Reviewed-by: Samuel Pitoiset
On 10/5/18 2:37 AM, Dave Airlie wrote:
From: Dave Airlie
There may be some security or sandbox reason this might fail, so
check and fail appropriately.
---
src/amd/vulkan/radv_meta.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/amd
Reviewed-by: Samuel Pitoiset
On 10/5/18 1:18 AM, Dave Airlie wrote:
From: Dave Airlie
Coverity pointed out we were copying 168 bytes here unnecessarily.
---
src/amd/vulkan/radv_pipeline.c | 13 ++---
1 file changed, 6 insertions(+), 7 deletions(-)
diff --git a/src/amd/vulkan/radv_
Patches 2-3 are:
Reviewed-by: Samuel Pitoiset
On 10/5/18 2:00 AM, Dave Airlie wrote:
From: Dave Airlie
The value is always >= 0 here.
Found by coverity
---
src/amd/vulkan/radv_nir_to_llvm.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/amd/vulkan/radv_nir_to_llv
Am Donnerstag, den 04.10.2018, 10:48 -0700 schrieb Gurchetan Singh:
>
> The idea is to get rid of any adjustments on both the Mesa /
> virglrenderer sides -- so transfer size is just what's needed to
> transfer the box size, and the offset is the just offset into the
> iovec from which the transfe
https://bugs.freedesktop.org/show_bug.cgi?id=97444
Michel Dänzer changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://bugs.freedesktop.org/show_bug.cgi?id=35268
Ross Burton changed:
What|Removed |Added
CC||raj.k...@gmail.com
--
You are receiving t
I'll take care of 18.2 releases series on Andres behalf.
CC: Andres Gomez
CC: Dylan Baker
CC: Emil Velikov
---
docs/release-calendar.html | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/docs/release-calendar.html b/docs/release-calendar.html
index f92f7ba00fb..379d1
From: Michel Dänzer
Signed-off-by: Michel Dänzer
---
src/gallium/state_trackers/dri/dri2.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/src/gallium/state_trackers/dri/dri2.c
b/src/gallium/state_trackers/dri/dri2.c
index b17c5e16ede..4efc4334b65 100644
--- a/src/gallium/st
From: Michel Dänzer
Signed-off-by: Michel Dänzer
---
src/gbm/backends/dri/gbm_dri.c | 4
1 file changed, 4 insertions(+)
diff --git a/src/gbm/backends/dri/gbm_dri.c b/src/gbm/backends/dri/gbm_dri.c
index b3d6ceb15a3..35ec3a1c3a2 100644
--- a/src/gbm/backends/dri/gbm_dri.c
+++ b/src/gbm/ba
Hello,
+ i...@freedesktop.org
and add missing bugzilla link
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108247
Regards,
Andrii.
On Thu, Oct 4, 2018 at 5:08 PM wrote:
> From: Andrii Simiklit
>
> GLSL 3.7 (Identifiers):
> "However, as noted in the specification, there are some cases
Am Donnerstag, den 04.10.2018, 17:16 + schrieb Roland Scheidegger:
> I've attached the diff (no guarantees, can't test it right now).
> nopt is useful because if there's tons of shaders to compile
> compilation is quite a bit faster (but generally you really don't
> want to do this).
> Not sure
https://bugs.freedesktop.org/show_bug.cgi?id=77449
Bug 77449 depends on bug 104602, which changed state.
Bug 104602 Summary: [apitrace] Graphical artifacts in Civilization VI on RX Vega
https://bugs.freedesktop.org/show_bug.cgi?id=104602
What|Removed |Added
Mesa 18.2.2 is now available.
In this release we have:
Different patches for the DirectX9 and DRI state trackers.
A patch to implement vkAcquireNextImage2 in the Intel and AMD vulkan drivers, as
well as a patch for adding support for protected memory properties in
GetPhysicalDeviceProperties2()
https://bugs.freedesktop.org/show_bug.cgi?id=77449
Bug 77449 depends on bug 104602, which changed state.
