The MSVC preprocessor doesnt understand #warning
---
src/util/process.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/util/process.c b/src/util/process.c
index 449c7fa77c2..6e6376986f3 100644
--- a/src/util/process.c
+++ b/src/util/process.c
@@ -107,7 +107,7 @@ __getProg
Hi Gert, Hi Nicolai,
I did play around with this quite a lot (mostly with the previous version) and
found it to be
stable (doesn't crash deus ex on start up from cold shader cache like NIR) or
leak memory
like llvm 7 (at least used to leak ~100MB with piglit shaders,
haven't given it another try
I thought this might be interesting because currently appveyor is only used
with MSVC and LLVM 3.3.1
They can however provide Ubuntu 16.04 LTS. Its free for OSS.
You only need some kind of repository.
Basic setup is easy
https://www.appveyor.com/docs/
The hard part is the appveyor.yml file, whi
Am 21.05.2018 um 19:21 schrieb Ian Romanick:
> On 05/10/2018 02:05 AM, Benedikt Schemmer wrote:
>> remove a memset too and yes, this is all functionally identical
>>
>> ---
>> src/mesa/main/shaderapi.c | 40
>> 1 f
en have an ubuntu image available
https://www.appveyor.com/docs/getting-started-with-appveyor-for-linux/
wonder why that isnt being used instead of msvc 2015 with llvm 3.3.1?
Am 21.05.2018 um 01:19 schrieb Timothy Arceri:
>
>
> On 20/05/18 22:16, Benedikt Schemmer wrote:
>> Ther
Am 21.05.2018 um 01:19 schrieb Timothy Arceri:
>
>
> On 20/05/18 22:16, Benedikt Schemmer wrote:
>> There is exactly one flock in mesa and it caused mesa not to build
>> on windows when shader cache was enabled.
>>
>> It should be possible to revert 9f8dc3b
There is exactly one flock in mesa and it caused mesa not to build
on windows when shader cache was enabled.
It should be possible to revert 9f8dc3bf03ec825bae7041858dda6ca2e9a34363
"utils: build sha1/disk cache only with Android/Autoconf" currently
guarding the offending code with ENABLE_SHADER_C
Thanks, fixed locally.
Am 20.05.2018 um 14:46 schrieb Mariusz Ceier:
> On 20 May 2018 at 14:16, Benedikt Schemmer wrote:
>> There is exactly one flock in mesa and it caused mesa not to build
>> on windows when shader cache was enabled.
>>
>> It s
There is exactly one flock in mesa and it caused mesa not to build
on windows when shader cache was enabled.
It should be possible to revert 9f8dc3bf03ec825bae7041858dda6ca2e9a34363
"utils: build sha1/disk cache only with Android/Autoconf" currently
guarding the offending code with ENABLE_SHADER_C
y: Marek Olšák mailto:marek.ol...@amd.com>>
>
> Marek
>
>
> On Thu, May 10, 2018 at 5:06 AM, Benedikt Schemmer <mailto:b...@besd.de>> wrote:
>
> Sometimes it is nice (and useful) to not just see if you've messed up
> but also if you
is the prerequisit for the others I just sent
CC: "Marek Olšák"
CC: Eric Anholt
CC: Emil Velikov
CC: Timothy Arceri
CC: Ilia Mirkin
Signed-off-by: Benedikt Schemmer
---
src/mesa/main/extensions_table.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src
extension was not designed.
Also this prevents extensions that might be problematic when they are always
enabled, to be enabled in the appropriate context.
Tested with several 64-bit games.
