On Thu, Sep 4, 2014 at 10:19 PM, Jason Ekstrand ja...@jlekstrand.net wrote:
This is the first chunk of patches in my work on adding instruction and
register widths to the fs backend. Eventually, this will allow us to more
easily emit 8-wide instructions in SIMD16 mode from the fs_visitor
https://bugs.freedesktop.org/show_bug.cgi?id=78773
Tapani Pälli lem...@gmail.com changed:
What|Removed |Added
Status|NEW |RESOLVED
On Fri, Jul 18, 2014 at 4:35 PM, Matt Turner matts...@gmail.com wrote:
On Tue, Jul 15, 2014 at 10:37 AM, Matt Turner matts...@gmail.com wrote:
On Mon, Jul 14, 2014 at 3:48 PM, Ian Romanick i...@freedesktop.org wrote:
Most of these patches have been sent to the list already in one form or
Jan Vesely jan.ves...@rutgers.edu writes:
It's unsafe as compat::string is not null terminated.
Fixes garbage log on successful build.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
It's obviously not needed in program::build, but I'm not sure
if runtime_error solution is the best
On mié, 2014-09-03 at 18:51 -0700, Jordan Justen wrote:
Rather than:
i965/gen6/gs: refactor gen6_gs_state
How about something like:
i965/gen6/gs: Skeleton for user GS program support
(more below)
On Thu, Aug 14, 2014 at 4:11 AM, Iago Toral Quiroga ito...@igalia.com wrote:
From:
Emil Velikov emil.l.veli...@gmail.com writes:
There is a recent bug in piglit (+ it's fix [1]) that causes the env vars used
not to be printed in the HTML test summary.
That would be a good bug fix. But the problem I had was that piglit
wasn't running this test at all, (so I wasn't even getting
On Thu, Sep 04, 2014 at 05:15:29AM +0100, Dave Airlie wrote:
This adds the guts of the fp64 implementation to the GLSL compiler.
- builtin double types
- double constant support
- lexer parsing for double types (lf, LF)
- enforcing flat on double fs inputs
- double operations (d2f,f2d,
On Fri, Sep 05, 2014 at 10:25:52AM +0300, Pohjolainen, Topi wrote:
On Thu, Sep 04, 2014 at 05:15:29AM +0100, Dave Airlie wrote:
This adds the guts of the fp64 implementation to the GLSL compiler.
- builtin double types
- double constant support
- lexer parsing for double types (lf, LF)
Instead we store a void pointer to the key, and cast it to
brw_wm_prog_key for fragment shader specific code paths.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 2 +-
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 10 +-
2
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/mesa/drivers/dri/i965/brw_context.h| 1 +
src/mesa/drivers/dri/i965/brw_fs.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 12 ++--
This matches backend_visitor, and will allow gl_program to be named prog.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 4 ++--
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 10 +-
2 files changed, 7 insertions(+), 7
Instead we store a brw_stage_prog_data pointer, and cast it to
brw_wm_prog_data for fragment shader specific code paths.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 2 +-
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 37
This will allow for stage specific code paths.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
git://people.freedesktop.org/~jljusten/mesa for-upstream
No piglit regressions seen on gen7.
src/mesa/drivers/dri/i965/brw_fs.h | 1 +
gl_program* is named prog similar to backend_visitor.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 2 +-
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 13 +++--
2 files changed, 8 insertions(+), 7 deletions(-)
diff --git
On Tue, Sep 02, 2014 at 09:34:13PM -0700, Matt Turner wrote:
---
src/mesa/drivers/dri/i965/brw_fs.cpp | 30 +-
src/mesa/drivers/dri/i965/brw_fs.h | 6 --
2 files changed, 21 insertions(+), 15 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp
On Tue, Sep 02, 2014 at 09:34:14PM -0700, Matt Turner wrote:
---
src/mesa/drivers/dri/i965/brw_vec4.cpp | 23 ---
1 file changed, 16 insertions(+), 7 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_vec4.cpp
b/src/mesa/drivers/dri/i965/brw_vec4.cpp
index
On Thu, Sep 4, 2014 at 11:17 PM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Thursday, September 04, 2014 04:16:47 PM Brian Paul wrote:
On 09/04/2014 03:17 PM, Kenneth Graunke wrote:
On Wednesday, September 03, 2014 06:50:39 PM Carl Worth wrote:
Carl Worth cwo...@cworth.org writes:
Just drop the conditional and simplify our build. This means that
it'll build every time, but it does not require any dependencies nor
does it take that long to compile 200 lines of boilerplate code.
