With the minor fix in 1/2, this series is
Reviewed-by: Ilia Mirkin
On Fri, Aug 3, 2018 at 10:52 PM, Karol Herbst wrote:
> mitigates hurt shaders after adding sqrt:
> total instructions in shared programs : 5456166 -> 5454825 (-0.02%)
> total gprs used in shared programs: 647522 -> 647551 (0
On Fri, Aug 3, 2018 at 10:52 PM, Karol Herbst wrote:
> ./GpuTest /test=pixmark_piano 1024x640 30sec:
> 301 -> 327 points
>
> shader-db:
> total instructions in shared programs : 5472103 -> 5456166 (-0.29%)
> total gprs used in shared programs: 647530 -> 647522 (-0.00%)
> total shared used in s
https://bugs.freedesktop.org/show_bug.cgi?id=107478
Chris Kawano changed:
What|Removed |Added
CC||ckaw...@oicinc.com
--
You are receiving
https://bugs.freedesktop.org/show_bug.cgi?id=107478
Chris Kawano changed:
What|Removed |Added
CC|ckaw...@oicinc.com |
--
You are receiving this mail because
https://bugs.freedesktop.org/show_bug.cgi?id=107478
Chris Kawano changed:
What|Removed |Added
CC||ckaw...@oicinc.com
--
You are receiving
mitigates hurt shaders after adding sqrt:
total instructions in shared programs : 5456166 -> 5454825 (-0.02%)
total gprs used in shared programs: 647522 -> 647551 (0.00%)
total shared used in shared programs : 389120 -> 389120 (0.00%)
total local used in shared programs : 21064 -> 21064 (0.0
A significant perf increase is expected for shaders doing a lot of sqrt
operations.
shader-db changes for complete series:
total instructions in shared programs : 5472103 -> 5454825 (-0.32%)
total gprs used in shared programs: 647530 -> 647551 (0.00%)
total shared used in shared programs : 38
https://bugs.freedesktop.org/show_bug.cgi?id=107478
Bug ID: 107478
Summary: Regression: Mesa compile broken (snprintf) with Visual
Studio 2013 (vs2013)
Product: Mesa
Version: unspecified
Hardware: x86 (IA32)
./GpuTest /test=pixmark_piano 1024x640 30sec:
301 -> 327 points
shader-db:
total instructions in shared programs : 5472103 -> 5456166 (-0.29%)
total gprs used in shared programs: 647530 -> 647522 (-0.00%)
total shared used in shared programs : 389120 -> 389120 (0.00%)
total local used in shar
On 08/02/2018 03:26 PM, Jason Ekstrand wrote:
> This one bothers me a bit. People don't usually expect integer
> calculations so be modified at all so to replace an expression that
> takes in an internet and sometimes returns a different one with a no-op
> seems rather unexpected. Then again, GLSL
https://bugs.freedesktop.org/show_bug.cgi?id=107477
--- Comment #1 from Clément Guérin ---
Forgot to mention that I'm using a R9 Fury.
--
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.___
mes
https://bugs.freedesktop.org/show_bug.cgi?id=107477
Bug ID: 107477
Summary: [DXVK] Setting high shader quality in GTA V results in
LLVM error
Product: Mesa
Version: git
Hardware: Other
OS: All
St
Acked-by: Ian Romanick
On 08/03/2018 04:03 PM, Marek Olšák wrote:
> From: Marek Olšák
>
> v2: use correct files
> ---
> include/GL/glcorearb.h | 76 ---
> include/GL/glext.h | 85 +++---
> include/GL/glxext.h| 22 +-
> include/GLES2/gl2.h| 8
This also looks like a harmless and useful refactory.
