conversion to double

2016-10-13 Thread Marek Olšák
For the series: Reviewed-by: Marek Olšák Marek On Thu, Oct 13, 2016 at 5:13 PM, Nicolai Hähnle wrote: > From: Nicolai Hähnle > > The corresponding opcodes for integers need to be treated the same as F2D. > > Fixes

[Mesa-dev] [PATCH 1/6] docs: Add EGL_MESA_platform_surfaceless.txt (v2)

2016-10-13 Thread Chad Versace
v2: - Assign enum values. - Define interactions with EGL_EXT_platform_base and EGL 1.4. - Add Gurchetan as contributor, as he implemented the pbuffer support. --- docs/specs/EGL_MESA_platform_surfaceless.txt | 120 +++ 1 file changed, 120 insertions(+) create

[Mesa-dev] [PATCH 0/6] EGL_MESA_platform_surfaceless (v2)

2016-10-13 Thread Chad Versace
Mesa's EGL has supported Chrome OS's "surfaceless" platform for many months, but the behavior of that platform has never been documented. Here's my attempt to fix that. I've already committed the draft extension spec into the private Khronos registry. After these patches land in master, I plan to

[Mesa-dev] [PATCH 4/6] egl: Move old EGL_MESA_screen_surface spec

2016-10-13 Thread Chad Versace
It was the lone file in src/egl/docs. Move it to where the other specs live, in $MESA_TOP/docs/specs. --- .../EGL_MESA_screen_surface => docs/specs/OLD/EGL_MESA_screen_surface.txt | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename src/egl/docs/EGL_MESA_screen_surface =>

[Mesa-dev] [PATCH 5/6] doc/specs: Reference the Khronos registry XML

2016-10-13 Thread Chad Versace
Years ago Khronos replaced the registry's spec files with newfangled XML files. Update the reference in doc/specs/enum.txt accordingly. --- docs/specs/enums.txt | 16 ++-- 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/docs/specs/enums.txt b/docs/specs/enums.txt index

[Mesa-dev] [PATCH 3/6] egl: Implement EGL_MESA_platform_surfaceless

2016-10-13 Thread Chad Versace
--- include/EGL/eglmesaext.h | 5 + src/egl/main/eglapi.c | 42 +- src/egl/main/egldisplay.c | 21 + src/egl/main/egldisplay.h | 5 + src/egl/main/eglglobals.c | 3 +++ 5 files changed, 75 insertions(+), 1 deletion(-)

[Mesa-dev] [PATCH 2/6] egl: Don't advertise unsupported platform extensions

2016-10-13 Thread Chad Versace
Mesa's set of supported platform extensions depends on the autoconf option --with-egl-platforms=foo,bar,baz. If --with-egl-platforms lacks foo, then eglGetPlatformDisplay(EGL_PLATFORM_FOO, ...) unconditonally fails. So, if --with-egl-platforms lacks foo, then remove EGL_VENDOR_platform_foo from

[Mesa-dev] [PATCH 6/6] docs/specs: Update allocated EGL enum values

2016-10-13 Thread Chad Versace
Document the EGL enum ranges for Mesa and those values allocated by the following extensions: EGL_MESA_drm_image EGL_MESA_platform_gbm EGL_MESA_platform_surfaceless EGL_WL_bind_wayland_display --- docs/specs/enums.txt | 27 ++- 1 file changed, 26

Re: [Mesa-dev] [RFC 00/12] egl, i965: Support EGL_ANDROID_native_fence_sync (v2)

2016-10-13 Thread Rob Clark
On Thu, Oct 13, 2016 at 4:37 PM, Rafael Antognolli wrote: > Hi all, > > On Tue, Oct 11, 2016 at 01:47:37PM -0400, Robert Foss wrote: >> >> >> On 2016-10-11 01:37 PM, Chad Versace wrote: >> > On Tue 11 Oct 2016, Emil Velikov wrote: >> > > On 10 October 2016 at 20:41,

Re: [Mesa-dev] [RFC 00/12] egl, i965: Support EGL_ANDROID_native_fence_sync (v2)

2016-10-13 Thread Rafael Antognolli
Hi all, On Tue, Oct 11, 2016 at 01:47:37PM -0400, Robert Foss wrote: > > > On 2016-10-11 01:37 PM, Chad Versace wrote: > > On Tue 11 Oct 2016, Emil Velikov wrote: > > > On 10 October 2016 at 20:41, Rob Clark wrote: > > > > On Mon, Oct 10, 2016 at 1:43 PM, Chad Versace > >

[Mesa-dev] [PATCH] radv/winsys: Fail early on overgrown cs.

