Chad Versace (2):
xf86-video-intel
dri: Do not tile stencil buffer
src/intel_dri.c | 16
mesa
intel: Fix stencil buffer to be W tiled
src/mesa/drivers/dri/intel/intel_clear.c |6 ++
src/mesa/drivers/dri/intel/intel_context.c |9 ++-
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM states that it is. However, it is actually
W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
4.5.2.1 W-Major Format:
W-Major Tile Format is used for separate stencil.
The GTT
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM states that it is. However, it is actually
W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
4.5.2.1 W-Major Format:
W-Major Tile Format is used for separate stencil.
The GTT
On 14/07/2011 02:05, Marek Olšák wrote:
> Module: Mesa
> Branch: master
> Commit: 02c8ee202f5a159659a027deae4721ff05cd1530
> URL:
> http://cgit.freedesktop.org/mesa/mesa/commit/?id=02c8ee202f5a159659a027deae4721ff05cd1530
>
> Author: Marek Olšák
> Date: Mon Jun 27 08:02:31 2011 +0200
>
> c
On 07/17/2011 03:50 PM, Marty Jack wrote:
Another problem in the RC1 tarballs appears to be src/mesa/depend. This
contains a pile of references to /usr/lib/gcc/x86_64-redhat-linux/4.6.0/include
and will fail if you are not on that system.
With that removed I am able to build the RC1 tarballs,
On 07/15/2011 02:59 PM, Pekka Paalanen wrote:
On Fri, 15 Jul 2011 12:22:41 -0600
Brian Paul wrote:
On 07/15/2011 10:07 AM, Dave Airlie wrote:
On Fri, Jul 15, 2011 at 4:10 AM, Brian
Paul wrote:
The map-texture-image-v4 branch that I just pushed implements
this change. It really cleaned t
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39286
---
src/mesa/state_tracker/st_cb_texture.c |8
1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/src/mesa/state_tracker/st_cb_texture.c
b/src/mesa/state_tracker/st_cb_texture.c
index 6907cfc..63cd142 100644
--- a/s
On Mon, Jul 18, 2011 at 8:11 AM, Vadim Girlin wrote:
> Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39286
> ---
> src/mesa/state_tracker/st_cb_texture.c | 8
> 1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/src/mesa/state_tracker/st_cb_texture.c
> b/src/mesa/
https://bugs.freedesktop.org/show_bug.cgi?id=39338
Summary: Mesa: LLVM dependency mishmash
Product: Mesa
Version: git
Platform: All
OS/Version: All
Status: NEW
Severity: blocker
Priority: medium
Component:
Commit 1a339b6c71ebab6e1a64f05b2e133022d3bbcd15 made
st_ChooseTextureFormat map GL_RGBA with type GL_UNSIGNED_BYTE
to PIPE_FORMAT_A8B8G8R8_UNORM.
The image format for ARGB pixmaps is PIPE_FORMAT_B8G8R8A8_UNORM
however. This mismatch caused the texture to be recreated in
st_finalize_texture.
NOTE:
On Mon, 2011-07-18 at 08:34 -0600, Brian Paul wrote:
> On Mon, Jul 18, 2011 at 8:11 AM, Vadim Girlin wrote:
> > Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39286
> > ---
> > src/mesa/state_tracker/st_cb_texture.c |8
> > 1 files changed, 8 insertions(+), 0 deletions(-)
> >
> >
We can't do try-map + create + map + dereference internally in a
driver. Creating a new buffer and replacing a pointer to the old one
may lead to the following issue. If a buffer pointer is replaced, it
doesn't necessarily update all the states the buffer is set in in all
existing contexts. Such st
https://bugs.freedesktop.org/show_bug.cgi?id=39338
--- Comment #1 from Droste 2011-07-18 10:10:50 PDT ---
Created an attachment (id=49256)
View: https://bugs.freedesktop.org/attachment.cgi?id=49256
Review: https://bugs.freedesktop.org/review?bug=39338&attachment=49256
Build fix with llvm 3.0sv
Improves glxgears performance 19.6% +/- 7.3% (second fps printout.
n=5).
