2016-10-13 1:49 GMT+08:00 Emil Velikov :
> This commit effectively reverts c10dcb2ce837922c6ee4e191e6d6202098a5ee10
> and fixes the typedef redefinition which inspired it.
>
> In order to prevent requiring X packages at build time earlier commit
> forward declared the
On 12.10.2016 20:12, Ilia Mirkin wrote:
Patches 1-3 are
Reviewed-by: Ilia Mirkin
Thanks.
Not comfortable enough with the vertex arrays code to review the last
one. I remember Dave had semi-good reasons for the if (!array) stuff
though.
As far as I understood the
Am Mittwoch, 12. Oktober 2016, 10:57:14 CEST schrieb Emil Velikov:
> On 12 October 2016 at 00:02, Tobias Droste wrote:
> > Each driver has to expllicitly call llvm_add_default_components to
> > add the shared components.
> > This way we can fail the build if a component is not
Am Mittwoch, 12. Oktober 2016, 19:30:51 CEST schrieb Emil Velikov:
> >>
> >> The function it quite "ugly" as-is and this patch changes things in a fun
> >> way.
> >>
> >> Namely: before you'll get the minimum required version check
> >> regardless of host_cpu for everyone, while now you get the
src2 was being given the wrong modifier, and we were not properly
managing the modifier on the SHL source either.
Signed-off-by: Ilia Mirkin
---
src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp | 12 +---
1 file changed, 5 insertions(+), 7 deletions(-)
On 12/10/16 19:30, Emil Velikov wrote:
On 12 October 2016 at 19:14, Tobias Droste wrote:
Am Mittwoch, 12. Oktober 2016, 10:42:19 CEST schrieb Emil Velikov:
On 12 October 2016 at 00:02, Tobias Droste wrote:
gallium_require_llvm() {
-if test
Am Mittwoch, 12. Oktober 2016, 19:38:56 CEST schrieb Emil Velikov:
> On 12 October 2016 at 19:01, Tobias Droste wrote:
> > Am Mittwoch, 12. Oktober 2016, 09:52:51 CEST schrieb Emil Velikov:
> >> On 12 October 2016 at 00:02, Tobias Droste wrote:
> >> > Add
On 10/12/2016 12:06 PM, Ilia Mirkin wrote:
Gallium is completely oblivious to whether the fbo is flipped or not.
Only flip the stipple pattern when the fbo is flipped as well. Otherwise
the driver has no idea when to unflip the pattern.
Fixes bin/gl-2.1-polygon-stipple-fs -fbo on nv50 and nvc0.
Am Mittwoch, 12. Oktober 2016, 19:39:47 CEST schrieb Jose Fonseca:
> >>
> >> Right now --enable-gallium-llvm will be set to no if it is "auto" on non
> >> x86.
> >>
> >> gallivm/llvm use the x86 target
> >
> > I haven't look too closely at the llvm code but there's nothing in
> > scons/automake
Patches 1-3 are
Reviewed-by: Ilia Mirkin
Not comfortable enough with the vertex arrays code to review the last
one. I remember Dave had semi-good reasons for the if (!array) stuff
though.
On Wed, Oct 12, 2016 at 1:50 PM, Nicolai Hähnle wrote:
> From:
Am Mittwoch, 12. Oktober 2016, 10:42:19 CEST schrieb Emil Velikov:
> On 12 October 2016 at 00:02, Tobias Droste wrote:
> >
> > gallium_require_llvm() {
> >
> > -if test "x$MESA_LLVM" = x0; then
> > -case "$host" in *gnux32) return;; esac
> > -case
From: Emil Velikov
The letter C was missing, thus in turn all the internal symbols were
exported.
As a result we hide ~150 symbols and cut ~36K from libvulkan_radeon.so.
Signed-off-by: Emil Velikov
---
My current testing involves static
From: Emil Velikov
At the moment (albeit unlikely) one could get link-time issues, since
libdrm_amdgpu.so is before it's users in the link chain.
Signed-off-by: Emil Velikov
---
As a nice bonus the diff vs the anv Makefile is quite small
From: Emil Velikov
Private/internal symbols should never not be exported. Using the
CXXFLAGS cuts ~300 exported symbols and ~23K from libvulkan_radeon.so.