2016-10-13 Thread Gustaw Smolarczyk
When !use_ib_bos, we can't easily chain ibs one to another. If the required cs size grows over 1Mi - 8 dwords just fail the cs so that we won't assert-fail in radv_amdgpu_winsys_cs_submit later on. --- Please, push the patch after it has been reviewed.

Re: [Mesa-dev] [RFC 00/12] egl, i965: Support EGL_ANDROID_native_fence_sync (v2)

2016-10-13 Thread Robert Foss
On 2016-10-13 04:37 PM, Rafael Antognolli wrote: Hi all, On Tue, Oct 11, 2016 at 01:47:37PM -0400, Robert Foss wrote: On 2016-10-11 01:37 PM, Chad Versace wrote: On Tue 11 Oct 2016, Emil Velikov wrote: On 10 October 2016 at 20:41, Rob Clark wrote: On Mon, Oct 10,

[Mesa-dev] [Bug 98222] [radv] LLVM ERROR: Cannot select: intrinsic %llvm.amdgcn.image.store

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98222 --- Comment #1 from Christoph Haag --- https://cgit.freedesktop.org/mesa/mesa/commit/?id=5c66d46d6ace0603227e4b919b132416444949e7 should fix it. -- You are receiving this mail because: You are the QA Contact for the bug.

[Mesa-dev] [PATCH v2] clover: Allow OpenCL version override

2016-10-13 Thread Vedran Miletić
CLOVER_CL_VERSION_OVERRIDE allows overriding default OpenCL version supported by Clover, analogous to MESA_GL_VERSION_OVERRIDE for OpenGL. CLOVER_CL_C_VERSION_OVERRIDE allows overridng default OpenCL C version. v2: - move version getters to version.hpp, simplify - set -cl-std= to

[Mesa-dev] [PATCH v3] doc/features.txt: factor out i965/hsw+ & radeonsi as GL45 complete

2016-10-13 Thread Edward O'Callaghan
V2. add i965/hsw+ to list V3. rebased on master. Signed-off-by: Edward O'Callaghan --- docs/features.txt | 18 +- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/docs/features.txt b/docs/features.txt index 0d6c16a..acea0cf 100644 ---

Re: [Mesa-dev] [PATCH] radeonsi: shorten "shader->selector" to "sel" in si_shader_create

2016-10-13 Thread Edward O'Callaghan
yes, nicer. Reviewed-by: Edward O'Callaghan On 10/14/2016 04:16 AM, Marek Olšák wrote: > From: Marek Olšák > > --- > src/gallium/drivers/radeonsi/si_shader.c | 15 --- > 1 file changed, 8 insertions(+), 7 deletions(-) > > diff

Re: [Mesa-dev] [PATCH 1/6] util: add vector util code.

2016-10-13 Thread Edward O'Callaghan
Too easy, one trivial comment below but either way: Reviewed-by: Edward O'Callaghan P.S. thanks for getting on top of this kind of maintainability stuff so fast ! On 10/14/2016 02:16 PM, Dave Airlie wrote: > From: Dave Airlie > > This is ported

[Mesa-dev] [PATCH] gallivm: print out time for jitting functions with GALLIVM_DEBUG=perf

2016-10-13 Thread sroland
From: Roland Scheidegger Compilation to actual machine code can easily take as much time as the optimization passes on the IR if not more, so print this out too. --- src/gallium/auxiliary/gallivm/lp_bld_init.c | 11 +++ 1 file changed, 11 insertions(+) diff --git

Re: [Mesa-dev] [PATCH v2] clover: Allow OpenCL version override

2016-10-13 Thread Jan Vesely
On Fri, 2016-10-14 at 02:34 +0200, Vedran Miletić wrote: > CLOVER_CL_VERSION_OVERRIDE allows overriding default OpenCL version > supported by Clover, analogous to MESA_GL_VERSION_OVERRIDE for OpenGL. > CLOVER_CL_C_VERSION_OVERRIDE allows overridng default OpenCL C version. > > v2: > - move

[Mesa-dev] [Bug 98241] dEQP GLES3.1 failure "Compute Shader should not have compiled with #version 300 es."