Bug 104602 Summary: [apitrace] Graphical artifacts in Civilization VI on RX Vega
https://bugs.freedesktop.org/show_bug.cgi?id=104602
What|Removed |Added
On Fri, Oct 5, 2018 at 3:38 AM Dave Airlie wrote:
>
> From: Dave Airlie
>
> There may be some security or sandbox reason this might fail, so
> check and fail appropriately.
> ---
> src/amd/vulkan/radv_meta.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/src/amd/vulk
Reviewed-by: Lionel Landwerlin
On 05/10/2018 01:13, Dave Airlie wrote:
From: Dave Airlie
If we fail the init path then don't call the free path.
Found by coverity
---
src/vulkan/wsi/wsi_common_wayland.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/vulkan/wsi
pt., 5 paź 2018 o 13:10 Grazvydas Ignotas napisał(a):
>
> On Fri, Oct 5, 2018 at 3:38 AM Dave Airlie wrote:
> >
> > From: Dave Airlie
> >
> > There may be some security or sandbox reason this might fail, so
> > check and fail appropriately.
> > ---
> > src/amd/vulkan/radv_meta.c | 4 +++-
> > 1
On Fri, 2018-10-05 at 10:13 +1000, Dave Airlie wrote:
> From: Dave Airlie
>
> If we fail the init path then don't call the free path.
>
> Found by coverity
> ---
> src/vulkan/wsi/wsi_common_wayland.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/src/vulkan/wsi/ws
Fixes the build for
$ meson -D dri-drivers=[] -D gallium-drivers=[] -D build-tests=true
Compiling all this unused code isn't an actual problem, until you also
try to build the tests, at which point you get this:
[213/705] Linking target src/mapi/glapi/glapi_static_check_table.
FAILED: src/map
Otherwise, they are removed during NIR linking or in some
lowering passes.
Cc: Jason Ekstrand
Cc: Timothy Arceri
Signed-off-by: Samuel Pitoiset
---
src/compiler/spirv/vtn_variables.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/compiler/spirv/vtn_variables.c
b/src/compiler/spirv/vt
On Friday, 2018-10-05 14:10:13 +0200, Philipp Zabel wrote:
> On Fri, 2018-10-05 at 10:13 +1000, Dave Airlie wrote:
> > From: Dave Airlie
> >
> > If we fail the init path then don't call the free path.
> >
> > Found by coverity
> > ---
> > src/vulkan/wsi/wsi_common_wayland.c | 4 ++--
> > 1 file
From: Gert Wollny
For testing it is of interest that all tests of dEQP pass, e.g. to test
virglrenderer on a host only providing software rendering like in a CI.
Hence make it possible to disable certain optimizations that make tests fail.
While we are there also add some documentation to the fl
On 10/03/2018 01:39 AM, Vadym Shovkoplias wrote:
> From Section 6.1.2 (Subroutines) of the GLSL 4.00 specification
>
> "A program will fail to compile or link if any shader
> or stage contains two or more functions with the same
> name if the name is associated with a subroutine type
Reviewed-by: Alejandro Piñeiro
On 05/10/18 02:00, Dave Airlie wrote:
> From: Dave Airlie
>
> Found by coverity
> ---
> src/mesa/main/glspirv.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/src/mesa/main/glspirv.c b/src/mesa/main/glspirv.c
> index fecf7384eb3..972989055e9 100644
> -
On 10/03/2018 11:52 PM, Iago Toral wrote:
> On Wed, 2018-10-03 at 16:24 +0300, Vadim Shovkoplias wrote:
>> Hi Iago,
>>
>> I also think that compiler is the best place for the fix. But from my
>> understanding compiler fix will be limited to distinct shader objects
>> (not the shader stage).
>> In G
This is different from the GL_ARB_spirv pass because it generates a much
simpler data structure that isn't tied to OpenGL and mtypes.h.