Signed-off-by: Emil Velikov
---
src/amd/Makefile.addrlib.am | 3 +++
1 file
From: Emil Velikov
Will allow us to catch issues (as fixed with previous patches) rather
than release a broken tarball.
Signed-off-by: Emil Velikov
---
Makefile.am | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
From: Emil Velikov
Use the shorthand make variable(s) as elsewhere in the build.
Signed-off-by: Emil Velikov
---
src/intel/Makefile.genxml.am | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
Am Mittwoch, 12. Oktober 2016, 12:03:38 CEST schrieb Emil Velikov:
> > Does this mean that the "default" components are required only by the
> > gallivm module ? Please rename the function to reflect that.
> >
> > Don't recall if swr driver is/was using any of it, but the nv30 path
> > of nouveau
On 12 October 2016 at 19:14, Tobias Droste wrote:
> Am Mittwoch, 12. Oktober 2016, 10:42:19 CEST schrieb Emil Velikov:
>> On 12 October 2016 at 00:02, Tobias Droste wrote:
>> >
>> > gallium_require_llvm() {
>> >
>> > -if test "x$MESA_LLVM" = x0; then
>> > -
Am Mittwoch, 12. Oktober 2016, 11:53:50 CEST schrieb Emil Velikov:
> >
> > +dnl LLVM versions
> > +LLVM_VERSION_REQUIRED_GALLIUM=3.3.0
> > +LLVM_VERSION_REQUIRED_LLVMPIPE=3.6.0
>
> Not exactly sure why/how llvmpipe got 3.6 since the driver reuses the
> "gallium" one.
Yes you're right. I'm going
On 12 October 2016 at 19:01, Tobias Droste wrote:
> Am Mittwoch, 12. Oktober 2016, 09:52:51 CEST schrieb Emil Velikov:
>> On 12 October 2016 at 00:02, Tobias Droste wrote:
>> > Add functions to add and check targets/components.
>> > Not used in this patch.
>> >
>>
On 12 October 2016 at 19:04, Tobias Droste wrote:
> Am Mittwoch, 12. Oktober 2016, 09:56:39 CEST schrieb Emil Velikov:
>> > fi
>> >
>> > +if test "x$enable_opencl" = xyes; then
>> > +llvm_check_version_for "3" "6" "0" "opencl"
>> > +
>> > +
On Wed, Oct 12, 2016 at 2:46 PM, Brian Paul wrote:
> On 10/12/2016 12:06 PM, Ilia Mirkin wrote:
>>
>> Gallium is completely oblivious to whether the fbo is flipped or not.
>> Only flip the stipple pattern when the fbo is flipped as well. Otherwise
>> the driver has no idea when
On 12 October 2016 at 19:44, Tobias Droste wrote:
> Am Mittwoch, 12. Oktober 2016, 19:38:56 CEST schrieb Emil Velikov:
>> On 12 October 2016 at 19:01, Tobias Droste wrote:
>> > Am Mittwoch, 12. Oktober 2016, 09:52:51 CEST schrieb Emil Velikov:
>> >> On 12 October
Am Mittwoch, 12. Oktober 2016, 20:21:29 CEST schrieb Emil Velikov:
> >>
> >> If things error out here, that means that one should finish the other
> >> refactoring before using these helpers ?
> >> If that's getting too hairy, just add keep the function a simple
> >> wrapper around
On 12 October 2016 at 19:15, Chih-Wei Huang wrote:
> 2016-10-13 1:49 GMT+08:00 Emil Velikov :
>> This commit effectively reverts c10dcb2ce837922c6ee4e191e6d6202098a5ee10
>> and fixes the typedef redefinition which inspired it.
