LGTM too
On April 21, 2017 5:29:33 PM Timothy Arceri wrote:
I don't entirely understand how this works, but it seems reasonable to me.
Acked-by: Timothy Arceri
Thanks for fixing this :)
On 21/04/17 19:13, Kenneth Graunke wrote:
opt_register_coalesce() was optimizing sequences such as:
On Thu, Apr 20, 2017 at 09:55:56AM -0700, Kristian H. Kristensen wrote:
> Rafael Antognolli writes:
>
> > Emit 3DSTATE_VS on Gen6+ using brw_batch_emit helper, that uses pack
> > structs from genxml.
> >
> > Signed-off-by: Rafael Antognolli
> > ---
> > src/mesa/drivers/dri/i965/Makefile.sources
I don't entirely understand how this works, but it seems reasonable to me.
Acked-by: Timothy Arceri
Thanks for fixing this :)
On 21/04/17 19:13, Kenneth Graunke wrote:
opt_register_coalesce() was optimizing sequences such as:
mul(8) acc0:D, attr18.xyyy:D, attr19.xyyy:D
mach(8) vgrf5.
https://bugs.freedesktop.org/show_bug.cgi?id=97524
Timothy Arceri changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
On 21 April 2017 at 20:01, Manolova, Plamena wrote:
> Looks good to me :)
>
> Reviewed-by: Plamena Manolova
>
... and pushed to master.
Thanks Elie, Plamena,
-Emil
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/m
On 21/04/17 20:51, Emil Velikov wrote:
On 21 April 2017 at 02:11, Timothy Arceri wrote:
Thanks for doing this. Looks like you forgot to updated the docs, with that
fixed.
Not sure what you mean here - are you thinking of a note to
docs/release/17.2.0.html?
Never mind I thought this was in e
https://bugs.freedesktop.org/show_bug.cgi?id=100613
--- Comment #5 from Roland Scheidegger ---
I suppose this patch m(In reply to Ray Strode [halfline] from comment #4)
> Created attachment 130977 [details] [review]
> patch to aid in troubleshooting
I suppose this makes sense. Sort of. I was mis
https://bugs.freedesktop.org/show_bug.cgi?id=100613
Ray Strode [halfline] changed:
What|Removed |Added
CC||rstr...@redhat.com
--- Comment
Both are
Reviewed-by: Jason Ekstrand
On Fri, Apr 21, 2017 at 12:05 PM, Manolova, Plamena <
plamena.manol...@intel.com> wrote:
> Hi Curro,
> This series looks good to me :)
>
> Reviewed-by: Plamena Manolova
>
> On Fri, Apr 21, 2017 at 11:49 AM, Francisco Jerez
> wrote:
>
>> Until now the spill
Hi Curro,
This series looks good to me :)
Reviewed-by: Plamena Manolova
On Fri, Apr 21, 2017 at 11:49 AM, Francisco Jerez
wrote:
> Until now the spilling cost calculation was neglecting the amount of
> data read from the register during the spilling cost calculation.
> This caused it to make s
Looks good to me :)
Reviewed-by: Plamena Manolova
On Fri, Apr 21, 2017 at 9:20 AM, Elie Tournier
wrote:
> Since commit ce562f9e3fa, two new files are generated.
> We don't want to track them.
>
> Signed-off-by: Elie Tournier
> ---
> src/egl/.gitignore | 2 ++
> 1 file changed, 2 insertions(+
This is what we use later on to compute the number of registers that
will actually get spilled to memory, so it's more likely to match
reality than the current open-coded approximation.
Cc:
---
src/intel/compiler/brw_fs_reg_allocate.cpp | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
di
Until now the spilling cost calculation was neglecting the amount of
data read from the register during the spilling cost calculation.
This caused it to make suboptimal decisions in some cases leading to
higher memory bandwidth usage than necessary.