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
configure.ac| 2 --
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
configure.ac | 1 +
1 file changed, 1 insertion(+)
diff --git a/configure.ac b/configure.ac
index d7511d7..e2ba485 100644
--- a/configure.ac
+++ b/configure.ac
@@ -1960,6 +1960,7 @@ if test -n $with_gallium_drivers; then
fi
With recent commit we removed the NEED_NONNULL_WINSYS checks when
selecting the hardware (inc svga) winsys. svga has only one winsys
that explicitly requires libdrm (via it's bundled version of
vmwgfx_drm.h) but configure.ac never really checks for it.
Add the check early to prevent people from
The variable was unused and gave false information. The need for nonnull
winsys currently does not relate as it used to. Nowadays one can mix and
match more freely with plenty of winsys' to make your head spin.
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
configure.ac | 54
Running fast clear glClear with SNB caused Valgrind to
complain about this.
v2: line 237 fixed glClear from leaking memory, other
strdups are also now changed to ralloc_strdups but I
don't know what effect those have. At least no changes in
my Piglit quick run.
Signed-off-by: Juha-Pekka Heikkila
On 03/09/14 23:21, Matt Turner wrote:
We're up to 16 pages of patches. Please take a look at your patches
and clear out the old ones.
http://patchwork.freedesktop.org/project/mesa/list/
If you're not an administrator, I believe you can still manage your
own patches. If you'd like to be an
On 01/09/14 17:08, Jonathan Gray wrote:
As older versions of gnu ld did not support --dynamic-list check to see
if it is supported before using it. Non gnu linkers such the apple one
likely lack this option as well.
Fixes the build on OpenBSD which has binutils 2.15 and 2.17.
The
I've cleaned out mine.
There's an 11 patch series from 7th of August that's left.
It's a v2/v3 of the algebraic optimizations I did back in July.
Not all patches gave improvements, but Eric was interested:
I'm definitely interested in seeing our optimizer gain features
like this, even if we
On Friday, September 05, 2014 09:30:52 AM Francisco Jerez wrote:
Jan Vesely jan.ves...@rutgers.edu writes:
It's unsafe as compat::string is not null terminated.
Fixes garbage log on successful build.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
It's obviously not needed in
On Mon, 2014-08-18 at 09:20 +0300, Francisco Jerez wrote:
EdB edb+m...@sigluy.net writes:
compact::string is not \0 terminated.
size() need to be used for std::string creation
---
src/gallium/state_trackers/clover/core/program.cpp | 4 ++--
1 file changed, 2 insertions(+), 2
On Fri, 2014-09-05 at 15:51 +0200, EdB wrote:
On Friday, September 05, 2014 09:30:52 AM Francisco Jerez wrote:
Jan Vesely jan.ves...@rutgers.edu writes:
It's unsafe as compat::string is not null terminated.
Fixes garbage log on successful build.
Signed-off-by: Jan Vesely
Hi guys!
Little recap on my project. I don't know if've been there recently but
there has been some nice changes (thanks for the llvmpipe and softpipe
cleanup by the way).
http://creak.foolstep.com/mesamatrix/
I'm wondering if something could be done for the r300 status. I think it
will
On 31/08/14 19:33, Matt Turner wrote:
Despite the comment above the function claiming otherwise, the function
did not reswizzle sources, which would lead to bad code generation.
Hi Matt,
Has this patch been superseded by another one or just missing review ?
Thanks
Emil
Bugzilla:
On 27/08/14 08:34, Andreas Pokorny wrote:
Hi,
This one liner should apply on top of master and the two release branches. It
fixes 'black windows' when importing prime fds on gpus using the i915 driver
paths.
Hello gents,
Can anyone review this trivial patch ?