Reviewed-by: Rafael Antognolli
On Thu, Aug 02, 2018 at 10:39:21AM +0100, Lionel Landwerlin wrote:
> Signed-off-by: Lionel Landwerlin
> ---
> src/intel/tools/aub_mem.c | 391
> src/intel/tools/aub_mem.h
Reviewed-by: Rafael Antognolli
On Thu, Aug 02, 2018 at 10:39:22AM +0100, Lionel Landwerlin wrote:
> ---
> src/intel/tools/aub_write.h | 8
> 1 file changed, 8 insertions(+)
>
> diff --git a/src/intel/tools/aub_write.h b/src/intel/tools/aub_write.h
> index b421679b9eb..6a09c1747b9 10064
On Thu, Aug 02, 2018 at 10:39:20AM +0100, Lionel Landwerlin wrote:
> Signed-off-by: Lionel Landwerlin
> ---
> src/util/rb_tree.h | 36
> 1 file changed, 36 insertions(+)
>
> diff --git a/src/util/rb_tree.h b/src/util/rb_tree.h
> index c77e9255ea2..df1a4197b8a
Looks like no functional change, and it's needed by the ui tool, so
Reviewed-by: Rafael Antognolli
On Thu, Aug 02, 2018 at 10:39:19AM +0100, Lionel Landwerlin wrote:
> Signed-off-by: Lionel Landwerlin
> ---
> src/intel/tools/aub_read.c | 307 ++
> src/intel/too
On Thu, Aug 2, 2018 at 8:52 PM, Dieter Nützel wrote:
> Am 02.08.2018 05:14, schrieb Marek Olšák:
>>
>> On Wed, Aug 1, 2018 at 5:02 PM, Dieter Nützel
>> wrote:
>>>
>>> Am 01.08.2018 19:53, schrieb Marek Olšák:
On Wed, Aug 1, 2018 at 12:02 PM, Dieter Nützel
wrote:
>
>
>
From: Marek Olšák
v2: use correct files
---
include/GL/glcorearb.h | 76 ---
include/GL/glext.h | 85 +++---
include/GL/glxext.h| 22 +-
include/GLES2/gl2.h| 8 +-
include/GLES2/gl2ext.h | 161 +++--
include
On Thu, Aug 02, 2018 at 10:39:13AM +0100, Lionel Landwerlin wrote:
> Up to now we've been lucky that the buffer returned was always exactly
> at the address we requested.
Looks like this needs to land, even if the rest of the series doesn't.
Reviewed-by: Rafael Antognolli
> Fixes: 144b40db5411
Oh. That's the base relative to the bo offset. Which is a u32. So it's
actually OK
On Fri, Aug 3, 2018 at 6:06 PM, Rhys Perry wrote:
> Yeah
>
> "base" in nve4_cp_launch_desc_set_cb() and
> gp100_cp_launch_desc_set_cb() are uint32_t too. They should probably
> be updated
>
> On Fri, Aug 3, 2018 at
Yeah
"base" in nve4_cp_launch_desc_set_cb() and
gp100_cp_launch_desc_set_cb() are uint32_t too. They should probably
be updated
On Fri, Aug 3, 2018 at 11:04 PM, Ilia Mirkin wrote:
> On Fri, Aug 3, 2018 at 5:57 PM, Rhys Perry wrote:
>> Previously, UBOs were implemented using global memory loads
On Fri, Aug 3, 2018 at 5:57 PM, Rhys Perry wrote:
> Previously, UBOs were implemented using global memory loads for compute on
> Kepler+ because it only supported 8 constant buffers on compute. This
> required bounds checking and expensive load instructions.
>
> However 6 of the constant buffer bi
Sorry, the patch is incorrect. It came from a stale repo of mine, not
from opengl-registry. I'll send a new patch.
Marek
On Fri, Aug 3, 2018 at 2:17 PM, Ian Romanick wrote:
> Actually, I'm going to NAK it. :P I was going to submit a Khronos bug,
> but I noticed that there are newer header files
Previously, UBOs were implemented using global memory loads for compute on
Kepler+ because it only supported 8 constant buffers on compute. This
required bounds checking and expensive load instructions.
However 6 of the constant buffer bindings were left unused, this uses them
instead of loading f
Oh. argparse.add_argument doesn't do what I thought it did. Neat. I
was going by memory, which most resembles a colander by now =/
This change is
Reviewed-by: Ilia Mirkin
On Fri, Aug 3, 2018 at 5:34 PM, Rhys Perry wrote:
> They should still be positional arguments.