Improves Unigine Heaven performance by ~4% on BD
https://bugs.freedesktop.org/show_bug.cgi?id=90264
Juan Pablo Ugarte changed:
What|Removed |Added
CC||juanpablouga...@gmail.com
--
You ar
On 04/21/2017 05:02 AM, Emil Velikov wrote:
On 21 April 2017 at 07:18, Jordan Justen wrote:
On 2017-04-20 18:49:57, Timothy Arceri wrote:
On 21/04/17 11:37, Jordan Justen wrote:
On 2017-04-20 12:33:45, Samuel Pitoiset wrote:
I have used it sometimes, but since VERBOSE_API is missing in a bun
Samuel Iglesias Gonsálvez writes:
> On Thu, 2017-04-20 at 10:26 -0700, Francisco Jerez wrote:
>> Samuel Iglesias Gonsálvez writes:
>>
>> > It was setting XYWZ swizzle to all uniforms, no matter if they were
>> > a vector or not.
>> >
>> > Signed-off-by: Samuel Iglesias Gonsálvez
>> > Cc: curr
Could do smarter stuff with the mappings. I decided not to.
Tested with kmscube. No current clients seem to use non-zero x0, y0, so
that's untested.
Cc: Tapani Pälli (AndroidIA?)
Cc: Emil Velikov
Signed-off-by: Ben Widawsky
---
src/mesa/drivers/dri/i965/intel_screen.c | 39 +++
On 2017-04-21 04:02:11, Emil Velikov wrote:
> On 21 April 2017 at 07:18, Jordan Justen wrote:
> > On 2017-04-20 18:49:57, Timothy Arceri wrote:
> >> On 21/04/17 11:37, Jordan Justen wrote:
> >> > On 2017-04-20 12:33:45, Samuel Pitoiset wrote:
> >> >> I have used it sometimes, but since VERBOSE_API
On 20/04/17 10:39, Emil Velikov wrote:
On 20 April 2017 at 00:36, Timothy Arceri wrote:
On 19/04/17 20:32, Emil Velikov wrote:
On 19 April 2017 at 08:45, Timothy Arceri wrote:
This reverts commit 458c7490c29ef2960a33a089f65490e044da5d27.
The commit did not revert cleanly so this was fixed
Since commit ce562f9e3fa, two new files are generated.
We don't want to track them.
Signed-off-by: Elie Tournier
---
src/egl/.gitignore | 2 ++
1 file changed, 2 insertions(+)
create mode 100644 src/egl/.gitignore
diff --git a/src/egl/.gitignore b/src/egl/.gitignore
new file mode 100644
index
I reread http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1570.pdf and
is not explicit.
But I checked C11 threads.h implementation and thrd_current also uses
DuplicateHandle.
So C11/threads_win32.h is correct. And ISO C11 standard commited
screwed up here. Maybe they'll introduce thread id
https://bugs.freedesktop.org/show_bug.cgi?id=100708
--- Comment #7 from Nikita Krupenko ---
I built and installed packages with newer libdrm.
Checked on master branch - same behaviour. And I still see either message like
this one:
XIO: fatal IO error 11 (Resource temporarily unavailable) on X
Hi Emil,
On 21 April 2017 at 14:43, Emil Velikov wrote:
> On 21 April 2017 at 13:26, Daniel Stone wrote:
>> On 20 April 2017 at 16:20, wrote:
>>> Trying to build the Vulkan WSI code results in a failure without DRI3.
>>> Make configure check that DRI3 is available before trying to build
>>> th
rasterizer.cpp takes a lot of resources to compile. Looks like the autotools
build runs -j2, which limits the exposure, because of fewer parallel jobs,
while scons, by default is more aggressive.
Suggest running scons -j 1 or scons -j 2 for swr.
Thanks,
George
> On Apr 21, 2017, at 7:08 AM,
No, I'm afraid the problem is still there AFAICS.
GetCurrentThread returns a magic constant (-2 if I'm not mistaken). So
it is (and always will be) useless for thrd_current.