Thanks
Emil
regards
A lot of GL3 and GL4 features actually only require GL2 hardware and
r300 should support them.
Marek
On Fri, Sep 5, 2014 at 4:57 PM, Romain Failliot
romain.faill...@foolstep.com wrote:
Hi guys!
Little recap on my project. I don't know if've been there recently but there
has been some nice
You can skip glx tests using these piglit-run parameters: -x glx -x makeCurrent
That should prevent X from crashing.
Marek
On Fri, Sep 5, 2014 at 7:07 AM, Alexandre Demers
alexandre.f.dem...@gmail.com wrote:
I've been testing your patch v2 on cayman.
Here are my results. For now, my piglit
Hmm the previous cover letter was not attached to any patches :\
Anywho, can someone take a few minutes to check this out ?
-Emil
On 27/08/14 08:36, Andreas Pokorny wrote:
Width and Height of the imported image was never initialized from the
imported bo.
Signed-off-by: Andreas Pokorny
https://bugs.freedesktop.org/show_bug.cgi?id=82882
Marek Olšák mar...@gmail.com changed:
What|Removed |Added
Status|NEW |RESOLVED
On Friday, September 05, 2014 12:44:12 AM Jordan Justen wrote:
This will allow for stage specific code paths.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
git://people.freedesktop.org/~jljusten/mesa for-upstream
No piglit regressions seen on gen7.
On Sunday, August 31, 2014 11:33:58 AM Matt Turner wrote:
Despite the comment above the function claiming otherwise, the function
did not reswizzle sources, which would lead to bad code generation.
Perhaps expand this to something like:
which lead to incorrect code generation since commit
In LLVM 3.4.1 and 3.4.2 the patch level was only set in config.h when
LLVM was built via autoconf and not when it was built with cmake.
Fall back to retrieving the patch level from llvm-config --version to
handle this case.
Cc: 10.2 10.3 mesa-sta...@lists.freedesktop.org
Signed-off-by: Jonathan
Mesa already defines _GNU_SOURCE for glibc based systems and defining
_GNU_SOURCE will break the Mesa build on other systems such as OpenBSD.
_GNU_SOURCE only seems to be included in llvm-config output when
LLVM is built via autoconf and not when it is built by cmake.
Cc: 10.2 10.3
On Sunday, August 31, 2014 11:34:11 AM Matt Turner wrote:
Before commit 04895f5c we would only reswizzle dot product instructions
(since they wrote the same value into all channels, and we didn't have
to think about anything else). That commit extended reswizzling to cases
when the swizzle was
On Thu, Sep 4, 2014 at 11:03 PM, Matt Turner matts...@gmail.com wrote:
On Thu, Sep 4, 2014 at 10:19 PM, Jason Ekstrand ja...@jlekstrand.net
wrote:
This is the first chunk of patches in my work on adding instruction and
register widths to the fs backend. Eventually, this will allow us to
Mesa 10.3 release candidate 3 is now available for testing. The current
plan of record is to have an additional release candidate each Friday
until the 10.3 release on Friday, September 12th.
The tag in the GIT repository for Mesa 10.3-rc3 is 'mesa-10.3-rc3'. I
have verified that the tag is in
https://bugs.freedesktop.org/show_bug.cgi?id=81680
--- Comment #32 from Marek Olšák mar...@gmail.com ---
Created attachment 105815
-- https://bugs.freedesktop.org/attachment.cgi?id=105815action=edit
hack
Could you please test this patch?
--
You are receiving this mail because:
You are the
https://bugs.freedesktop.org/show_bug.cgi?id=81680
--- Comment #33 from Eugene ken20...@ukr.net ---
Sorry, but again I don't know how. Do I need to compile MESA from source to try
your patch ?
--
You are receiving this mail because:
You are the assignee for the bug.
https://bugs.freedesktop.org/show_bug.cgi?id=81680
--- Comment #34 from Marek Olšák mar...@gmail.com ---
Yes.
I cannot reproduce this bug, so there's nothing else I can do.
--
You are receiving this mail because:
You are the assignee for the bug.
___
This patch has basically already been committed, so ignore it here.