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98241 Bug ID: 98241 Summary: dEQP GLES3.1 failure "Compute Shader should not have compiled with #version 300 es." Product: Mesa Version: git Hardware: x86-64 (AMD64)

[Mesa-dev] [Bug 98241] dEQP GLES3.1 failure "Compute Shader should not have compiled with #version 300 es."

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98241 Randy changed: What|Removed |Added Blocks||94448 Referenced Bugs:

[Mesa-dev] [Bug 98241] dEQP GLES3.1 failure "Compute Shader should not have compiled with #version 300 es."

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98241 Tapani Pälli changed: What|Removed |Added Resolution|--- |DUPLICATE

[Mesa-dev] [Bug 98132] #version 300 es compute shaders should not be possible

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98132 Tapani Pälli changed: What|Removed |Added CC||randy...@intel.com ---

Re: [Mesa-dev] [PATCH v3 25/25] configure.ac: Add required LLVM versions to the top

2016-10-13 Thread Michel Dänzer
On 14/10/16 09:54 AM, Michel Dänzer wrote: > On 13/10/16 11:16 PM, Mike Lothian wrote: >> I think there were some issues with some programs / games that had >> clashing symbols but I'm not sure if that's still an issue > > I'm only aware of such issues with the Steam runtime, which are arguably >

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #13 from shinji.suz...@gmail.com --- > So the second fence_reference call will find so->fence == NULL and do nothing. Thank you. I failed to see that fence_reference(screen,,0) is no-op thanks to "if (ptr != refererence)" in

[Mesa-dev] [PATCH 2/6] anv: port to using new u_vector shared helper.

2016-10-13 Thread Dave Airlie
From: Dave Airlie This just removes the anv vector code and uses the new helper. Signed-off-by: Dave Airlie --- src/intel/vulkan/anv_allocator.c | 8 ++-- src/intel/vulkan/anv_batch_chain.c | 40 ++-- src/intel/vulkan/anv_private.h

[Mesa-dev] [PATCH 3/6] radv: make use of shared vector helper.

2016-10-13 Thread Dave Airlie
From: Dave Airlie This removes the vector code from radv in favour of sharing code with anv. Signed-off-by: Dave Airlie --- src/amd/vulkan/radv_private.h | 53 src/amd/vulkan/radv_util.c| 74

[Mesa-dev] [PATCH 5/6] radv: drop local MIN/MAX macros.

2016-10-13 Thread Dave Airlie
From: Dave Airlie Use the ones in macros.h instead. Signed-off-by: Dave Airlie --- src/amd/vulkan/radv_private.h | 5 + src/amd/vulkan/radv_wsi_x11.c | 2 +- 2 files changed, 2 insertions(+), 5 deletions(-) diff --git

[Mesa-dev] [PATCH 6/6] anv: drop local MIN/MAX macros.

2016-10-13 Thread Dave Airlie
From: Dave Airlie Use the ones from mesa, most places already did. Signed-off-by: Dave Airlie --- src/intel/vulkan/anv_image.c | 2 +- src/intel/vulkan/anv_private.h | 5 + 2 files changed, 2 insertions(+), 5 deletions(-) diff --git

[Mesa-dev] [PATCH 1/6] util: add vector util code.

2016-10-13 Thread Dave Airlie
From: Dave Airlie This is ported from anv, both anv and radv can share this. Signed-off-by: Dave Airlie --- src/util/Makefile.sources | 4 +- src/util/u_vector.c | 98 +++ src/util/u_vector.h | 85

[Mesa-dev] [PATCH 4/6] util: move min/max/clamp macros to util macros.h

2016-10-13 Thread Dave Airlie
From: Dave Airlie Although the vulkan drivers include mesa macros.h, for radv I'd like to move away from that. Signed-off-by: Dave Airlie --- src/mesa/main/macros.h | 13 - src/util/macros.h | 13 + 2 files changed, 13

Re: [Mesa-dev] [PATCH v3 25/25] configure.ac: Add required LLVM versions to the top