>
> For example, the output
They should still be positional arguments.
For example, the output of --help (after the 5th patch) is:
> usage: nv-report.py [-h] [--top ATTR N] [--smallest ATTR N] before after
>
> Shows changes between two shader-db runs for the nouveau drivers.
>
> positional arguments:
> before
I'd rather keep these as positional args. You can still use argparse
for other optional items.
On Fri, Aug 3, 2018 at 3:53 PM, Rhys Perry wrote:
> Signed-off-by: Rhys Perry
> ---
> nv-report.py | 13 +++--
> 1 file changed, 11 insertions(+), 2 deletions(-)
>
> diff --git a/nv-report.py
FWIW I've avoided looking at the "affected" stats, since IMHO they're
meaningless. A lot of people like to look at them to increase the
apparent significance of their optimization. If a change reduces a
shader's instructions by 99%, it's not that big of a deal if it's the
only shader where it does
both look good to me
On Fri, Aug 3, 2018 at 10:20 PM, Ilia Mirkin wrote:
> On Fri, Aug 3, 2018 at 3:53 PM, Rhys Perry wrote:
>> Signed-off-by: Rhys Perry
>> ---
>> nv-report.py | 49 +
>> 1 file changed, 33 insertions(+), 16 deletions(-)
>>
>> di
On Fri, Aug 3, 2018 at 3:53 PM, Rhys Perry wrote:
> Signed-off-by: Rhys Perry
> ---
> nv-report.py | 49 +
> 1 file changed, 33 insertions(+), 16 deletions(-)
>
> diff --git a/nv-report.py b/nv-report.py
> index 3f0be63557..487074fdd6 100644
> ---
Reviewed-by: Ilia Mirkin
On Fri, Aug 3, 2018 at 3:53 PM, Rhys Perry wrote:
> Signed-off-by: Rhys Perry
> ---
> nv-report.py | 53 +++--
> 1 file changed, 19 insertions(+), 34 deletions(-)
>
> diff --git a/nv-report.py b/nv-report.py
> index bdc20
Signed-off-by: Rhys Perry
---
nv-report.py | 13 +++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/nv-report.py b/nv-report.py
index 079b2e08a3..1c1491af0a 100644
--- a/nv-report.py
+++ b/nv-report.py
@@ -12,11 +12,15 @@ fields, except for the type, are optional.
imp
Signed-off-by: Rhys Perry
---
nv-report.py | 28 +++-
1 file changed, 19 insertions(+), 9 deletions(-)
diff --git a/nv-report.py b/nv-report.py
index 487074fdd6..079b2e08a3 100644
--- a/nv-report.py
+++ b/nv-report.py
@@ -46,11 +46,13 @@ class Stats(object):
asse
Signed-off-by: Rhys Perry
---
nv-report.py | 53 +++--
1 file changed, 19 insertions(+), 34 deletions(-)
diff --git a/nv-report.py b/nv-report.py
index bdc2093a73..3f0be63557 100644
--- a/nv-report.py
+++ b/nv-report.py
@@ -14,6 +14,9 @@ import re
This adds two new features to nv-report:
- additionally reporting changes of only affected programs
- an option to print smallest affected programs or most affected programs
Rhys Perry (5):
nv-report: deduplicate list of fields
nv-report: move code out of main()
nv-report: additionally repor
Signed-off-by: Rhys Perry
---
nv-report.py | 53 +++--
1 file changed, 51 insertions(+), 2 deletions(-)
diff --git a/nv-report.py b/nv-report.py
index 1c1491af0a..703fcf4571 100644
--- a/nv-report.py
+++ b/nv-report.py
@@ -81,13 +81,16 @@ def read_
Signed-off-by: Rhys Perry
---
nv-report.py | 49 +
1 file changed, 33 insertions(+), 16 deletions(-)
diff --git a/nv-report.py b/nv-report.py
index 3f0be63557..487074fdd6 100644
--- a/nv-report.py
+++ b/nv-report.py
@@ -45,8 +45,14 @@ class Stats(o
Quoting Gert Wollny (2018-08-03 11:12:47)
> Am Freitag, den 03.08.2018, 08:47 -0700 schrieb Dylan Baker:
> > Quoting Gert Wollny (2018-08-03 02:55:36)
> > > From: Gert Wollny
> > >
> > > os.path.exists doesn't return True for stale symlinks, but they are
> > > in
> > > the way later, when a link/
Actually, I'm going to NAK it. :P I was going to submit a Khronos bug,
but I noticed that there are newer header files (July 25, 2018 vs May
25, 2018) that don't have this weird issue. Can we have those instead? :)
See
https://github.com/KhronosGroup/OpenGL-Registry/blob/master/api/GL/glext.h.