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
On 05/09/14 16:43, Jonathan Gray wrote:
In LLVM 3.4.1 and 3.4.2 the patch level was only set in config.h when
LLVM was built via autoconf and not when it was built with cmake.
Fall back to retrieving the patch level from llvm-config --version to
handle this case.
Hi Jonathan,
Can you please
So right now, if it's red it really means that the r300 could do it, but
it's not been implemented yet?
2014-09-05 11:12 GMT-04:00 Marek Olšák mar...@gmail.com:
A lot of GL3 and GL4 features actually only require GL2 hardware and
r300 should support them.
Marek
On Fri, Sep 5, 2014 at
On 05/09/14 16:44, Jonathan Gray wrote:
Mesa already defines _GNU_SOURCE for glibc based systems and defining
_GNU_SOURCE will break the Mesa build on other systems such as OpenBSD.
_GNU_SOURCE only seems to be included in llvm-config output when
LLVM is built via autoconf and not when it is
IMHO all the red blocks should stand for not implemented. Whether $driver
could do it or not that is highly individual and depends on the
feature/extension and the actual hardware.
-Emil
P.S. I don't work on the radeon drivers.
On 05/09/14 20:54, Romain Failliot wrote:
So right now, if it's red
Like Emil said, it depends on the feature. For example, the r300
hardware can do GL_ARB_clip_control from GL 4.5, but cannot do
GL_EXT_texture_array from GL 3.0. The driver doesn't support
GL_ARB_clip_control yet.
Marek
On Fri, Sep 5, 2014 at 9:54 PM, Romain Failliot
romain.faill...@foolstep.com
https://bugs.freedesktop.org/show_bug.cgi?id=77449
Bug 77449 depends on bug 74700, which changed state.
Bug 74700 Summary: Dota2 shimmering artifacts on Mesa 10.0.1 on Iris Pro
graphics
https://bugs.freedesktop.org/show_bug.cgi?id=74700
What|Removed |Added
https://bugs.freedesktop.org/show_bug.cgi?id=81680
yas...@windowslive.com changed:
What|Removed |Added
CC||yas...@windowslive.com
--
You
https://bugs.freedesktop.org/show_bug.cgi?id=80848
--- Comment #24 from Emil Velikov emil.l.veli...@gmail.com ---
Gents just pushed a few commits which explicitly checks for xcb and links the
relevant libraries against it. This is required as soon (tm) libxcb will remote
it's Requires to
The scale factors for the resolve rectangle change for BDW and we have
to look at bdw-gen now to figure out how big it should be.
Fixes: https://bugs.freedesktop.org/attachment.cgi?id=105777
Cc: 10.3 mesa-sta...@lists.freedesktop.org
Signed-off-by: Kristian Høgsberg k...@bitplanet.net
https://bugs.freedesktop.org/show_bug.cgi?id=81680
--- Comment #35 from yas...@windowslive.com ---
(In reply to comment #34)
Yes.
I cannot reproduce this bug, so there's nothing else I can do.
Steps to reproduce the bug:
-Set layers.acceleration.force-enabled to true
-Set
I'm not sure if this is correct or not I'll leave that for someone else
to comment on, but there doesn't seem to be any reason for you to move
the location of the code. Especially the GL_KHR_debug (GL 4.3)/
GL_ARB_debug_output stuff this should only need to be a one line patch.
On Sun, 2014-08-31
Uniform values are in the UNIFORM register file, not the GRF register file.
Looking in virtual_grf_sizes makes no sense and only makes the output of
dump_instructions confusint.
Signed-off-by: Jason Ekstrand jason.ekstr...@intel.com
---
src/mesa/drivers/dri/i965/brw_fs.cpp | 3 +--
1 file
Mesa 10.2.7 has been released. Mesa 10.2.7 is a bug fix release
fixing bugs since the 10.2.6 release, (see below for a list of
changes).
The tag in the git repository for Mesa 10.2.7 is 'mesa-10.2.7'.
Mesa 10.2.7 is available for download at
ftp://freedesktop.org/pub/mesa/10.2.7/
SHA-256
On Fri, Sep 5, 2014 at 5:09 PM, Jason Ekstrand ja...@jlekstrand.net wrote:
Uniform values are in the UNIFORM register file, not the GRF register file.