2016-10-13 Thread Michel Dänzer
On 13/10/16 07:14 PM, Emil Velikov wrote: > On 13 October 2016 at 04:07, Michel Dänzer wrote: >> On 13/10/16 03:37 AM, Tobias Droste wrote: >>> Am Mittwoch, 12. Oktober 2016, 11:53:50 CEST schrieb Emil Velikov: > > +LLVM_VERSION_REQUIRED_OPENCL=3.6.0 >

Re: [Mesa-dev] [PATCH v3 25/25] configure.ac: Add required LLVM versions to the top

2016-10-13 Thread Michel Dänzer
On 13/10/16 11:16 PM, Mike Lothian wrote: > If you recompile llvm and mesa each time, are there any downsides of > using the shared libs? No particular downsides come to mind, rather upsides: * Generally no requirement to rebuild Mesa after changes to LLVM. * Both LLVM and Mesa take

[Mesa-dev] [Bug 98172] Concurrent call to glClientWaitSync results in segfault in one of the waiters.

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98172 --- Comment #12 from Michel Dänzer --- It should be fine. You're right that fence_finish can run concurrently in multiple threads, but even if it returns true in multiple threads, the fence_reference calls are serialized by

[Mesa-dev] [PATCH] draw: improved handling of undefined inputs

2016-10-13 Thread sroland
From: Roland Scheidegger Previous attempts to zero initialize all inputs were not really optimal (though no performance impact was measurable). In fact this is not really necessary, since we know the max number of inputs used. Instead, just generate fetch for up to max inputs

[Mesa-dev] [Bug 98223] dEQP GLES3.1 program_interface_query failures w/ error "could not find target resource"

2016-10-13 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=98223 --- Comment #7 from Randy --- (In reply to Alejandro Piñeiro (freenode IRC: apinheiro) from comment #6) > > The problem is that the spec is contradictory. The patch on that link (that > btw, the author is Kenneth, that wrote

[Mesa-dev] [PATCH] docs: Fix GL status of radeonsi

2016-10-13 Thread Andreas Boll
Currently radeonsi doesn't advertise GLSL 4.40 and "GL 4.4, GLSL 4.40 -- all DONE" means the driver actually advertises GL 4.4 and GLSL 4.40. So as long as radeonsi doesn't enable GLSL >= 4.40 it's not "all DONE". Fixes: 789119d21 ("st/mesa: enable ARB_enhanced_layouts and turn the cap on")

Re: [Mesa-dev] [PATCH v2] doc/features.txt: factor out nvc0/radeonsi as GL45 complete

2016-10-13 Thread Andreas Boll
nak, neither radeonsi nor i965 advertise GLSL 4.50. Nicolai hasn't pushed the patch to enable GLSL 4.50 [1]. I'm not sure what's the plan for nouveau is [2]. See also https://patchwork.freedesktop.org/patch/115441/ Thanks, Andreas [1] https://patchwork.freedesktop.org/patch/114577/ [2]

[Mesa-dev] [PATCH] doc/features.txt: factor out nvc0/radeonsi as GL45 complete

2016-10-13 Thread Edward O'Callaghan
Signed-off-by: Edward O'Callaghan --- docs/features.txt | 18 +- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/docs/features.txt b/docs/features.txt index ec2634f..b15ec35 100644 --- a/docs/features.txt +++ b/docs/features.txt @@

Re: [Mesa-dev] [PATCH 1/2] st/mesa: Take local references for sync object fences

2016-10-13 Thread Nicolai Hähnle
On 13.10.2016 05:45, Michel Dänzer wrote: On 12/10/16 08:52 PM, Nicolai Hähnle wrote: On 12.10.2016 11:31, Michel Dänzer wrote: diff --git a/src/mesa/state_tracker/st_cb_syncobj.c b/src/mesa/state_tracker/st_cb_syncobj.c index 123925a..de01880 100644 ---

[Mesa-dev] [PATCH 1/7] nir: move nir_shader_info to a common compiler header

2016-10-13 Thread Timothy Arceri
This will allow use to stop copying values between structs and will also simplify handling handling these values in the shader cache. --- src/compiler/nir/nir.h | 98 +--- src/compiler/shader_info.h | 123

[Mesa-dev] RFC - shared shader info (Part 1)