CC: "Marek Olšák"
CC: Eric Anholt
CC: Emil Velikov
CC: Timothy Arceri
CC: Ilia Mirkin
Benedikt S
Velikov
CC: Timothy Arceri
CC: Ilia Mirkin
Signed-off-by: Benedikt Schemmer
---
src/mesa/main/extensions_table.h | 98 ++--
1 file changed, 53 insertions(+), 45 deletions(-)
diff --git a/src/mesa/main/extensions_table.h b/src/mesa/main/extensions_table.h
index 93
This just makes the constants more descriptive
CC: "Marek Olšák"
CC: Eric Anholt
CC: Emil Velikov
CC: Timothy Arceri
CC: Ilia Mirkin
Signed-off-by: Benedikt Schemmer
---
src/mesa/main/extensions_table.h | 833 ---
1 file changed, 417 insertio
Am 18.05.2018 um 20:05 schrieb Ilia Mirkin:
> On Fri, May 18, 2018 at 1:54 PM, Benedikt Schemmer wrote:
>> so the gll is ok?
>>
>> my idea with the glc is that this particular extension would no longer be
>> advertised as soon as the app
>> request a context w
the year) so that
GLSL version overriding has a similar effect?
Am 18.05.2018 um 19:32 schrieb Ilia Mirkin:
> On Fri, May 18, 2018 at 1:27 PM, Benedikt Schemmer wrote:
>> Am I on the right track to assume that Timothy's patch should read
>>
s_table.h"
#undef EXT
Am 18.05.2018 um 14:35 schrieb Ilia Mirkin:
> On Fri, May 18, 2018 at 8:20 AM, Benedikt Schemmer wrote:
>> Indeed it does, I've written a little test program where I can request
>> specific OpenGL versions.
>> I think that used to work
ia Mirkin:
> The check is against what the driver max is, not what the application
> requested.
>
> Using overrides messes up this logic, it seems.
>
> On Fri, May 18, 2018, 06:14 Benedikt Schemmer <mailto:b...@besd.de>> wrote:
>
> Yes its backwards and it
he driver, just because an extension can be
> advertised at a certain level doesn't mean you must (or
> can) do it.
>
> Tim
>
> On 18/05/18 19:13, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>> I was looking into something similar myself.
>> Great t
A list of extension requirements and interactions extracted from the GL_* files
on Khronos
so this is messy but relatively complete
I've started to introduce key phrases like "This extension requires" so that
this
could become machine readable in the future.
---
./WGL_ARB_make_current_read.txt
Hi Timothy,
I was looking into something similar myself.
Great to see someone working on getting legacy extensions available
to the applications that can actually use them!
Another thing I noticed is that extensions aren't advertised at the
correct feature level.
The code below works and does a b
_FLOCK_L_TYPE */
# error "This platform lacks flock function, and Gnulib doesn't provide a
replacement. This is a bug in Gnulib."
# endif /* !HAVE_STRUCT_FLOCK_L_TYPE */
#endif /* !Windows */
Am 18.05.2018 um 07:34 schrieb Tapani Pälli:
>
>
> On 05/17/2018 02
2:28 schrieb Tapani Pälli:
>
>
> On 05/17/2018 11:38 AM, Benedikt Schemmer wrote:
>> Thanks for reviewing!
>>
>> Am 17.05.2018 um 08:42 schrieb Tapani Pälli:
>>>
>>>
>>> On 05/10/2018 12:05 PM, Benedikt Schemmer wrote:
>>>> Move s
w until we get to a proper fix.
Fixes: d1efa09d342 "util: import sha1 implementation from OpenBSD"
Signed-off-by: Emil Velikov
Am 17.05.2018 um 12:28 schrieb Tapani Pälli:
>
>
> On 05/17/2018 11:38 AM, Benedikt Schemmer wrote:
>> Thanks for reviewing!
>>
>&
Am 17.05.2018 um 09:11 schrieb Tapani Pälli:
> some nitpicking below .. I'll do some testing to see that things work as
> before but overall these changes look good to me. Would be nice to hear
> comments from active shader-db users.