Am Freitag, den 03.08.2018, 08:47 -0700 schrieb Dylan Baker:
> Quoting Gert Wollny (2018-08-03 02:55:36)
> > From: Gert Wollny
> >
> > os.path.exists doesn't return True for stale symlinks, but they are
> > in
> > the way later, when a link/file with the same name is to be
> > created.
> > For in
GLXCreate{,New}Context, like most X resource creation requests, does not
emit a reply and therefore is emitted into the X stream asynchronously.
However, unlike most resource creation requests, the GLXContext we
return is a handle to library state instead of an XID. So if context
creation fails for
On 2 August 2018 at 20:31, Mauro Rossi wrote:
> A problem was reported with arm,arm64 targets build due to missing
> libLLVM shared library dependency with AOSP; to avoid this issue vulkan.radv
> is built conditionally only when radeonsi is in BOARD_GPU_DRIVERS
>
> Fixes: 0ca153f869 ("android: rad
On 08/02/2018 05:52 PM, Timothy Arceri wrote:
> I thought you were trying to move these type of opts to your new pass?
>
> Or should I be splitting out and resending some of the bcsel opts in my
> dirt showdown patch [1]? And trying again to get the other patches from
> the series reviewed.
The n
Tested-by: Aaron Watry
Yay, I can finally use gnome/wayland with the Slack snap again without
insta-crashing my session.
--Aaron
On Thu, Aug 2, 2018 at 7:29 AM, Olivier Fourdan wrote:
> st_renderbuffer_delete() can segfault if we get a non-NULL context
> pointer but if the st_context is NULL:
if statement should check return value
of drmSyncobjImportSyncFile function.
Fixes CID 1438127
Signed-off-by: Sagar Ghuge
---
src/gallium/drivers/v3d/v3d_fence.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/drivers/v3d/v3d_fence.c
b/src/gallium/drivers/v3d/v3
https://bugs.freedesktop.org/show_bug.cgi?id=107474
Bug ID: 107474
Summary: RADV does not support a dedicated transfer queue
Product: Mesa
Version: unspecified
Hardware: All
OS: All
Status: NEW
Severity: e
Quoting Gert Wollny (2018-08-03 02:55:36)
> From: Gert Wollny
>
> os.path.exists doesn't return True for stale symlinks, but they are in
> the way later, when a link/file with the same name is to be created.
> For instance it is conceivable that the pointed to file is replaced by
> a file with a
Am 03.08.2018 um 07:00 schrieb Dave Airlie:
> On 3 August 2018 at 14:02, Roland Scheidegger wrote:
>> Am 03.08.2018 um 05:10 schrieb Dave Airlie:
>>> From: Dave Airlie
>>>
>>> This fixes some hangs with the arb_shader_image_load_store-atomicity tests
>>> on evergreen/cayman GPUs.
>>>
>>> I'm not
Hi,
I am trying to update mesa from 17.2.0-rc4 to latest 18.1.4 in Android N os
for AMD RAVEN GPU.
I could able to compile successfully and executed. But my issue is Window
is not completely up.
Means only one GLConsumer of android could open GPU device. All other
GLConsumers could not. So NO displ
From: Emil Velikov
Add LookupDevices/FindDevice helpers, alongside their first user -
software device ;-)
This adds support to both dri2 and haiku backends.
Signed-off-by: Emil Velikov
---
How can we split this up for platforms that support SW and HW devices?