Looking in virtual_grf_sizes makes no sense and only makes the output of
dump_instructions confusint.
typo
Signed-off-by: Jason Ekstrand
On Fri, Sep 5, 2014 at 5:25 PM, Matt Turner matts...@gmail.com wrote:
On Fri, Sep 5, 2014 at 5:09 PM, Jason Ekstrand ja...@jlekstrand.net
wrote:
Uniform values are in the UNIFORM register file, not the GRF register
file.
Looking in virtual_grf_sizes makes no sense and only makes the output
On 06/09/14 09:33, Timothy Arceri wrote:
I'm not sure if this is correct or not I'll leave that for someone else
to comment on, but there doesn't seem to be any reason for you to move
the location of the code. Especially the GL_KHR_debug (GL 4.3)/
GL_ARB_debug_output stuff this should only
In commit 567e2769b81863b6dffdac3826a6b729ce6ea37c (ra: make the p, q
test more efficient) I unknowingly introduced a new requirement to the
register allocator API: the user must set the register class of all
nodes before setting up their interferences, because
ra_add_conflict_list() now uses the
As noted in the previous commit, this was introduced in
567e2769b81863b6dffdac3826a6b729ce6ea37c (ra: make the p, q test more
efficient), but I forgot to mention it.
Signed-off-by: Connor Abbott cwabbo...@gmail.com
---
src/mesa/program/register_allocate.h | 5 -
1 file changed, 4
q_total should never go below 0 (which is why it's defined as unsigned),
and if it does, then something is seriously wrong.
Signed-off-by: Connor Abbott cwabbo...@gmail.com
---
src/mesa/program/register_allocate.c | 1 +
1 file changed, 1 insertion(+)
diff --git
https://bugs.freedesktop.org/show_bug.cgi?id=83549
Priority: medium
Bug ID: 83549
Keywords: regression
CC: emil.l.veli...@gmail.com
Assignee: mesa-dev@lists.freedesktop.org
Summary: nv50/nv50_draw.c:79:25: error: 'struct
Fixes commit b6e9b54d066d29d250c4502e0005b317589dd8b3 .
Signed-off-by: David Heidelberger david.heidelber...@ixit.cz
---
src/mesa/program/register_allocate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/mesa/program/register_allocate.c
Thanks Marek, you were right.
So, on cayman, after comparing results from both runs, it seems there is
only one regression which is gs-atan-vec2 under glsl-1.50:
piglit/bin/shader_runner
On Fri, Sep 05, 2014 at 08:48:17PM +0100, Emil Velikov wrote:
On 05/09/14 16:43, Jonathan Gray wrote:
In LLVM 3.4.1 and 3.4.2 the patch level was only set in config.h when
LLVM was built via autoconf and not when it was built with cmake.
Fall back to retrieving the patch level from
On Fri, Sep 05, 2014 at 08:59:30PM -0400, Connor Abbott wrote:
In commit 567e2769b81863b6dffdac3826a6b729ce6ea37c (ra: make the p, q
test more efficient) I unknowingly introduced a new requirement to the
register allocator API: the user must set the register class of all
nodes before setting
On Sat, 2014-09-06 at 10:52 +1000, Matthew Waters wrote:
On 06/09/14 09:33, Timothy Arceri wrote:
I'm not sure if this is correct or not I'll leave that for someone else
to comment on, but there doesn't seem to be any reason for you to move
the location of the code. Especially the
https://bugs.freedesktop.org/show_bug.cgi?id=83549
Ilia Mirkin imir...@alum.mit.edu changed:
What|Removed |Added
Status|NEW |RESOLVED
On Fri, Sep 05, 2014 at 09:00:55PM +0100, Emil Velikov wrote:
On 05/09/14 16:44, Jonathan Gray wrote:
Mesa already defines _GNU_SOURCE for glibc based systems and defining
_GNU_SOURCE will break the Mesa build on other systems such as OpenBSD.
_GNU_SOURCE only seems to be included in
73 matches
Mail list logo