---
src/compiler/Makefile.sources | 4 +-
src/compiler/nir/meson.build | 2 +
src/compiler/nir/nir_gather_xfb_info.c | 150
https://bugs.freedesktop.org/show_bug.cgi?id=108250
Bug ID: 108250
Summary: [GLSL] layout-location-struct.shader_test fails to
link
Product: Mesa
Version: 18.2
Hardware: Other
OS: All
Status: NEW
Section 7.4.1 (Shader Interface Matching) of the OpenGL 4.30 spec says:
"Variables or block members declared as structures are considered
to match in type if and only if structure members match in name,
type, qualification, and declaration order."
Fixes:
* layout-location-struc
https://bugs.freedesktop.org/show_bug.cgi?id=108250
--- Comment #1 from vadym ---
Patch: https://patchwork.freedesktop.org/patch/255084/
--
You are receiving this mail because:
You are the assignee for the bug.___
mesa-dev mailing list
mesa-dev@lists.
On 05/10/18 16:13, Jason Ekstrand wrote:
> This is different from the GL_ARB_spirv pass because it generates a much
> simpler data structure that isn't tied to OpenGL and mtypes.h.
I have just skimmed it (don't have time right now for a full check, will
take a deeper look next Monday), but FWIW, t
On Fri, Oct 5, 2018 at 10:34 AM Alejandro Piñeiro
wrote:
> On 05/10/18 16:13, Jason Ekstrand wrote:
> > This is different from the GL_ARB_spirv pass because it generates a much
> > simpler data structure that isn't tied to OpenGL and mtypes.h.
>
> I have just skimmed it (don't have time right now
Reviewed-by: Jason Ekstrand
On Fri, Oct 5, 2018 at 7:39 AM Samuel Pitoiset
wrote:
> Otherwise, they are removed during NIR linking or in some
> lowering passes.
>
> Cc: Jason Ekstrand
> Cc: Timothy Arceri
> Signed-off-by: Samuel Pitoiset
> ---
> src/compiler/spirv/vtn_variables.c | 1 +
> 1
On 05/10/18 17:44, Jason Ekstrand wrote:
> On Fri, Oct 5, 2018 at 10:34 AM Alejandro Piñeiro
> mailto:apinhe...@igalia.com>> wrote:
>
> On 05/10/18 16:13, Jason Ekstrand wrote:
> > This is different from the GL_ARB_spirv pass because it
> generates a much
> > simpler data structure
From: Boyuan Zhang
vlVaGetImage should respect the width, height, and coordinates x and y that
passed in. Therefore, pipe_box should be created with the passed in values
instead of surface width/height.
Signed-off-by: Boyuan Zhang
---
src/gallium/state_trackers/va/image.c | 9 ++---
1 fil
https://bugs.freedesktop.org/show_bug.cgi?id=108095
Michel Dänzer changed:
What|Removed |Added
Version|DRI git |git
Product|DRI
This is an improvement, but I think you need to clip the box to
1. Size of the surface
2. Size of the image
I think that there are clipping helpers available to do that (maybe
pipe_box_clip or so? I forget, check the auxiliary dir). Christian -
does that make sense to you?
Cheers,
-ilia
On Fr
https://bugs.freedesktop.org/show_bug.cgi?id=108095
--- Comment #10 from Michel Dänzer ---
Specifically, with xfwm4, loader_dri3_wait_x() is currently a no-op, because
draw->have_fake_front is false.
--
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assigne
Quoting Eric Engestrom (2018-10-05 05:19:34)
> Fixes the build for
> $ meson -D dri-drivers=[] -D gallium-drivers=[] -D build-tests=true
>
> Compiling all this unused code isn't an actual problem, until you also
> try to build the tests, at which point you get this:
>
> [213/705] Linking target
If that also fixes the problem, then yeah that makes perfect sense to me
as well.
Christian.