Imagine that:
- thread #1 invokes thrd_current() and gets -2
- thread #2 invokes thrd_current() and gets -2
- a thread
On Fri, Apr 21, 2017 at 4:35 PM, Mauro Rossi wrote:
>
>
> 2017-04-20 18:51 GMT+02:00 Mauro Rossi :
>>
>>
>>
>> 2017-04-20 4:18 GMT+02:00 Tomasz Figa :
>>> I'd also like to hear from Mauro if this version, combined with the
>>> patch to use cancelBuffer instead of queueBuffer for destroy surface,
>
On 21 April 2017 at 13:26, Daniel Stone wrote:
> Hi Ville,
>
> On 20 April 2017 at 16:20, wrote:
>> Trying to build the Vulkan WSI code results in a failure without DRI3.
>> Make configure check that DRI3 is available before trying to build
>> the Vulkan drivers.
>
> It shouldn't. Could you plea
Use a lower make j number and make sure you're not using LTO
On Fri, 21 Apr 2017, 2:33 pm Emil Velikov, wrote:
> On 21 April 2017 at 13:25, Gustaw Smolarczyk wrote:
> > 2017-04-21 14:08 GMT+02:00 Emil Velikov :
> >> Note: GCC 5/6 crashes with internal compiler error at seemingly random
> >> sta
On 20 April 2017 at 22:50, Vinson Lee wrote:
> Signed-off-by: Vinson Lee
Cc:
Reviewed-by: Emil Velikov
-Emil
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
On 21 April 2017 at 13:31, Christian König wrote:
> Am 21.04.2017 um 14:11 schrieb Emil Velikov:
>>
>> Both headers are used everywhere, plus they both provide interop
>> mechanism.
>>
>> Since they define [consecutive] VDPAU driver funcs, we really want the
>> definitions side by side. Otherwise
On 21 April 2017 at 13:29, Christian König wrote:
> Am 21.04.2017 um 14:11 schrieb Emil Velikov:
>>
>> Split the helper in two:
>> - an "official" VDP format to pipe
>> - and a custom VDP format to pipe
>>
>> The latter is DMAbuf specific and does not belong with the rest.
>> That is, until no
On 21 April 2017 at 13:25, Gustaw Smolarczyk wrote:
> 2017-04-21 14:08 GMT+02:00 Emil Velikov :
>> Note: GCC 5/6 crashes with internal compiler error at seemingly random
>> stages of the build. Unless any of the SWR devs sort it out, we might
>> want to disable it for now.
>>
>> A sample report ca
On 21 April 2017 at 14:24, Mike Lothian wrote:
> Is nine and clover built with this?
>
Yes they are. Since building clover is slow we have separate "ST
Clover" and "ST Other" targets.
-Emil
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https:
Is nine and clover built with this?
On Fri, 21 Apr 2017, 1:25 pm Gustaw Smolarczyk,
wrote:
> 2017-04-21 14:08 GMT+02:00 Emil Velikov :
> > Note: GCC 5/6 crashes with internal compiler error at seemingly random
> > stages of the build. Unless any of the SWR devs sort it out, we might
> > want to
On 20 April 2017 at 18:29, Brian Paul wrote:
> In the subject line "compare"
>
Fixed and pushed.
Thanks for the confirmation everybody!
Emil
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Am 21.04.2017 um 14:11 schrieb Emil Velikov:
Both headers are used everywhere, plus they both provide interop
mechanism.
Since they define [consecutive] VDPAU driver funcs, we really want the
definitions side by side. Otherwise we risk clashing multiple functions
at the same address.
Cc: Christ
Am 21.04.2017 um 14:11 schrieb Emil Velikov:
Split the helper in two:
- an "official" VDP format to pipe
- and a custom VDP format to pipe
The latter is DMAbuf specific and does not belong with the rest.
That is, until not until the VDP_RGBA_FORMAT_R8{,G8} formats become part
of libvdpau.
W
Hi Ville,
On 20 April 2017 at 16:20, wrote:
> Trying to build the Vulkan WSI code results in a failure without DRI3.