>>
>> In order to
On 12 October 2016 at 20:28, Bas Nieuwenhuizen wrote:
> On Wed, Oct 12, 2016 at 7:05 PM, Emil Velikov
> wrote:
>> From: Emil Velikov
>>
>> Private/internal symbols should never not be exported. Using the
>>
I'm using Mesa in a "superbuild" scenario where CFLAGS and CXXFLAGS are set
by a top level project that builds many subprojects. They typical settings
for release builds with this is adding O3 to CFLAGS and CXXFLAGS, which in
turn cause Mesa to build with O3. I haven't noticed any runtime issues
On 12 October 2016 at 19:58, Tobias Droste wrote:
> Am Mittwoch, 12. Oktober 2016, 19:51:21 CEST schrieb Emil Velikov:
>> On 12 October 2016 at 19:04, Tobias Droste wrote:
>> > Am Mittwoch, 12. Oktober 2016, 09:56:39 CEST schrieb Emil Velikov:
>> >> > fi
>> >> >
On Wed, Oct 12, 2016 at 7:05 PM, Emil Velikov wrote:
> From: Emil Velikov
>
> Private/internal symbols should never not be exported. Using the
> CXXFLAGS cuts ~300 exported symbols and ~23K from libvulkan_radeon.so.
Double negative? I think
Am Mittwoch, 12. Oktober 2016, 20:28:03 CEST schrieb Emil Velikov:
> On 12 October 2016 at 19:58, Tobias Droste wrote:
> > Am Mittwoch, 12. Oktober 2016, 19:51:21 CEST schrieb Emil Velikov:
> >> On 12 October 2016 at 19:04, Tobias Droste wrote:
> >> > Am Mittwoch,
On Wed, Oct 12, 2016 at 12:40 PM, Chuck Atkins wrote:
> I'm using Mesa in a "superbuild" scenario where CFLAGS and CXXFLAGS are set
> by a top level project that builds many subprojects. They typical settings
> for release builds with this is adding O3 to CFLAGS and
Fixes: b3bd8bb611bb465d2e5e ("swr: [rasterizer core] add support
for "RAW" surface format")
CovID: 1373647
Signed-off-by: Eric Engestrom
---
src/gallium/drivers/swr/rasterizer/common/formats.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
On Wed, Oct 12, 2016 at 8:52 PM, Ilia Mirkin wrote:
> On Wed, Oct 12, 2016 at 2:46 PM, Brian Paul wrote:
>> On 10/12/2016 12:06 PM, Ilia Mirkin wrote:
>>>
>>> Gallium is completely oblivious to whether the fbo is flipped or not.
>>> Only flip the stipple
Reviewed-by: Marek Olšák
Marek
On Wed, Oct 12, 2016 at 8:05 PM, Emil Velikov wrote:
> From: Emil Velikov
>
> Private/internal symbols should never not be exported. Using the
> CXXFLAGS cuts ~300 exported symbols and
Reviewed-by: Marek Olšák
Marek
On Wed, Oct 12, 2016 at 7:42 PM, Nicolai Hähnle wrote:
> From: Nicolai Hähnle
>
> Fixes GL45-CTS.shader_image_load_store.basic-allTargets-atomic*
> ---
> src/gallium/drivers/radeonsi/si_shader.c
For 1-3:
Reviewed-by: Marek Olšák
Marek
On Wed, Oct 12, 2016 at 7:50 PM, Nicolai Hähnle wrote:
> From: Nicolai Hähnle
>
> ---
> src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 6 +++---
> 1 file changed, 3 insertions(+), 3
From: Marek Olšák
The table was copied from the Vulkan driver. The comment lines are as long
as the table for cosmetic reasons.
---
src/gallium/drivers/radeonsi/si_shader.h| 1 -
src/gallium/drivers/radeonsi/si_state_shaders.c | 53 +++--
2
From: Marek Olšák
This is a serious performance fix. Discovered by luck.
Cc: 12.0
---
src/gallium/drivers/radeonsi/si_state_shaders.c | 11 ---
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git
On Wed, Oct 12, 2016 at 10:49 AM, 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 earlier commit
> forward
On Tue, Oct 11, 2016 at 12:26 PM, Topi Pohjolainen
wrote:
> From: Ben Widawsky
>
> Signed-off-by: Ben Widawsky
> ---
> src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 6 +++---
> 1 file changed, 3 insertions(+), 3
https://bugs.freedesktop.org/show_bug.cgi?id=96770
--- Comment #5 from Mauro Rossi ---
Hi,
I need to report that Android build has been broken by Vinson's commit
mentioned at Comment #3.
Mauro
In file included from
https://bugs.freedesktop.org/show_bug.cgi?id=97260
--- Comment #62 from Michel Dänzer ---
(In reply to alvarex from comment #60)
> I tried bisecting mesa but I failed sometimes it won't compile sorry I'm not
> good with git.