Am 05.10.2018 um 18:11 schrieb Ilia Mirkin:
This is an improvement, but I think you need to clip the box to
1. Size of the surface
2. Size of the image
I think that there are clipping helpers availab
On 10/04/2018 11:25 PM, Ilia Mirkin wrote:
> This happens in situations where we might do
>
> vec.wzyx[i] = ...
Ouch.
I haven't looked on the piglit list yet, but is there a piglit test for
this?
> The swizzle would get effectively ignored because of the interaction
> between how ir_assignmen
On Fri, Oct 5, 2018 at 12:55 PM Ian Romanick wrote:
>
> On 10/04/2018 11:25 PM, Ilia Mirkin wrote:
> > This happens in situations where we might do
> >
> > vec.wzyx[i] = ...
>
> Ouch.
>
> I haven't looked on the piglit list yet, but is there a piglit test for
> this?
Not yet, but will be tonigh
On 10/03/2018 08:21 PM, Alan Coopersmith wrote:
> On 10/ 3/18 03:15 PM, Roland Mainz wrote:
>> On Wed, Oct 3, 2018 at 11:51 PM Alan Coopersmith
>> wrote:
>>>
>>> Introduced-by: commit 04bdbbcab3c4862bf3f54ce60fcc1d2007776f80
Is this formatting something you need for your internal systems? The
fo
https://bugs.freedesktop.org/show_bug.cgi?id=108250
vadym changed:
What|Removed |Added
CC||vadym.shovkoplias@globallog
|
On Fri, Oct 5, 2018 at 1:04 AM Gert Wollny wrote:
>
> Am Donnerstag, den 04.10.2018, 10:48 -0700 schrieb Gurchetan Singh:
> >
> > The idea is to get rid of any adjustments on both the Mesa /
> > virglrenderer sides -- so transfer size is just what's needed to
> > transfer the box size, and the off
Otherwise building just vulkan (among other things) will build these
tests, pull in a bunch of stuff they shouldn't, and potentially fail to
compile.
---
src/mapi/glapi/meson.build| 2 +-
src/mapi/shared-glapi/meson.build | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git
Since there are no other users of the glsl compiler.
---
src/compiler/glsl/glcpp/meson.build | 2 +-
src/compiler/glsl/meson.build | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/compiler/glsl/glcpp/meson.build
b/src/compiler/glsl/glcpp/meson.build
index 769406f53
This has always been a requirement, it's just somehow been missed in the
meson build.
---
src/meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/meson.build b/src/meson.build
index af881cff70b..73146d37143 100644
--- a/src/meson.build
+++ b/src/meson.build
@@ -82,7
avoid misinterpretation of encoded immediate values in instruction and
disassembled output.
Signed-off-by: Sagar Ghuge
---
While encoding the immediate floating point values in instruction we use
values upto precision 8, but while disassembling, we print precision to
6 places, which round up the
Reviewed-by: Leo Liu
-Original Message-
From: Zhang, Boyuan
Sent: Friday, October 05, 2018 12:01 PM
To: mesa-dev@lists.freedesktop.org
Cc: Liu, Leo ; imir...@alum.mit.edu;
ckoenig.leichtzumer...@gmail.com; Zhang, Boyuan
Subject: [PATCH] st/va: use provided sizes and coords for getimage
On Tue, Oct 02, 2018 at 04:59:17PM -0700, Nanley Chery wrote:
> On Wed, Sep 26, 2018 at 04:31:02PM -0700, Nanley Chery wrote:
> > The current workaround has two issues. It causes significant slow-downs [1]
> > in
> > application startup times and uses the modified ASTC blocks for non-sampling
> >
On 10/ 5/18 10:04 AM, Ian Romanick wrote:
On 10/03/2018 08:21 PM, Alan Coopersmith wrote:
On 10/ 3/18 03:15 PM, Roland Mainz wrote:
On Wed, Oct 3, 2018 at 11:51 PM Alan Coopersmith
wrote:
Introduced-by: commit 04bdbbcab3c4862bf3f54ce60fcc1d2007776f80
Is this formatting something you need f
These patches are part of a proposed new interface for supporting variable
refresh rate via DRM properties.