---
src/mesa/drivers/dri/intel/intel_clear.c |5 -
1 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/src/mesa/drivers/dri/intel/intel_clear.c
b/src/mesa/drivers/dri/intel/intel_clear.c
index dfca03c..35a
On Mon, 18 Jul 2011 00:55:03 -0700, Chad Versace wrote:
> diff --git a/src/mesa/drivers/dri/intel/intel_fbo.c
> b/src/mesa/drivers/dri/intel/intel_fbo.c
> index 1669af2..507cc33 100644
> --- a/src/mesa/drivers/dri/intel/intel_fbo.c
> +++ b/src/mesa/drivers/dri/intel/intel_fbo.c
> @@ -173,6 +173,9
This cuts out a large portion of the overhead of glClear() from
resetting the texenv state and recomputing the fixed function
programs. It also means less use of fixed function internally in our
GLES2 drivers, which is rather bogus.
---
src/mesa/drivers/common/meta.c | 166 ++
The classic nouveau driver supports nv04, nv1x, nv2x IIRC. I'm
definitely not the right person for the job, but I'll look at it if
nobody else can.
~ C.
On Mon, Jul 18, 2011 at 7:09 AM, Brian Paul wrote:
> On 07/15/2011 02:59 PM, Pekka Paalanen wrote:
>>
>> On Fri, 15 Jul 2011 12:22:41 -0600
>>
On Wed, 13 Jul 2011 13:51:44 -0700, Ben Widawsky wrote:
> The debugger shared memory needs to be a fixed size. Since this is
> scratch memory that is already used by register spilling, add
> appropriate hooks to do the right thing when debugging.
>
> v2: Include the bytes for a known good system
On Wed, 13 Jul 2011 16:17:51 -0700, Chad Versace wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> On 07/12/2011 03:22 PM, Eric Anholt wrote:
> > ---
> > src/mesa/drivers/dri/i915/i830_vtbl.c | 46
> > +++-
> > src/mesa/drivers/dri/i915/i915_vtbl.c | 4
On Sun, 17 Jul 2011 16:25:42 -0700, Ben Widawsky wrote:
> Upload the system routine as part of the invariant state if debugging.
>
> Remove SIP setting if not debugging to make it more friendly for others
> that may be debugging shaders or media kernels.
>
> v2: removed comment per Chris
This p
On Mon, 18 Jul 2011 09:19:32 +0800, "Zou, Nanhai" wrote:
> >>-Original Message-
> >>From: Eric Anholt [mailto:e...@anholt.net]
> >>Sent: 2011年7月17日 8:08
> >>To: Zou, Nanhai; mesa-dev@lists.freedesktop.org
> >>Subject: Re: [Mesa-dev] [PATCH] i965: fix timer query on gen6+
> >>
> >>On Fri, 1
Hi,
On Monday, July 18, 2011 18:28:26 Marek Olšák wrote:
> We can't do try-map + create + map + dereference internally in a
> driver. Creating a new buffer and replacing a pointer to the old one
> may lead to the following issue. If a buffer pointer is replaced, it
> doesn't necessarily update al
This is a re-send of the unit tests for lower_jumps.cpp which I sent
to the mailing list on July 8. I haven't received any comments yet,
and since this is a rewrite of some patches that proved controversial,
I'd appreciate some comments as to whether the issues from my previous
submission have bee
Several headers redundantly define the INLINE macro. Adding this
guard prevents the compiler from complaining about macro redefinition.
---
src/mesa/main/compiler.h | 42 ++
1 files changed, 22 insertions(+), 20 deletions(-)
diff --git a/src/mesa/main/co
This function is used by main.cpp to initialize a context to a default
configuration for use in compiling builtins. Moved the bulk of it to
glsl_parser_extras.cpp so that it can be re-used in testing.
---
src/glsl/glsl_parser_extras.cpp | 33 +
src/glsl/glsl_pars
This patch adds a new build artifact, glsl_test, which can be used for
testing optimization passes in isolation.
I'm hoping that we will be able to add other useful standalone tests
to this executable in the future. Accordingly, it is built in a
modular fashion: the main() function uses its first
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/17/2011 04:41 PM, Henri Verbeet wrote:
> This fixes a regression introduced by commit
> a26121f37530619610a78a5fbe5ef87e44047fda (fd.o bug #39219).