Just run "git bisect skip" for commits which
Unfortunately, this introduces some regressions:
spec/arb_gpu_shader5/texturegatheroffset/fs-r-none-shadow-2d: fail
spec/arb_gpu_shader5/texturegatheroffset/fs-r-none-shadow-2darray: fail
spec/arb_gpu_shader5/texturegatheroffset/fs-r-none-shadow-2drect: fail
On Wed, Oct 12, 2016 at 12:11:01PM +0300, Pohjolainen, Topi wrote:
> On Fri, Oct 07, 2016 at 09:41:13PM -0700, Jason Ekstrand wrote:
> > Signed-off-by: Jason Ekstrand
> > ---
> > src/intel/vulkan/anv_blorp.c | 60
> > src/intel/vulkan/anv_meta_clear.c
On Tuesday, 11 October 2016 20:10:27 CEST Brian Paul wrote:
> Rename to gl_vertex_attrib_array::BufferBindingIndex because this field
> is an index into the array of buffer binding points. This makes some
> code a little easier to follow since there's also a "VertexBinding" field
> in
Hi,
On Tuesday, 11 October 2016 20:10:26 CEST Brian Paul wrote:
> Use 'vao' instead of 'obj' to be consistent with other code.
> Plus, add a comment.
One question in patch #2.
Either way, the series is:
Reviewed-by: Mathias Fröhlich
> ---
>
On 12.10.2016 01:49, Tom Stellard wrote:
On Tue, Oct 11, 2016 at 03:21:24PM +0200, Nicolai Hähnle wrote:
On 11.10.2016 07:36, Dave Airlie wrote:
On 11 October 2016 at 12:13, Dave Airlie wrote:
On 11 October 2016 at 11:42, Dave Airlie wrote:
On 11
That's a nice improvement. For the series:
Reviewed-by: Nicolai Hähnle
On 11.10.2016 16:48, Marek Olšák wrote:
From: Marek Olšák
---
src/gallium/drivers/radeonsi/si_compute.c| 2 +-
src/gallium/drivers/radeonsi/si_state.c | 12
For the series:
Reviewed-by: Nicolai Hähnle
On 11.10.2016 17:00, Marek Olšák wrote:
From: Marek Olšák
---
src/gallium/winsys/amdgpu/drm/amdgpu_cs.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git
https://bugs.freedesktop.org/show_bug.cgi?id=98172
--- Comment #5 from shinji.suz...@gmail.com ---
Hello Michael-san.
I'm afraid my testing was not enough and I spoke too soon. Today I've run my
app remotely through ssh session by pointing DISPLAY to :0 as I'm away from my
home machine. Now the
On 12 October 2016 at 00:02, Tobias Droste wrote:
> This way LLVM_CONFIG can bet set from an env variable if it's outside
> the $llvm_prefix.
>
> This is not a must, but it helps testing.
>
Indeed. Nicely done !
> Signed-off-by: Tobias Droste
> ---
>
On 12 October 2016 at 00:02, Tobias Droste wrote:
> This is handled by "llvm_check_version_for" a few lines below.
>
> Signed-off-by: Tobias Droste
Reviewed-by: Emil Velikov
-Emil
___
On 12 October 2016 at 00:02, Tobias Droste wrote:
> The openCL checks don't need to be inside the LLVM version check.
> "llvm_check_version_for" also works if LLVM wasn't found.
>
> Signed-off-by: Tobias Droste
> ---
> configure.ac | 33
On Fri, Oct 07, 2016 at 09:41:13PM -0700, Jason Ekstrand wrote:
> Signed-off-by: Jason Ekstrand
> ---
> src/intel/vulkan/anv_blorp.c | 60
> src/intel/vulkan/anv_meta_clear.c | 141
> --
> 2 files changed, 60
Reviewed-by: Marek Olšák
On Oct 12, 2016 4:08 AM, "Brian Paul" wrote:
> Use a bitmask to indicate which color buffers are integer-valued, rather
> than a bool. Also, the old field was mis-computed. If an integer buffer
> was followed by a non-integer
If pPhysicalDevices is too small for all physical devices,
the driver must return VK_INCOMPLETE. Since only a single
physical device is supported, this is only the case when
pPhysicalDeviceCount == 0 && pPhysicalDevices != NULL.