> Make configure check that DRI3 is available before trying to build
> the Vulkan drivers.
It shouldn't. Could you please try this - fixing the existing
conditional - instead? ht
2017-04-21 14:08 GMT+02:00 Emil Velikov :
> Note: GCC 5/6 crashes with internal compiler error at seemingly random
> stages of the build. Unless any of the SWR devs sort it out, we might
> want to disable it for now.
>
> A sample report can be seen at
> https://travis-ci.org/evelikov/Mesa/builds/22
The variable is propagated as constant, yet we explicitly cast away the
constness as we get to use it.
Make it clearer and drop the const notation all together.
Cc: Dave Airlie
Signed-off-by: Emil Velikov
---
src/gallium/drivers/llvmpipe/lp_texture.c | 4 ++--
src/gallium/drivers/soft
Both headers are used everywhere, plus they both provide interop
mechanism.
Since they define [consecutive] VDPAU driver funcs, we really want the
definitions side by side. Otherwise we risk clashing multiple functions
at the same address.
Cc: Christian König
Signed-off-by: Emil Velikov
---
sr
Split the helper in two:
- an "official" VDP format to pipe
- and a custom VDP format to pipe
The latter is DMAbuf specific and does not belong with the rest.
That is, until not until the VDP_RGBA_FORMAT_R8{,G8} formats become part
of libvdpau.
With this commit we effectively revert "st/vdpau:
From: Emil Velikov
Signed-off-by: Emil Velikov
---
.travis.yml | 39 ---
1 file changed, 36 insertions(+), 3 deletions(-)
diff --git a/.travis.yml b/.travis.yml
index aa2a55d7bb4..1b0368f0a5e 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -31,6 +31,9 @@ matri
From: Emil Velikov
Some of the libraries may be dlopened, which may not always work due to
the non-standard prefix that we're using.
Signed-off-by: Emil Velikov
---
Adding this is a good idea imho, regardless if we opt for or against
libtxc_dxtn.
---
.travis.yml | 1 +
1 file changed, 1 insert
From: Emil Velikov
Requires GCC 5.0 (due to the C++14 requirement) and LLVM 3.9.
v2: Enable the target, add libedit-dev, rework check target.
Cc: Tim Rowley
Cc: George Kyriazis
Signed-off-by: Emil Velikov
---
Note!
With this the Travis instance seems to be having Internal Compiler Error
at v
From: Emil Velikov
The former does not require any LLVM, while the latter uses LLVM 3.3.
This way we'll quickly catch any LLVM 3.3+ functionality that gets
introduced where it shouldn't.
Add the full list of addons for each build permutation.
v2: Keep libedit-dev, rework check target.
Signed-
From: Emil Velikov
Split the "if test" blocks so that we get more sensible output in case
of a failure.
Signed-off-by: Emil Velikov
---
.travis.yml | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/.travis.yml b/.travis.yml
index 8921429c7e9..a4fe00d8023 100644
--- a/.trav
From: Emil Velikov
We effectively override libdrm-dev and libxcb-dri2-0-dev since we build
and install the package locally.
Signed-off-by: Emil Velikov
---
.travis.yml | 2 --
1 file changed, 2 deletions(-)
diff --git a/.travis.yml b/.travis.yml
index f34b762a4e5..8921429c7e9 100644
--- a/.tr
From: Emil Velikov
Split the target to allow faster builds for each run.
The overall build time will be more, yet Travis runs multiple builds in
parallel so we're limited by the slowest one.
Things are split roughly as:
- DRI loaders, classic DRI drivers, classic OSMesa, make check
- All Gall
From: Emil Velikov
Signed-off-by: Emil Velikov
---
.travis.yml | 20
1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/.travis.yml b/.travis.yml
index c4bb3330d07..ea2bfea4a0b 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -22,6 +22,7 @@ env:
- LIBTXC_DXTN_
From: Andres Gomez
The txc-dxtn library implements the patented S3 Texture Compression
algorithm.