>
> On 05/10/2018 12:05 PM, Benedikt S
Am 17.05.2018 um 08:59 schrieb Tapani Pälli:
>
>
> On 05/10/2018 12:05 PM, Benedikt Schemmer wrote:
>> remove a memset too and yes, this is all functionally identical
>>
>> ---
>> src/mesa/main/shaderapi.c | 40
&
Thanks for reviewing!
Am 17.05.2018 um 08:42 schrieb Tapani Pälli:
>
>
> On 05/10/2018 12:05 PM, Benedikt Schemmer wrote:
>> Move shader-cache code from back to front and make generate_sha1() usable
>> unconditionally to avoid code duplication in the following patch
>>
Sometimes it is nice (and useful) to not just see if you've messed up
but also if you've made an improvement.
---
si-report.py | 24
1 file changed, 24 insertions(+)
diff --git a/si-report.py b/si-report.py
index fba652c..d0e0686 100755
--- a/si-report.py
+++ b/si-report
When using sha as part of the captured shader name that name can get so
long that part of it is cut off and not visible anymore (although this
also happens when using long directory names etc.) So this is a general
improvement.
---
si-report.py | 72 ++-
Move shader-cache code from back to front and make generate_sha1() usable
unconditionally to avoid code duplication in the following patch
---
src/mesa/main/shaderapi.c | 228 +++---
1 file changed, 116 insertions(+), 112 deletions(-)
diff --git a/src/mesa
It is inconvenient to capture shaders by program name alone because this does
not allow to capture shaders that get overwritten by shaders with the same
program name (ie games when you change settings or piglit).
---
src/mesa/main/shaderapi.c | 47 ---
remove a memset too and yes, this is all functionally identical
---
src/mesa/main/shaderapi.c | 40
1 file changed, 20 insertions(+), 20 deletions(-)
diff --git a/src/mesa/main/shaderapi.c b/src/mesa/main/shaderapi.c
index e8acca4490..1d0ca5374b 100644
--
When capturing shaders with MESA_SHADER_CAPTURE_PATH, shaders can and will
be overwritten if they have the same program number.
This happens when you're capturing shaders from games especially after
changing graphics settings and piglit shaders (I use that as a convenient way
for testing radeonsi
Am 29.04.2018 um 21:44 schrieb Gert Wollny:
> Hello Benedict,
>
> thanks for all the testing!
thanks for all the developing ;)
>
> On 29.04.2018 12:12, Benedikt Schemmer wrote:
>>> Which are the names of these test? I'd like to check this on r600,
>
% -0.58 % .7.41 %
6.67 %0.26 % -0.89 % .
--
Total 52662 .0.01 % -0.19 % .1.34 %
1.09 %0.01 % . .
Am 29.04.2018 um 11:34 schrieb Gert Wollny:
> Am Sonntag, den 29.04.2018, 10
Am Sonntag, den 29.04.2018, 10:43 +0200 schrieb Benedikt Schemmer:
>> Hi Gert,
>>
>> couldn't resist at least to try what would happen if I enable
>> register merge for radeonsi:
>>
>> PERCENTAGE DELTASShaders SGPRs VGPRs SpillSGPR
>>
Am 29.04.2018 um 11:34 schrieb Gert Wollny:
> Am Sonntag, den 29.04.2018, 10:43 +0200 schrieb Benedikt Schemmer:
>> Hi Gert,
>>
>> couldn't resist at least to try what would happen if I enable
>> register merge for radeonsi:
>>
>> PERCENTAGE DEL
hen your code makes it worse and when its
an improvement this would be really nice.
Really interesting.