---
src/amd/vulkan/radv_device.c | 2 ++
1 file changed, 2
Hi Nicolas,
On 6 October 2016 at 20:21, Nicolas Koch wrote:
> If pPhysicalDevices is too small for all physical devices,
> the driver must return VK_INCOMPLETE.
> Since only a single physical device is supported, this is only the case
> when pPhysicalDeviceCount == 0 &&
On 12 October 2016 at 00:02, Tobias Droste wrote:
> The gallium checks don't need to be inside the LLVM version check.
> If "enable-gallium-llvm" ist set this is called after the LLVM version check.
>
> Signed-off-by: Tobias Droste
> ---
> configure.ac | 18
On 12 October 2016 at 00:02, Tobias Droste wrote:
> A function with the LLVM version checked is moved to the top.
> The function is called where the old code was.
Replace the second line with "... in order to reuse/rework X"
> No functional change.
>
> Signed-off-by: Tobias
Sounds reasonable.
Reviewed-by: Samuel Pitoiset
On 10/12/2016 02:51 AM, Ilia Mirkin wrote:
The offset needs to be properly copied over to the phi value, otherwise
it will get assigned to the base of the merge instead of the proper
location.
Signed-off-by: Ilia
On Fri, Oct 07, 2016 at 09:41:18PM -0700, Jason Ekstrand wrote:
> Now that we no longer have meta, all pipelines get created via the normal
> Vulkan pipeline creation mechanics. There is no more need for this bit of
> extra magic data that we've been passing around.
>
> Signed-off-by: Jason
On 12.10.2016 11:31, Michel Dänzer wrote:
From: Michel Dänzer
Fixes a race condition:
Thread AThread B
Test if so->fence != NULL
=> true
Set so->fence = NULL
Dereference
On Fri, Oct 07, 2016 at 09:41:15PM -0700, Jason Ekstrand wrote:
> Signed-off-by: Jason Ekstrand
> ---
> src/intel/vulkan/anv_blorp.c | 113
> ++
> src/intel/vulkan/anv_meta_clear.c | 24
> 2 files changed, 113
On Fri, Oct 07, 2016 at 09:41:16PM -0700, Jason Ekstrand wrote:
> Signed-off-by: Jason Ekstrand
> ---
> src/intel/vulkan/anv_blorp.c | 81 +++
> src/intel/vulkan/anv_meta_clear.c | 298
> --
> 2 files changed, 81
On 12 October 2016 at 00:02, Tobias Droste wrote:
> Each driver has to expllicitly call llvm_add_default_components to
> add the shared components.
> This way we can fail the build if a component is not found and avoid
> the recursive solution from a previous version of the pach
Specifically, the point is that size_t is for CPU sizes, while this code
is about GPU sizes. The patch is
Reviewed-by: Nicolai Hähnle
On 11.10.2016 19:57, Axel Davy wrote:
size_t could be uint32_t when compiling 32 bits.
However that would lead to overflow, which
https://bugs.freedesktop.org/show_bug.cgi?id=98172
--- Comment #6 from shinji.suz...@gmail.com ---
Created attachment 127237
--> https://bugs.freedesktop.org/attachment.cgi?id=127237=edit
Diff that went into my copy of mesa-source
--
You are receiving this mail because:
You are the QA Contact
Reviewed-by: Nicolai Hähnle
On 12.10.2016 11:31, Michel Dänzer wrote:
From: Michel Dänzer
Duplicate code cleanup, no functional change intended.
Signed-off-by: Michel Dänzer
---
src/mesa/state_tracker/st_cb_syncobj.c
https://bugs.freedesktop.org/show_bug.cgi?id=97260
--- Comment #63 from alvarex ---
If I remove the line "adev->ddev->mode_config.async_page_flip = true;" from the
files on comment 17, and boot with radeon.dpm=0 radeon.runpm=0. That will
trigger the bug, on
On 12 October 2016 at 00:02, Tobias Droste wrote:
> Consolidate the required LLVM versions at the top where the other
> versions for dependencies are listed.
> Rework the LLVM_VERSION and LLVM_VERSION_INT calculation to make the
> format "x.y.z." possible.