By default it won't be used but we add the possibility of setting the
USE_TXC_DXTN variable to yes in the travis web UI so it will be
installed and used for the scons tests.
Cc: Eric Anholt
Cc: Rh
From: Emil Velikov
Split into OpenCL and others, since the former is quite time consuming.
v2:
- explicitly enable/disable components
- build libvdpau 1.1 requirement
- enable st/vdpau
- build libva 1.6.2 (API 0.38) requirement
Signed-off-by: Emil Velikov
---
.travis.yml | 83
From: Emil Velikov
With next commits we'll add a couple of more options.
v2: Rework check target.
Signed-off-by: Emil Velikov
---
.travis.yml | 18 +-
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/.travis.yml b/.travis.yml
index a4fe00d8023..d95d4a74650 100644
Hi all,
Bit of an update since last round:
- misc fixes throughout
- add st/vdpau, ANV and wayland support to the build
I.e. we finally build _everything_
- added SWR + Scons on Linux
Note: GCC 5/6 crashes with internal compiler error at seemingly random
stages of the build. Unless any of the
From: Andres Gomez
Trusty's LLVM toochain repository was whitelisted some time ago. See:
https://github.com/travis-ci/apt-source-whitelist/commit/479067c5e74cb0c1e2419209179b1afe2edce274
Signed-off-by: Andres Gomez
[Emil Velikov]
- set sudo to false
- reference the Trusty change (Rhys)
- kee
From: Emil Velikov
According to the manual
"If you are using ccache, use:
language: c # or other C/C++ variants
cache: ccache
to cache $HOME/.ccache and automatically add /usr/lib/ccache to your
$PATH."
Signed-off-by: Emil Velikov
---
.travis.yml | 4 +---
1 file changed, 1 insertion(+
From: Emil Velikov
Signed-off-by: Emil Velikov
---
.travis.yml | 1 +
1 file changed, 1 insertion(+)
diff --git a/.travis.yml b/.travis.yml
index e317a027233..061aed1bc7c 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -4,6 +4,7 @@ sudo: false
dist: trusty
cache:
+ apt: true
directories:
Series is,
Reviewed-by: Edward O'Callaghan
On 04/21/2017 08:44 PM, Samuel Iglesias Gonsálvez wrote:
> Series is,
>
> Reviewed-by: Samuel Iglesias Gonsálvez
>
> Sam
>
> On Fri, 2017-04-21 at 11:53 +0200, Samuel Pitoiset wrote:
>> The glsl_type helpers interface is quite nice and this (trivial
On 21 April 2017 at 07:18, Jordan Justen wrote:
> On 2017-04-20 18:49:57, Timothy Arceri wrote:
>> On 21/04/17 11:37, Jordan Justen wrote:
>> > On 2017-04-20 12:33:45, Samuel Pitoiset wrote:
>> >> I have used it sometimes, but since VERBOSE_API is missing in a bunch of
>> >> places, that's quite u
On 21 April 2017 at 02:11, Timothy Arceri wrote:
> Thanks for doing this. Looks like you forgot to updated the docs, with that
> fixed.
>
Not sure what you mean here - are you thinking of a note to
docs/release/17.2.0.html?
Thanks
Emil
___
mesa-dev mail
On 21 April 2017 at 09:08, Tapani Pälli wrote:
>
> On 04/21/2017 10:33 AM, Timothy Arceri wrote:
>>
>>
>>
>> On 21/04/17 17:31, Timothy Arceri wrote:
>>>
>>>
>>>
>>> On 21/04/17 17:08, Timothy Arceri wrote:
Currently we were only making sure types were the same within a
single stage
Series is,
Reviewed-by: Samuel Iglesias Gonsálvez
Sam
On Fri, 2017-04-21 at 11:53 +0200, Samuel Pitoiset wrote:
> The glsl_type helpers interface is quite nice and this (trivial)
> series tends to make use of the is_XXX() functions everywhere it's
> possible.