2016-10-13 Thread Timothy Arceri
Having these fields directly in nir_shader makes things difficult for the shader cache. This series creates a shared shader info and makes a start at the large amount of clean-ups this will enable. I decided to send this before getting too carried away to make sure people are happy with this

[Mesa-dev] [PATCH 7/7] mesa/i965: eliminate gl_tess_ctrl_program and use new shared shader_info

2016-10-13 Thread Timothy Arceri
--- src/mesa/drivers/dri/i965/brw_context.h | 6 ++--- src/mesa/drivers/dri/i965/brw_draw.c | 2 +- src/mesa/drivers/dri/i965/brw_program.c | 2 +- src/mesa/drivers/dri/i965/brw_tcs.c | 32 ++-

[Mesa-dev] [PATCH 3/7] compiler: add fields for tes metadata to shader info

2016-10-13 Thread Timothy Arceri
--- src/compiler/shader_info.h | 7 +++ 1 file changed, 7 insertions(+) diff --git a/src/compiler/shader_info.h b/src/compiler/shader_info.h index 68800a1..8266fc3 100644 --- a/src/compiler/shader_info.h +++ b/src/compiler/shader_info.h @@ -117,6 +117,13 @@ typedef struct shader_info {

[Mesa-dev] [PATCH 4/7] compiler: add input primative field for gs in shader info

2016-10-13 Thread Timothy Arceri
--- src/compiler/shader_info.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/compiler/shader_info.h b/src/compiler/shader_info.h index 8266fc3..502c7c8 100644 --- a/src/compiler/shader_info.h +++ b/src/compiler/shader_info.h @@ -78,6 +78,9 @@ typedef struct shader_info {

[Mesa-dev] [PATCH 5/7] compiler: add additional cs metadata fields to shader info

2016-10-13 Thread Timothy Arceri
--- src/compiler/shader_info.h | 10 ++ 1 file changed, 10 insertions(+) diff --git a/src/compiler/shader_info.h b/src/compiler/shader_info.h index 502c7c8..66e06ad 100644 --- a/src/compiler/shader_info.h +++ b/src/compiler/shader_info.h @@ -114,6 +114,16 @@ typedef struct shader_info {

[Mesa-dev] [PATCH 6/7] glsl: add a shader info field to the gl_program type

2016-10-13 Thread Timothy Arceri
And use this field as the source for shader info in the nir_shader this will allow us to set some of these fields from GLSL directly. It will also simplify restoring from shader cache and allow the removal of duplicate fields from GLSL. --- src/compiler/glsl/glsl_to_nir.cpp | 3 ++-

Re: [Mesa-dev] Mesa (master): st/glsl_to_tgsi: simpler fixup of empty writemasks

2016-10-13 Thread Michel Dänzer
On 13/10/16 04:26 PM, Nicolai Hähnle wrote: > Hi Michel, > > On 13.10.2016 08:42, Michel Dänzer wrote: >> On 13/10/16 01:50 AM, Nicolai Hähnle wrote: >>> Module: Mesa >>> Branch: master >>> Commit: f5f3cadca3809952288e3726ed5fde22090dc61d >>> URL: >>>

[Mesa-dev] [PATCH v2] doc/features.txt: factor out nvc0/radeonsi as GL45 complete

2016-10-13 Thread Edward O'Callaghan
Signed-off-by: Edward O'Callaghan --- docs/features.txt | 18 +- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/docs/features.txt b/docs/features.txt index ec2634f..a5c8750 100644 --- a/docs/features.txt +++ b/docs/features.txt @@

Re: [Mesa-dev] [PATCH] gallium/hud: bugfix: 68169 - Sensor extensions segfaults.

2016-10-13 Thread Nicolai Hähnle
On 12.10.2016 19:02, Steven Toth wrote: https://bugs.freedesktop.org/show_bug.cgi?id=98169 Really a two part bug. 1. The recent extensions to the HUB framework were tested exclusively with glxgears/demo/head. None of these tools exercise the free_query_data() code path. When these codepaths

Re: [Mesa-dev] [PATCH 1/5] st/va: Return more useful config attributes

2016-10-13 Thread Mark Thompson
On 13/10/16 08:20, Christian König wrote: > Am 13.10.2016 um 00:52 schrieb Mark Thompson: >> The encoder attributes are needed for a user of the encoder to be >> able to configure it sensibly without internal knowledge. > > Reviewed-by: Christian König for the whole