>
Not sure how others
From: Michel Dänzer
Duplicate code cleanup, no functional change intended.
Signed-off-by: Michel Dänzer
---
src/mesa/state_tracker/st_cb_syncobj.c | 28 +---
1 file changed, 5 insertions(+), 23 deletions(-)
diff --git
From: Michel Dänzer
Fixes a race condition:
Thread AThread B
Test if so->fence != NULL
=> true
Set so->fence = NULL
Dereference so->fence
=> NULL dereference
Also, taking a
On 12 October 2016 at 00:02, Tobias Droste wrote:
> There's already a function for gallium LLVM checks.
> Move the remaining code to this function and call it for every driver
> that uses LLVM.
>
> Signed-off-by: Tobias Droste
> ---
> configure.ac | 31
On Wednesday, 24 August 2016 08:32:10 CEST mathias.froehl...@gmx.net wrote:
> From: Mathias Fröhlich
>
> Hi all,
>
> kind of a ping with a rephrased commit message
> and the style change.
> Please review.
Ping.
best
Mathias
>
> Thanks!
>
> Mathias
>
>
> When
On 12 October 2016 at 00:02, Tobias Droste wrote:
> Add functions to add and check targets/components.
> Not used in this patch.
>
> The error message in llvm_add_component is disabled until it doesn't break
> the build anymore.
> This is the same functionality as before where
On 12 October 2016 at 00:02, Tobias Droste wrote:
> Signed-off-by: Tobias Droste
> ---
> configure.ac | 6 +-
> 1 file changed, 1 insertion(+), 5 deletions(-)
>
> diff --git a/configure.ac b/configure.ac
> index 69421ff..e48ed57 100644
> --- a/configure.ac
>
On Wed, Oct 12, 2016 at 10:32 AM, Emil Velikov wrote:
> Hi Nicolas,
>
> On 6 October 2016 at 20:21, Nicolas Koch wrote:
>> If pPhysicalDevices is too small for all physical devices,
>> the driver must return VK_INCOMPLETE.
>> Since only a single
https://bugs.freedesktop.org/show_bug.cgi?id=98172
--- Comment #4 from Michel Dänzer ---
Shinji-san,
(In reply to shinji.suzuki from comment #3)
> Yes, it does! Thank you for your interest and effort in solving the issue.
Great, thank you for testing the patch.
> My itch
On 12 October 2016 at 10:57, Emil Velikov wrote:
> On 12 October 2016 at 00:02, Tobias Droste wrote:
>> Each driver has to expllicitly call llvm_add_default_components to
>> add the shared components.
>> This way we can fail the build if a component is
Recent fix for non-const offsets broke the case of a single offset (vs 4
offsets). The later code relies on the offs array to contain null values
to tell whether they should be added onto the srcs list.
Fixes: 5239bd592 ("nvc0/ir: fix overwriting of value backing non-constant
gather offset")
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 will be read
by a shader. There is no guarantee about what will actually happen.
This is a
https://bugs.freedesktop.org/show_bug.cgi?id=98169
Steven Toth changed:
What|Removed |Added
Attachment #127217|0 |1
is
On 10/12/2016 01:43 AM, Mathias Fröhlich wrote:
On Tuesday, 11 October 2016 20:10:27 CEST Brian Paul wrote:
Rename to gl_vertex_attrib_array::BufferBindingIndex because this field
is an index into the array of buffer binding points. This makes some
code a little easier to follow since there's
https://bugs.freedesktop.org/show_bug.cgi?id=97260
--- Comment #64 from Alex Deucher ---
(In reply to alvarex from comment #63)
> If I remove the line "adev->ddev->mode_config.async_page_flip = true;" from
> the files on comment 17, and boot with radeon.dpm=0
https://bugs.freedesktop.org/show_bug.cgi?id=98169
--- Comment #25 from Steven Toth ---
Christoph,
Please remove the prior sensor-cleanup.patch, apply the attached
sensor-cleanupv2.patch.
This should prevent freeing of an object while its being used by a second hud
On 11/10/16 23:04, srol...@vmware.com wrote:
From: Roland Scheidegger
This should make the code more robust if a shader tries to use inputs which
aren't defined by the vertex element layout (which usually shouldn't happen).