Cheers, Benedikt
Am 29.04.2018 um 09:55 schrieb Gert Wollny:
> Hello Benedikt,
>
> Am Sonntag, den 29.04.2018, 00:06 +0200 schrieb Benedikt Schemmer:
>> Hi Gert
>>
>&g
Hi Gert
Am 28.04.2018 um 23:51 schrieb Gert Wollny:
> Am Samstag, den 28.04.2018, 22:43 +0200 schrieb Benedikt Schemmer:
>> The patches apply cleanly, however I just did a shader-db test run
>> and can't find a difference with your patch
>> applied, am I doing something
The patches apply cleanly, however I just did a shader-db test run and can't
find a difference with your patch
applied, am I doing something wrong?
compile times went up though:
before:
Thread 3 took 113.72 seconds and compiled 17899 shaders (not including SIMD16)
with 2232 GL context switches
Change from a purely number.shader_test naming scheme to sha_number.shader_test
because especially games often use the same number for entirely different
shaders
based on graphics settings etc. and then already captured shaders get
overwritten.
It is also useful for capturing shaders from applica
For the series:
Tested-by: Benedikt Schemmer
Tested with
Series starting with [Mesa-dev] [PATCH 1/7] ac/surface: handle DCC
subresource fast clear restriction on VI
Series starting with [Mesa-dev] [PATCH 1/9] radeonsi: remove
r600_pipe_common.h
Series starting with [Mesa-dev] [PATCH 0/9
Hi all,
I did some more test runs with NIR and shader-db.
This time they also include piglit shaders.
There was only one crash that was explicitly related to NIR but also
quite a number of regressions that might be worth looking at.
I've put the results up on
https://github.com/bendat78/mymesa/tre
Hi Marek,
I just tried different combinations: this and with meta
ctx->flags |= SI_CONTEXT_INV_ICACHE |
SI_CONTEXT_INV_SMEM_L1 |
SI_CONTEXT_INV_VMEM_L1 |
SI_CONTEXT_INV_GLOBAL_L2 |
SI_CONTEXT_START_PIPELINE_ST
RX460 ;)
For the series:
Tested-by: Benedikt Schemmer
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Hi Marek,
your patches dont apply cleanly
(src/gallium/drivers/radeonsi/si_gfx_cs.c in particular)
Rebase?
Other than that:
Dirt Rally (avg,min,max)
before 54/35/73
after 54/41/73
My system is GPU bound. RX560 OC with 4GB (aka reflashed RX460 ;)
For the series:
Tested-by: Benedikt Schemmer
Depending on the application captured shaders may or may not be the same
when they have the same number.
In Games it often depends on the users graphics options.
This is a problem when running shader-db and looking at the output.
There the same shader number can mean entirely different shaders
dep
RX460 4GB
Am 04.04.2018 um 15:59 schrieb Timothy Arceri:
> On 04/04/18 22:53, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>> another game that is behaving strangly is Metro 2033 Redux, also crashes
>> earlier with apitrace attached. This time TGSI and NIR crash at about
apitrace
Am 04.04.2018 um 14:06 schrieb Timothy Arceri:
> On 04/04/18 21:51, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>> thanks for looking into this.
>>
>> Dead Island still crashes for me with NIR.
>
> Just to make sure, are you building a 32bit version of M
bDivisor(index = 4, divisor = 0) // incomplete
Am 04.04.2018 um 14:06 schrieb Timothy Arceri:
> On 04/04/18 21:51, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>> thanks for looking into this.
>>
>> Dead Island still crashes for me with NIR.
>
> Just to make s
ause mesa git feels a bit
broken. If I switch to stable versions like 17.3.8 most problems
disappear (but so does NIR I guess ;)
Am 04.04.2018 um 14:06 schrieb Timothy Arceri:
> On 04/04/18 21:51, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>> thanks for looking into this.
>&g
ctx =
0xce69b30) = True
294175 glXSwapIntervalMESA(interval = 0) = 0
294176 glXMakeCurrent(dpy = 0xcbbc820, drawable = 0, ctx = NULL) = True
Am 04.04.2018 um 06:05 schrieb Timothy Arceri:
> On 31/03/18 02:44, Benedikt Schemmer wrote:
>> Hi all,
>>
>> I did some more testing
Hi Timothy,
I put up another run for you with llvm 7 and mesa git of this morning.