FindDevice on a HW device will fai
From: Emil Velikov
Add implementation based around the drmDevice API. As such it's only
available only when building with libdrm. With the latter already a
requirement when using !SW code paths in the platform code.
Note: the current code will work if a device is hot-plugged. Yet
hot-unplugged i
From: Emil Velikov
The EGL_EXT_platform_device extension will be coming shortly. It will
effectively require comparing the platform specific options, passed via
the attribs array.
Signed-off-by: Emil Velikov
---
I'm not a huge fan of this. Yet again the other platform (x11) that uses
these has
From: Emil Velikov
The DRI module provides the correct format. If there is a cornercase
where it does the wrong thing, we should address it.
Signed-off-by: Emil Velikov
---
src/egl/drivers/dri2/platform_surfaceless.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/egl/d
From: Emil Velikov
The helper calls a single function - there's limited benefit in having
it as-is.
Inline, as with follow-up patch we'll tweak the createImage invocation.
Signed-off-by: Emil Velikov
---
src/egl/drivers/dri2/platform_surfaceless.c | 20 +---
1 file changed, 5
From: Emil Velikov
This new 'platform' is added by default with no guards.
It is effectively a copy of the surfaceless one, with updated function
names and brand new probe function.
Due to the reuse, some of the ifdef HAVE_SURFACELESS_PLATFORM guards
have been dropped.
A worthy mention are the
From: Emil Velikov
The only user was removed with previous patch.
Signed-off-by: Emil Velikov
---
src/egl/drivers/dri2/egl_dri2.h | 1 -
src/egl/drivers/dri2/platform_surfaceless.c | 7 ---
2 files changed, 8 deletions(-)
diff --git a/src/egl/drivers/dri2/egl_dri2.h b/src/egl/
From: Emil Velikov
One cannot have haiku and surfaceless,x11,etc.
Group things up, which will make the addition of platform_device a bit
easier.
Signed-off-by: Emil Velikov
---
View with git show -w to omit the whitespace changes.
---
src/egl/meson.build | 67 +++--
From: Emil Velikov
Introduce the API for device query and enumeration. Those at the moment
produce nothing useful since zero devices are actually available.
With later commits we'll add support for software (always) and hardware
devices. Each one exposing the respective extension string.
Note:
Hi all,
This series implements the following extensions:
- EGL_EXT_device_base
- EGL_MESA_device_software
- EGL_EXT_device_drm
- EGL_platform_device
As you know the APIs are used to enumerate, select and use EGLDevices.
The second extension (proposed by Ajax) defines a 'software' device,
alon
From: Jonny Lamb
The extension is effectively a combination of device_enumerate and
device_query.
This adds the stubbed entry points alongside the EGL_KHR_debug
boilerplate.
v2 (Emil) - rebase, EGL_KHR_debug boilerplate, remove ifdef guards
Signed-off-by: Jonny Lamb
Signed-off-by: Emil Veliko
https://bugs.freedesktop.org/show_bug.cgi?id=106163
Gert Wollny changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
https://bugs.freedesktop.org/show_bug.cgi?id=106163
--- Comment #2 from mirh ---
This.. seems to have been fixed, I think?
My own case at least does.
--
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.
One tiny nit below but for patches 3,4 and 5:
Reviewed-by: Tapani Pälli
Special thanks for the documentation. I want to go through rest of
changes within Android but I'm currently experiencing some horrible
issues with the image building. I'm hoping to resolve those soon and get
back to busi
Reviewed-by: Tapani Pälli
On 07/31/2018 09:18 PM, Chad Versace wrote:
There exist *two* queryable EGL_RENDER_BUFFER states in EGL:
eglQuerySurface(EGL_RENDER_BUFFER) and
eglQueryContext(EGL_RENDER_BUFFER).
These changes eliminate potentially very fragile code in the upcoming
EGL_KHR_mutable_re
Reviewed-by: Tapani Pälli
On 07/31/2018 09:18 PM, Chad Versace wrote:
This pulls an 'else' block into the function's main body, making the
code easier to follow.
Without this change, the upcoming EGL_KHR_mutable_render_buffer patch
transforms dri2_make_current() into spaghetti.