No piglit change.
---
On 11.10.2016 23:45, Tom Stellard wrote:
---
src/gallium/drivers/radeonsi/si_shader.c | 59 +---
1 file changed, 46 insertions(+), 13 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_shader.c
b/src/gallium/drivers/radeonsi/si_shader.c
index
yep!
Reviewed-by: Edward O'Callaghan
On 10/12/2016 10:04 PM, Nicolai Hähnle wrote:
> That's a nice improvement. For the series:
>
> Reviewed-by: Nicolai Hähnle
>
> On 11.10.2016 16:48, Marek Olšák wrote:
>> From: Marek Olšák
On 08.10.2016 12:58, Marek Olšák wrote:
From: Marek Olšák
---
src/compiler/glsl/glcpp/glcpp-lex.l | 2 +-
src/compiler/glsl/glcpp/glcpp-parse.y | 203 +++---
src/compiler/glsl/glcpp/glcpp.h | 1 +
3 files changed, 94 insertions(+),
On Wed, Oct 12, 2016 at 3:11 PM, Nicolai Hähnle wrote:
> On 08.10.2016 12:58, Marek Olšák wrote:
>>
>> From: Marek Olšák
>>
>> ---
>> src/compiler/glsl/glcpp/glcpp-lex.l | 2 +-
>> src/compiler/glsl/glcpp/glcpp-parse.y | 203
>>
Nice work! I sent some comments on patches 6 & 7. Patches 1-4 and 7-15 are
Reviewed-by: Nicolai Hähnle
assuming you checked them with a Piglit run in addition to the shader-db.
Something that LLVM does for its intermediate representations is using
Recycler objects.
On 08.10.2016 12:58, Marek Olšák wrote:
From: Marek Olšák
It would be useful to hook this up with Valgrind and address sanitizer.
Some drivers already do that; basically, there are macros
VALGRIND_MAKE_MEM_{NOACCESS, UNDEFINED, DEFINED} in valgrind/memcheck.h,
and
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 footprint for Z16.
The format promotion is
On 13 October 2016 at 03:42, Nicolai Hähnle wrote:
> From: Nicolai Hähnle
>
> Fixes GL45-CTS.shader_image_load_store.basic-allTargets-atomic*
Reviewed-by: Dave Airlie
> ---
> src/gallium/drivers/radeonsi/si_shader.c | 9
On Mon, Oct 10, 2016 at 9:19 PM, Gustaw Smolarczyk wrote:
> 2016-10-10 22:04 GMT+02:00 Bas Nieuwenhuizen :
>> Hi Gustaw,
>>
>> The patch is
>> reviewed-by: Bas Nieuwenhuizen
>
> Thanks, please push it since I don't have
Hi Marek
This and the patch before are
Tested-by: Edmondo Tommasina
Tested with:
* The Witcher 2
* Two Worlds
* The Talos Principle
No regressions. Performance seems similar. I'm quite CPU limited
(Phenom II X3 720) so I don't expect a measurable performance
On 13 October 2016 at 03:50, Nicolai Hähnle wrote:
> Hi all,
>
> this is admittedly a pretty random series, but that's what tends to happen
> when you go through test suite failures. Please review!
All look good to me, though it would be good to make sure llvmpipe doesn't
Reviewed-by: Marek Olšák
Marek
On Wed, Oct 12, 2016 at 4:08 AM, Brian Paul wrote:
> None of the drivers which implement this hook do anything with the
> texture parameter value. Drivers just look at the pname and set a
> dirty flag if needed.
>
> We
On Tue 11 Oct 2016, Emil Velikov wrote:
> From: Emil Velikov
>
> Signed-off-by: Emil Velikov
> ---
> src/intel/isl/isl.c | 4 ++--
> src/intel/isl/isl.h | 10 --
> 2 files changed, 2 insertions(+), 12 deletions(-)
Patches 1-5 are
https://bugs.freedesktop.org/show_bug.cgi?id=98169
--- Comment #26 from Christoph Haag ---
Created attachment 127250
--> https://bugs.freedesktop.org/attachment.cgi?id=127250=edit
success
With the patches
https://bugs.freedesktop.org/show_bug.cgi?id=98169#c19
1 - 100 of 168 matches
Mail list logo