It adds a new option for supporting adaptive sync to drirc along with the
implementation of notifying the window manager/DDX of the support via a window
property.
The option is enabled b
This option lets the user decide whether mesa should notify the
window manager / DDX driver that the current application is adaptive
sync capable.
It's off by default.
Signed-off-by: Nicholas Kazlauskas
---
src/util/xmlpool/t_options.h | 5 +
1 file changed, 5 insertions(+)
diff --git a/sr
Some programs start with the path and command line arguments in
argv[0] (program_invocation_name). Chromium is an example of
an application using mesa that does this.
This tries to query the real path for the symbolic link /proc/self/exe
to find the program name instead. It only uses the realpath
Applications that don't present at a predictable rate (ie. not games)
shouldn't have adapative sync enabled. This list covers some of the
common desktop compositors, web browsers and video players.
Signed-off-by: Nicholas Kazlauskas
---
src/util/00-mesa-defaults.conf | 79 +++
The DDX driver can be notified of adaptive sync suitability by
flagging the application's window with the _VARIABLE_REFRESH property.
This property is set on the first swap the application performs
when adaptive_sync_enable is set to true in the drirc.
It's performed here instead of when the load
It's better to let most applications make use of adaptive sync
by default. Problematic applications can be placed on the blacklist
or the user can manually disable the feature.
Signed-off-by: Nicholas Kazlauskas
---
src/gallium/drivers/radeonsi/driinfo_radeonsi.h | 4
1 file changed, 4 inse
Whiskey Lake uses the same gen graphics as Coffe Lake, including some
ids that were previously marked as reserved on Coffe Lake, but that
now are moved to WHL page.
This follows the ids and approach used on kernel's commit
b9be78531d27 ("drm/i915/whl: Introducing Whiskey Lake platform")
and commit
Reviewed-by: Marek Olšák
Marek
On Thu, Oct 4, 2018 at 8:55 PM Dave Airlie wrote:
>
> From: Dave Airlie
>
> Found by coverity
> ---
> src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
> b/src/mesa/stat
Reviewed-by: Marek Olšák
Marek
On Fri, Oct 5, 2018 at 6:22 AM Michel Dänzer wrote:
>
> From: Michel Dänzer
>
> Signed-off-by: Michel Dänzer
> ---
> src/gallium/state_trackers/dri/dri2.c | 13 +
> 1 file changed, 13 insertions(+)
>
> diff --git a/src/gallium/state_trackers/dri/dri2
Reviewed-by: Marek Olšák
Marek
On Fri, Oct 5, 2018 at 6:23 AM Michel Dänzer wrote:
>
> From: Michel Dänzer
>
> Signed-off-by: Michel Dänzer
> ---
> src/gbm/backends/dri/gbm_dri.c | 4
> 1 file changed, 4 insertions(+)
>
> diff --git a/src/gbm/backends/dri/gbm_dri.c b/src/gbm/backends/dri
Looks alright to me. I'm not quite sold on the "safemath" name though,
since "safe math" is usually associated with floating point
optimizations, and this here is just filtering hacks. Maybe something
like disable_filtering_hacks would be more fitting?