Re: [Mesa-dev] [PATCH] gallium: add PIPE_RESOURCE_FLAG_TEXTURING_MORE_LIKELY

2016-10-13 Thread Marek Olšák
On Thu, Oct 13, 2016 at 2:09 AM, Roland Scheidegger wrote: > Am 12.10.2016 um 15:29 schrieb Marek Olšák: >> From: Marek Olšák >> >> For performance tuning in drivers. It filters out window system >> framebuffers and OpenGL renderbuffers. >> >> radeonsi

Re: [Mesa-dev] Mesa (master): st/glsl_to_tgsi: simpler fixup of empty writemasks

2016-10-13 Thread Michel Dänzer
Hi Nicolai, On 13/10/16 01:50 AM, Nicolai Hähnle wrote: > Module: Mesa > Branch: master > Commit: f5f3cadca3809952288e3726ed5fde22090dc61d > URL: > http://cgit.freedesktop.org/mesa/mesa/commit/?id=f5f3cadca3809952288e3726ed5fde22090dc61d > > Author: Nicolai Hähnle

Re: [Mesa-dev] [PATCH] mesa_glinterop: allow building without X and related headers

2016-10-13 Thread Tapani Pälli
Thanks Emil! Reviewed-by: Tapani Pälli On 10/12/2016 08:49 PM, Emil Velikov wrote: This commit effectively reverts c10dcb2ce837922c6ee4e191e6d6202098a5ee10 and fixes the typedef redefinition which inspired it. In order to prevent requiring X packages at build time

Re: [Mesa-dev] Mesa (master): st/glsl_to_tgsi: simpler fixup of empty writemasks

2016-10-13 Thread Nicolai Hähnle
Hi Michel, On 13.10.2016 08:42, Michel Dänzer wrote: On 13/10/16 01:50 AM, Nicolai Hähnle wrote: Module: Mesa Branch: master Commit: f5f3cadca3809952288e3726ed5fde22090dc61d URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=f5f3cadca3809952288e3726ed5fde22090dc61d Author: Nicolai

Re: [Mesa-dev] [PATCH] doc/features.txt: factor out nvc0/radeonsi as GL45 complete

2016-10-13 Thread Edward O'Callaghan
On 10/13/2016 06:24 PM, Nicolai Hähnle wrote: > On 13.10.2016 08:34, Edward O'Callaghan wrote: >> Signed-off-by: Edward O'Callaghan > > I think i965 can get the same treatment. And the mesamatrix people need OK, I just was not sure about 'i965/hsw+' vs. 'i965' as

Re: [Mesa-dev] [PATCH 1/5] intel: automake: replace direct basename $@ invokation with $(@F)

2016-10-13 Thread Lionel Landwerlin
Reviewed-by: Lionel Landwerlin On 12/10/16 19:05, Emil Velikov wrote: From: Emil Velikov Use the shorthand make variable(s) as elsewhere in the build. Signed-off-by: Emil Velikov ---

Re: [Mesa-dev] i965: Fast color clear for mipmapped and arrayed

2016-10-13 Thread Pohjolainen, Topi
On Tue, Oct 11, 2016 at 10:26:32PM +0300, Topi Pohjolainen wrote: > The entire series is mostly bookkeeping changes. It switches to track > color and state per slice instead of per miptree (such as we already > do for depth and hiz). > > Patch 22 is still work in progress. We have all the details

[Mesa-dev] [PATCH v2.1] i965/vec4: teach register coalescing about 64-bit

2016-10-13 Thread Iago Toral Quiroga
Specifically, at least for now, we don't want to deal with the fact that channel sizes for fp64 instructions are twice the size, so prevent coalescing from instructions with a different type size. Also, we should check that if we are coalescing a register from another MOV we should be writing the

Re: [Mesa-dev] [PATCH] doc/features.txt: factor out nvc0/radeonsi as GL45 complete

2016-10-13 Thread Nicolai Hähnle
On 13.10.2016 08:34, Edward O'Callaghan wrote: Signed-off-by: Edward O'Callaghan I think i965 can get the same treatment. And the mesamatrix people need to check their parsing code for OpenGL 4.4 ;) Cheers, Nicolai --- docs/features.txt | 18

  1   2   >