---
src/egl/d
Yeah, just rename.
Regards,
Qiang
From: Emil Velikov
Sent: Friday, August 3, 2018 6:38:18 PM
To: Yu, Qiang
Cc: ML mesa-dev; Adam Jackson; Michel Dänzer; Eric Engestrom
Subject: Re: [PATCH v2 2/3] driconf: move ${sysconfdir}/drirc to
${datadir}/drirc.d/00
Hi Qiang Yu,
Couple of suggestions mentioned before apply here as well.
On 3 August 2018 at 09:19, Qiang Yu wrote:
> v2:
> add kernel_driver device attribute to specify kernel
> driver name instead of reuse driver attribute.
>
Please move to the end of the commit message.
Can you split thin
Hi Emil,
Thanks for the review, I indeed miss the dirent free.
>> +/** \brief Parse configuration files in a directory */
>> +static void
>> +parseConfigDir(struct OptConfData *data, const char *dirname)
>> +{
>> +int i, count;
>> +struct dirent **entries = NULL;
>> +
>> +count = scan
On 3 August 2018 at 09:19, Qiang Yu wrote:
> ${sysconfdir} is for store admin config files, so move
> this mesa default config file to ${datadir}/drirc.d.
>
> Signed-off-by: Qiang Yu
> ---
> src/util/00-mesa-defaults.conf | 314
> +
> src/util/Makefile.am
Hi Qiang Yu,
Couple of minor suggestions.
On 3 August 2018 at 09:19, Qiang Yu wrote:
> v2:
> drop /etc/drirc.d
>
Revision log is at the end of the commit message. See for example
61a02729f749add535ad9d18c62f65641e428cfb - git log has many more.
> Driver and application can put their drirc fil
From: Gert Wollny
os.path.exists doesn't return True for stale symlinks, but they are in
the way later, when a link/file with the same name is to be created.
For instance it is conceivable that the pointed to file is replaced by
a file with a new name, and then the symlink is dead.
To handle thi
https://bugs.freedesktop.org/show_bug.cgi?id=107457
Timo Aaltonen changed:
What|Removed |Added
CC||tjaal...@ubuntu.com
--
You are receivi
v2:
add kernel_driver device attribute to specify kernel
driver name instead of reuse driver attribute.
drirc implementation of MESA_LOADER_DRIVER_OVERRIDE which can be
used to override dri driver to load.
Usage:
override dri driver for device with spec kernel driver name:
or
S
${sysconfdir} is for store admin config files, so move
this mesa default config file to ${datadir}/drirc.d.
Signed-off-by: Qiang Yu
---
src/util/00-mesa-defaults.conf | 314 +
src/util/Makefile.am | 5 +-
src/util/drirc | 314 --
v2:
drop /etc/drirc.d
Driver and application can put their drirc files in
${datadir}/drirc.d/ with name xxx.conf. Config files
will be read and applied in file name alphabete order.
So there are three places for drirc listed in order:
1. /usr/share/drirc.d/
2. /etc/drirc
3. ~/.drirc
Signed-off
v2:
1. remove /etc/drirc.d
2. move orginal /etc/drirc to /usr/share/drirc.d
3. add kernel_driver device attribute
This patch series is discussed in thread:
https://lists.freedesktop.org/archives/mesa-dev/2018-July/201348.html
amdgpu-pro driver can use drirc config to override the mesa radeo
Hi Rob,
Il giorno gio 2 ago 2018 alle ore 22:33 Rob Herring
ha scritto:
> On Thu, Aug 2, 2018 at 12:47 AM Mauro Rossi wrote:
> >
> > Hi,
> >
> > Il giorno gio 2 ago 2018 alle ore 02:06 John Stultz <
> john.stu...@linaro.org> ha scritto:
> >>
> >> Just as a heads up, with mesa3d master (plus my o
https://bugs.freedesktop.org/show_bug.cgi?id=106644
v...@endrift.com changed:
What|Removed |Added
CC||v...@endrift.com
--- Comment #42 from
79 matches
Mail list logo