Reviewed-by: Roland Scheidegger
On 10/05
On 10/04/2018 07:08 AM, asimiklit.w...@gmail.com wrote:
> From: Andrii Simiklit
>
> GLSL 3.7 (Identifiers):
> "However, as noted in the specification, there are some cases where
> previously declared variables can be redeclared to change or add some
> property, and predeclared "gl_" names are all
On 10/05/2018 03:02 PM, Ian Romanick wrote:
> On 10/04/2018 07:08 AM, asimiklit.w...@gmail.com wrote:
>> From: Andrii Simiklit
>>
>> GLSL 3.7 (Identifiers):
>> "However, as noted in the specification, there are some cases where
>> previously declared variables can be redeclared to change or add so
v2: check for lstat() failing
Fixes: 04bdbbcab3c "xmlconfig: read more config files from drirc.d/"
Signed-off-by: Alan Coopersmith
Reviewed-by: Roland Mainz
Reviewed-by: Ian Romanick
---
src/util/xmlconfig.c | 8
1 file changed, 8 insertions(+)
diff --git a/src/util/xmlconfig.c b/src
Thanks. I pushed the series.
Please see also this commit:
https://cgit.freedesktop.org/mesa/mesa/commit/?id=86f004bdfcc1c14ace99f5fccf540c0d813d2254
I added it because the logic was broken for 64-bit reg_saved.
Marek
On Wed, Oct 3, 2018 at 11:55 AM Sonny Jiang wrote:
>
> Signed-off-by: Sonny Ji
Series is
Reviewed-by: Ian Romanick
On 10/03/2018 06:12 AM, Emil Velikov wrote:
> From: Emil Velikov
>
> Write down both X and GLX visual types when mapping from one to the
> other. Makes grepping through the code a tiny bit easier.
>
> Signed-off-by: Emil Velikov
> ---
> src/glx/glxext.c |
Reviewed-by: Ian Romanick
On 09/25/2018 07:04 AM, Tapani Pälli wrote:
> v2: check types of both operands (Ian)
>
> Cc: mesa-sta...@lists.freedesktop.org
> Signed-off-by: Tapani Pälli
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108012
> ---
> src/compiler/glsl/ast_to_hir.cpp | 6 ++
On Tuesday, October 2, 2018 11:06:23 AM PDT Chris Wilson wrote:
> Reuse the same query object buffer for multiple queries within the same
> batch.
>
> A task for the future is propagating the GL_NO_MEMORY errors.
>
> Signed-off-by: Chris Wilson
> Cc: Kenneth Graunke
> Cc: Matt Turner
> ---
>
On 10/02/2018 11:06 AM, Chris Wilson wrote:
> If we know the bo is idle (that is we have no submitted a command buffer
> referencing this bo since the last query) we can skip asking the kernel.
> Note this may report a false negative if the target is being shared
> between processes (exported via d
Patches 2 and 3 are
Reviewed-by: Ian Romanick
On 10/02/2018 11:06 AM, Chris Wilson wrote:
> Lots of places open-coded the assumed layout of the predicate/results
> within the query object, replace those with simple helpers.
>
> v2: Fix function decl style.
>
> Signed-off-by: Chris Wilson
> Cc
On 10/02/2018 11:06 AM, Chris Wilson wrote:
> If we always write the 'available' flag after writing the final result
> of the query, we can probe that predicate to quickly query whether the
> result is ready from userspace. The primary advantage of checking the
> predicate is that it allows for mor
From: Marek Olšák
Compilers can use this to generate optimal code for integer division
by a constant.
Additionally, an unsigned division by a uniform that is constant but not
known at compile time can still be optimized by passing 2-4 division
factors to the shader as uniforms and executing one
From: Marek Olšák
---
src/util/fast_idiv_by_const.c | 21 +
src/util/fast_idiv_by_const.h | 5 +++--
2 files changed, 24 insertions(+), 2 deletions(-)
diff --git a/src/util/fast_idiv_by_const.c b/src/util/fast_idiv_by_const.c
index 65a9e640789..7b93316268c 100644
--- a/src/
There's nothing inherently fixed-width in the code. All that's required
to generalize it is to make everything internally 64-bit and pass
UINT_BITS in as a parameter to util_compute_fast_[us]div_info. With
that, it can now handle 8, 16, 32, and 64-bit integer division by a
constant.