>
> Please review,
> Thanks!
>
>
FWIW, I think this series can land, because glthread is not enabled by
default, and the libX11 issue is unrelated.
Marek
On Apr 21, 2017 4:22 AM, "Michel Dänzer" wrote:
On 21/04/17 09:01 AM, Marek Olšák wrote:
> On Thu, Apr 20, 2017 at 9:44 PM, gregory hainaut
> wrote:
>> On Thu, 20 Apr 2017
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/ast_to_hir.cpp | 3 +--
src/compiler/glsl/glsl_to_nir.cpp | 4 ++--
src/compiler/glsl/ir_validate.cpp | 28 ++--
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 2 +-
4 files changed, 18 inser
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/ast_function.cpp | 2 +-
src/compiler/glsl/ast_to_hir.cpp | 6 ++--
src/compiler/glsl/builtin_functions.cpp | 12 +++
src/compiler/glsl/ir.cpp | 2 +-
src/compiler/glsl/ir_equals.cpp
---
src/compiler/glsl_types.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/src/compiler/glsl_types.h b/src/compiler/glsl_types.h
index b98ce66c48..403663f7b8 100644
--- a/src/compiler/glsl_types.h
+++ b/src/compiler/glsl_types.h
@@ -491,8 +491,7 @@ struct glsl_type {
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl_types.h | 8
1 file changed, 8 insertions(+)
diff --git a/src/compiler/glsl_types.h b/src/compiler/glsl_types.h
index 62af1496da..b98ce66c48 100644
--- a/src/compiler/glsl_types.h
+++ b/src/compiler/glsl_types.h
@@ -479,6 +479,14 @@ s
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/ast_function.cpp | 5 ++-
src/compiler/glsl/ast_to_hir.cpp | 6 ++--
src/compiler/glsl/ir.cpp | 7 ++--
src/compiler/glsl/ir_expression_operation.py | 9 ++---
src/compiler/glsl/ir_function.cpp
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/hir_field_selection.cpp | 3 +--
src/compiler/glsl/ir.cpp | 2 +-
src/compiler/glsl/ir_print_visitor.cpp | 3 +--
src/mesa/program/ir_to_mesa.cpp| 2 +-
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 4 ++--
5 f
---
src/compiler/glsl/lower_int64.cpp | 11 ++-
1 file changed, 2 insertions(+), 9 deletions(-)
diff --git a/src/compiler/glsl/lower_int64.cpp
b/src/compiler/glsl/lower_int64.cpp
index 0a19ce429e..4e2e311e02 100644
--- a/src/compiler/glsl/lower_int64.cpp
+++ b/src/compiler/glsl/lower_int
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/ast_to_hir.cpp | 2 +-
src/compiler/glsl/ir_print_visitor.cpp| 2 +-
src/compiler/glsl/link_varyings.cpp | 2 +-
src/compiler/glsl/tests/uniform_initializer_utils.cpp | 2 +-
src/mesa/state
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/ast_to_hir.cpp | 5 ++---
src/compiler/glsl/linker.cpp | 3 +--
src/compiler/glsl_types.h| 2 +-
3 files changed, 4 insertions(+), 6 deletions(-)
diff --git a/src/compiler/glsl/ast_to_hir.cpp b/src/compiler/glsl/ast_to_hir.cpp
inde
The glsl_type helpers interface is quite nice and this (trivial)
series tends to make use of the is_XXX() functions everywhere it's
possible.
Please review,
Thanks!