Also I pruned duplicate shaders (especially Metro Redux had significant
amounts of those) so we're down to ~24k unique shaders.
I'm going to use release builds for testing because there is no
difference to the debu
From: Benedikt Schemmer
Date: Sun, 1 Apr 2018 13:18:02 +0200
Subject: [PATCH] Disable bindless textures with radeonsi NIR until there
is support for them.
- Allows to build and use a debug build of mesa with
shader-db (crashes otherwise with affected shaders)
- Dirt Rally doesnt care about
Hi all,
I did some more testing with NIR and wanted to share the results.
https://github.com/bendat78/mymesa/tree/mymesa2/testresults
Overall it seems nir could be better than tgsi, but there are some
shaders with significant regressions and I think that introduces
framerate drops that are quite
Hi all,
while looking for problems with NIR, I noticed that in the Bioshock
Infinite and also Dirt Rallye Benchmark there are times when GPU shader
utilization drops to 80% (Bioshock) or 60% (Dirt Rallye).
In Bioshock this directly associated with fps drops, in Dirt Rallye it
is not so obvious bu
: some crashes during shader generation
in Deus Ex, Dying Light and Dead Island (Original) (see nir test) and
TressFX not working correctly in Tomb Raider.
Am 07.03.2018 um 05:16 schrieb Timothy Arceri:
> On 06/03/18 23:49, Benedikt Schemmer wrote:
>> Hi Timothy,
>>
>>
Hi all,
I thought I'd do some testing on my machine to see if there are any
problems with nir (enabled with R600_DEBUG=nir)
Tested on Ubuntu 17.10 with mesa git of today
(oibaf for 32 bit, mine with LLVM5.0 on 64 bit because I cant figure out
how to build 32bit drivers on ubuntu :( )
Xeon 126
From: Benedikt Schemmer
Date: Sat, 30 Sep 2017 18:05:43 +0200
Subject: [PATCH V3] radeonsi/uvd: fix planar formats broken since
f70f6baaa3bb0f8b280ac2eaea69bb
V1: Marek Olsak
V2: remove code duplication and one unnecessary variable, minor whitespace fix
V3: ensure valid resource format when
yes
Am 30.09.2017 um 17:53 schrieb Marek Olšák:
> On Sat, Sep 30, 2017 at 5:47 PM, Christian König
> wrote:
>> Am 30.09.2017 um 12:57 schrieb Benedikt Schemmer:
>>>>
>>>> It should be handled as an error if resource_formats[0] is
>>>>
>
> It should be handled as an error if resource_formats[0] is PIPE_FORMAT_NONE.
>
> Better write this as:
>
> if (resource_formats[i] == PIPE_FORMAT_NONE) {
> if (i == 0)
> gotot error;
> continue;
> }
> ...
>
I dont think it can be zero.
In the beginning of vl_video_buffer.c form
From: Benedikt Schemmer
Date: Fri, 29 Sep 2017 21:02:13 +0200
Subject: [PATCH V2] radeonsi/uvd: fix planar formats broken since
f70f6baaa3bb0f8b280ac2eaea69bb
V2: remove code duplication and one unnessecary variable, minor whitespace fix
---
src/gallium/drivers/radeonsi/si_uvd.c | 40
Would it be ok to rewrite the patch like below?
Avoids code duplication and one variable.
Otherwise your original patch works fine so FWIW you have my
Tested-by: Benedikt Schemmer
Cheers,
Benedikt
diff --git a/src/gallium/drivers/radeonsi/si_uvd.c
b/src/gallium/drivers/radeonsi/si_uvd.c
There is a small typo that prevents compilation.
Please kindly review and push.