We also add
---
src/util/Makefile.am | 1 +
src/util/Makefile.sources | 1 +
src/util/bigmath.h| 112 ++
src/util/meson.build | 1 +
4 files changed, 115 insertions(+)
create mode 100644 src/util/bigmath.h
diff --git a/src/util/Makefile.am b/src/ut
While I generally trust rediculousfish to have done his homework, we've
made some adjustments to suite the needs of mesa and it'd be good to
test those. Also, there's no better place than unit tests to clearly
document the different edge cases of the different methods.
---
configure.ac
From: Marek Olšák
---
src/util/fast_idiv_by_const.h | 44 +++
1 file changed, 44 insertions(+)
diff --git a/src/util/fast_idiv_by_const.h b/src/util/fast_idiv_by_const.h
index ac10cf79ba8..1ba9f9a20b8 100644
--- a/src/util/fast_idiv_by_const.h
+++ b/src/util/fast
It's reasonably well-known that you can replace an unsigned integer
division by a constant with a sequence of multiply (by constant), adds, and
shifts. This little series implements such an optimization for NIR. This
second version of the series is based on the integer division by constant
code b
---
src/compiler/nir/nir_constant_expressions.py | 1 +
src/compiler/nir/nir_opcodes.py | 43 ++--
2 files changed, 40 insertions(+), 4 deletions(-)
diff --git a/src/compiler/nir/nir_constant_expressions.py
b/src/compiler/nir/nir_constant_expressions.py
index 118af9
---
src/compiler/nir/nir.h | 1 +
src/compiler/nir/nir_lower_int64.c | 54 ++
2 files changed, 55 insertions(+)
diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h
index e0df95c391c..9bb9f50b9b4 100644
--- a/src/compiler/nir/nir.h
+++ b/src/compil
From: Ian Romanick
---
src/intel/compiler/brw_fs_nir.cpp | 5 +
1 file changed, 5 insertions(+)
diff --git a/src/intel/compiler/brw_fs_nir.cpp
b/src/intel/compiler/brw_fs_nir.cpp
index 7f453d75b64..f042eac432d 100644
--- a/src/intel/compiler/brw_fs_nir.cpp
+++ b/src/intel/compiler/brw_fs_n
Shader-db results on Sky Lake:
total instructions in shared programs: 15105795 -> 15111403 (0.04%)
instructions in affected programs: 72774 -> 78382 (7.71%)
helped: 0
HURT: 265
Note that hurt here actually means helped because we're getting rid of
integer quotient operations (whic
From: Ian Romanick
---
src/compiler/nir/nir_opcodes.py | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/compiler/nir/nir_opcodes.py b/src/compiler/nir/nir_opcodes.py
index 209f0c5509b..1a52b3b4228 100644
--- a/src/compiler/nir/nir_opcodes.py
+++ b/src/compiler/nir/nir_opcodes.py
@@ -437
It's a reasonably well-known fact in the world of compilers that integer
divisions by constants can be replaced by a multiply, an add, and some
shifts. This commit adds such an optimization to NIR for easiest case
of udiv. Other division operations will be added in following commits.
In order to
This happens in situations where we might do
vec.wzyx[i] = ...
The swizzle would get effectively ignored because of the interaction
between how ir_assignment->set_lhs works and overwriting the write_mask.
There are two cases, one where i is a constant, and another where i is
variable. We have t
On Sat, Oct 6, 2018 at 1:24 AM Ilia Mirkin wrote:
>
> This happens in situations where we might do
>
> vec.wzyx[i] = ...
>
> The swizzle would get effectively ignored because of the interaction
> between how ir_assignment->set_lhs works and overwriting the write_mask.
> There are two cases, one
https://bugs.freedesktop.org/show_bug.cgi?id=108135
--- Comment #6 from Kenneth Graunke ---
This does point out an even more serious issue though:
Megadrivers interacts very badly with drivers that have global C++
initializers. When dlopen'ing a DRI driver, C++ global constructors will run
for
88 matches
Mail list logo