Samuel Pitoiset (11):
glsl: add glsl_type::is_atomic_uint() helper
glsl: make use glsl_type::is_atomic_uint()
glsl: make use o
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl/hir_field_selection.cpp | 2 +-
src/compiler/glsl/shader_cache.cpp| 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/compiler/glsl/hir_field_selection.cpp
b/src/compiler/glsl/hir_field_selection.cpp
index eab08a
Signed-off-by: Samuel Pitoiset
---
src/compiler/glsl_types.h | 8
1 file changed, 8 insertions(+)
diff --git a/src/compiler/glsl_types.h b/src/compiler/glsl_types.h
index 7709556fe0..10a22aee0a 100644
--- a/src/compiler/glsl_types.h
+++ b/src/compiler/glsl_types.h
@@ -669,6 +669,14 @@ s
On 04/21/2017 11:47 AM, Timothy Arceri wrote:
On 21/04/17 17:50, Samuel Pitoiset wrote:
I would definitely prefer to see _mesa_hash_table instead here. That
way you can get rid of the unnecessary locking at the same time.
I actually use _mesa_hash_table for the resident texture/image handles
On 21/04/17 17:50, Samuel Pitoiset wrote:
I would definitely prefer to see _mesa_hash_table instead here. That way
you can get rid of the unnecessary locking at the same time.
I actually use _mesa_hash_table for the resident texture/image handles
as part of ARB_bindless_texture. Though, I use
On 21/04/17 17:54, Samuel Pitoiset wrote:
The commit message looks a bit vague because this deleteFlag pattern is
used in many places. Here, you are *only* removing it for unlocked
objects, I think the title should be slightly improved.
Yeah. This was originally from a series that replaced a
opt_register_coalesce() was optimizing sequences such as:
mul(8) acc0:D, attr18.xyyy:D, attr19.xyyy:D
mach(8) vgrf5.xy:D, attr18.xyyy:D, attr19.xyyy:D
mov(8) m4.zw:F, vgrf5.xxxy:F
into:
mul(8) acc0:D, attr18.xyyy:D, attr19.xyyy:D
mach(8) m4.zw:D, attr18.xxxy:D, attr19.xxxy:D
This
On Monday, April 17, 2017 10:52:23 PM PDT Timothy Arceri wrote:
> From: Timothy Arceri
>
> If packing doesn't cross locations we can easily make use of
> ARB_enhanced_layouts to do packing rather than using the GLSL IR
> lowering pass lower_packed_varyings().
>
> Shader-db Broadwell results:
>
Current Anv allocator assign memory in terms of a fixed block size.
But there can be cases where this block is not enough for a memory
request, and thus several blocks must be assigned in a row.
This commit adds support for specifying how many blocks of memory must
be assigned.
This fixes a numb
On 04/21/2017 10:33 AM, Timothy Arceri wrote:
On 21/04/17 17:31, Timothy Arceri wrote:
On 21/04/17 17:08, Timothy Arceri wrote:
Currently we were only making sure types were the same within a
single stage. This looks to have regressed with 953a0af8e3f73.
V2: move the SamplersValidated res
On 21.04.2017 09:08, Timothy Arceri wrote:
Currently we were only making sure types were the same within a
single stage. This looks to have regressed with 953a0af8e3f73.
V2: move the SamplersValidated reset into the common linker
code.
https://bugs.freedesktop.org/show_bug.cgi?id=97524
Cc:
On 21.04.2017 07:20, Timothy Arceri wrote:
From Chapter 5 'Shared Objects and Multiple Contexts' of
the OpenGL 4.5 spec:
"Objects which contain references to other objects include
framebuffer, program pipeline, query, transform feedback,
and vertex array objects. Such objects are call
Patches 1-3 are:
Reviewed-by: Samuel Pitoiset
On 04/21/2017 07:20 AM, Timothy Arceri wrote:
From: Matt Turner
From Chapter 5 'Shared Objects and Multiple Contexts' of
the OpenGL 4.5 spec:
"Objects which contain references to other objects include
framebuffer, program pipeline, quer
The commit message looks a bit vague because this deleteFlag pattern is
used in many places. Here, you are *only* removing it for unlocked
objects, I think the title should be slightly improved.
On 04/21/2017 07:20 AM, Timothy Arceri wrote:
From: Matt Turner
---
src/mesa/main/arrayobj.c
I would definitely prefer to see _mesa_hash_table instead here. That way
you can get rid of the unnecessary locking at the same time.