Cheers,
Benedikt
Signed-off-by: Benedikt Schemmer
diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c
b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c
--- a/src/gallium/drivers/nouveau/nvc0
Wrong commit, correct one is:
gallium: add PIPE_QUERY_OCCLUSION_PREDICATE_CONSERVATIVE
Am 18.09.2017 um 15:27 schrieb Benedikt Schemmer:
> There is a small typo that prevents compilation.
> Please kindly review and push.
> Cheers,
> Benedikt
>
> Signed-off-by: Benedikt Schemme
There is a small typo that prevents compilation.
Please kindly review and push.
Cheers,
Benedikt
Signed-off-by: Benedikt Schemmer
---
diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c
b/src/gallium/drivers/nouveau/nvc0/nvc0_query_hw.c
--- a/src/gallium/drivers/nouveau/nvc0
This series aims to fix hundreds of missing initializer warnings in generated
header files
when compiling with -Wextra
V1: Fix the old fashioned way by adding 0s where needed
V2: switch to designated initializers (Emil), didnt send
V3: add some layout so its easier to read and create a new versio
Fix missing initializer warning in egd_tables.h by adding appropriate
designated initializers in egd_tables.py
---
src/gallium/drivers/r600/egd_tables.py | 15 ++-
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/gallium/drivers/r600/egd_tables.py
b/src/gallium/dri
Fix missing initializer warning in vk_format_table.h by changing to a default
initializer in vk_format_table.py
and correct the autogenerated from message
---
src/amd/vulkan/vk_format_table.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/amd/vulkan/vk_format_table.
Fix missing initializer warning in sid_tables.h by adding appropriate
designated initializers in sid_tables.py
---
src/amd/common/sid_tables.py | 15 ++-
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/amd/common/sid_tables.py b/src/amd/common/sid_tables.py
index f
initializer is redundant and unnecessary. Empty
>> initializer {} is also commonly used instead of memset.
>>
>> You need to suppress this warning if you don't want to see it.
>>
>> Marek
>>
>> On Sun, Jun 4, 2017 at 9:45 AM, Benedikt Schemmer
Fix missing initializer warning in vk_format_table.h by adding appropriate
default fields in vk_format_table.py
and correct the autogenerated from message
---
src/amd/vulkan/vk_format_table.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/amd/vulkan/vk_format_table.
Fix missing initializer warning in sid_tables.h by adding appropriate default
fields in sid_tables.py
---
src/amd/common/sid_tables.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/amd/common/sid_tables.py b/src/amd/common/sid_tables.py
index fd88d3c9d5..691d766b08
Fix missing initializer warning in egd_tables.h by adding appropriate default
fields in egd_tables.py
---
src/gallium/drivers/r600/egd_tables.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/r600/egd_tables.py
b/src/gallium/drivers/r600/egd_tables.p
s frequent errors.
This also fixes a small copy and paste error in vk_format_table.py where
it isnt clear how it is autogenerated from where.
Please kindly review and push if you find this useful.
Thanks,
Benedikt
Benedikt Schemmer (3):
Fix missing initializer warning in sid_tables.h by adding a
From: Benedikt Schemmer
This patch sets the allow_glsl_builtin_variable_redeclaration for Dead
Island Riptide Definitive Edition.
Tested with Mesa git as of today and Dying Light, Dead Island Definitive
Edition and Dead Island Riptide Definite Edition
v2: set only required option and
Hi all,
Dying Light really works (Radeon RX460 4GB, Ubuntu 17.04, Mesa git
patches applied plus some minor changes that i have locally to get rid
of compiler warnings).
I did have to jump through some additional hoops to get it working:
Dying Light only launches from within Steam, Launch Opti
From: Benedikt Schemmer
This patch sets the allow_glsl_builtin_redeclaration for Dead Island
Riptide Definitive Edition.
This requires the patch series:
[Mesa-dev] [PATCH 0/2] Allow redeclaration of GLSL builtins; fixes Dying
Light and Dead Island Definitive Edition
by John Brooks sent Fri
78 matches
Mail list logo