I actually use _mesa_hash_table for the resident texture/image handles
as part of ARB_bindless_texture. Though, I use the locked hash_table
implementation insid
For the series:
Reviewed-by: Nicolai Hähnle
On 20.04.2017 19:02, Samuel Pitoiset wrote:
Hi,
Found by inspection while working on the GLSL bindless alternative
(ie. don't introduce new base types). A full piglit run doesn't
report any regressions using RadeonSI.
Please review,
Thanks!
Samuel
+1 on the idea as well, that would help in order to remove some
unnecessary locking, although not in a critical path.
On 04/19/2017 09:45 AM, Timothy Arceri wrote:
This extension is not supported by GLVND, also as far
as I can tell this extension requires us to do extra
locking for objects that
On 20.04.2017 18:32, Emil Velikov wrote:
On 18 April 2017 at 13:55, Pekka Paalanen wrote:
On Mon, 27 Feb 2017 13:26:11 +
Emil Velikov wrote:
Hi Julien,
On 27 February 2017 at 12:08, Julien Isorce wrote:
Hi,
Since 2012 commit ccff74971203b533bf16b46b49a9e61753f75e6c it is said:
"sampl
On 20.04.2017 12:21, Kai Wasserbäch wrote:
The OpenGL extension KHR_no_error is exposed since commit
d42d150ad26e29d9e894ba9f9e28f8134e2e5393 by Timothy Arceri. Therefore it
should be marked as "started" in the features.txt
Signed-off-by: Kai Wasserbäch
---
Hey,
this is a rather trivial "patch
2017-04-20 18:51 GMT+02:00 Mauro Rossi :
>
>
> 2017-04-20 4:18 GMT+02:00 Tomasz Figa :
>
>> On Wed, Apr 19, 2017 at 11:51 PM, Emil Velikov
>> wrote:
>> > Hi Tomasz,
>> >
>> > On 19 April 2017 at 08:00, Tomasz Figa wrote:
>> >> Android buffer queues can be abandoned, which results in failing to
>
On 21/04/17 17:31, Timothy Arceri wrote:
On 21/04/17 17:08, Timothy Arceri wrote:
Currently we were only making sure types were the same within a
single stage. This looks to have regressed with 953a0af8e3f73.
V2: move the SamplersValidated reset into the common linker
code.
https://bu
On Monday, April 17, 2017 10:52:24 PM PDT Timothy Arceri wrote:
> From: Timothy Arceri
>
> V2: leave copy propagation to avoid interpolation regressions
>
> IVB is running into some spilling issues with the loop
> removed so we leave it there for gen7 and below for now.
>
> Run time for shader-
On 21/04/17 17:08, Timothy Arceri wrote:
Currently we were only making sure types were the same within a
single stage. This looks to have regressed with 953a0af8e3f73.
V2: move the SamplersValidated reset into the common linker
code.
https://bugs.freedesktop.org/show_bug.cgi?id=97524
Cc:
On 04/21/2017 03:37 AM, Jordan Justen wrote:
On 2017-04-20 12:33:45, Samuel Pitoiset wrote:
I have used it sometimes, but since VERBOSE_API is missing in a bunch of
places, that's quite useless. :)
I also use MESA_VERBOSE=api every few months or so. I've found it
pretty useful, but frustrat
On 21/04/17 16:02, Tapani Pälli wrote:
On 04/19/2017 05:15 AM, Timothy Arceri wrote:
Hi Tapani,
You seem to have been one of the last people to touch this code. Do
you think you could take a look at this patch?
OK sure.
The validation loop looks fine to me but I find it strange that we ne
Currently we were only making sure types were the same within a
single stage. This looks to have regressed with 953a0af8e3f73.
V2: move the SamplersValidated reset into the common linker
code.
https://bugs.freedesktop.org/show_bug.cgi?id=97524
Cc: Tapani Pälli
---
src/mesa/main/uniform_quer
98 matches
Mail list logo