>
> Since the __glXInitialize() call should be unnecessary anyway, this is
> probably a nicer fix
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/18/2011 12:55 AM, Chad Versace wrote:
> Until now, the stencil buffer was allocated as a Y tiled buffer, because
> in several locations the PRM states that it is. However, it is actually
> W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part
On 07/18/2011 10:33 AM, Eric Anholt wrote:
This cuts out a large portion of the overhead of glClear() from
resetting the texenv state and recomputing the fixed function
programs. It also means less use of fixed function internally in our
GLES2 drivers, which is rather bogus.
Looks good. Only
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/18/2011 12:55 AM, Chad Versace wrote:
> Until now, the stencil buffer was allocated as a Y tiled buffer, because
> in several locations the PRM states that it is. However, it is actually
> W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part
On 15 July 2011 03:16, Kenneth Graunke wrote:
> When parsing S-Expressions, we need to store nul-terminated strings for
> Symbol nodes. Prior to this patch, we called ralloc_strndup each time
> we constructed a new s_symbol. It turns out that this is obscenely
> expensive.
>
> Instead, copy the
On 07/18/2011 12:37 PM, Paul Berry wrote:
Several headers redundantly define the INLINE macro. Adding this
guard prevents the compiler from complaining about macro redefinition.
Other Mesa headers? Or system headers? Or?
Reviewed-by: Brian Paul
---
src/mesa/main/compiler.h | 42
On Mon, 18 Jul 2011 08:09:17 -0600
Brian Paul wrote:
> On 07/15/2011 02:59 PM, Pekka Paalanen wrote:
> > On Fri, 15 Jul 2011 12:22:41 -0600
> > Brian Paul wrote:
> >
> >> On 07/15/2011 10:07 AM, Dave Airlie wrote:
> >>> On Fri, Jul 15, 2011 at 4:10 AM, Brian
> >>> Paul wrote:
>
>
> >
On 07/18/2011 12:47 PM, Brian Paul wrote:
On 07/18/2011 10:33 AM, Eric Anholt wrote:
This cuts out a large portion of the overhead of glClear() from
resetting the texenv state and recomputing the fixed function
programs. It also means less use of fixed function internally in our
GLES2 drivers, w
On 07/18/2011 01:00 PM, Pekka Paalanen wrote:
On Mon, 18 Jul 2011 08:09:17 -0600
Brian Paul wrote:
On 07/15/2011 02:59 PM, Pekka Paalanen wrote:
On Fri, 15 Jul 2011 12:22:41 -0600
Brian Paul wrote:
On 07/15/2011 10:07 AM, Dave Airlie wrote:
On Fri, Jul 15, 2011 at 4:10 AM, Brian
Paul
On 07/18/2011 11:45 AM, Ian Romanick wrote:
> On 07/18/2011 12:55 AM, Chad Versace wrote:
>> Until now, the stencil buffer was allocated as a Y tiled buffer, because
>> in several locations the PRM states that it is. However, it is actually
>> W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Par
2011/7/18 Mathias Fröhlich :
>
> Hi,
>
> On Monday, July 18, 2011 18:28:26 Marek Olšák wrote:
>> We can't do try-map + create + map + dereference internally in a
>> driver. Creating a new buffer and replacing a pointer to the old one
>> may lead to the following issue. If a buffer pointer is replac
On 07/18/2011 11:49 AM, Ian Romanick wrote:
> On 07/18/2011 12:55 AM, Chad Versace wrote:
>> Until now, the stencil buffer was allocated as a Y tiled buffer, because
>> in several locations the PRM states that it is. However, it is actually
>> W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Par
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/18/2011 12:28 PM, Chad Versace wrote:
> On 07/18/2011 11:45 AM, Ian Romanick wrote:
>> On 07/18/2011 12:55 AM, Chad Versace wrote:
>>> Until now, the stencil buffer was allocated as a Y tiled buffer, because
>>> in several locations the PRM state
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/18/2011 01:54 PM, Chad Versace wrote:
> On 07/18/2011 11:49 AM, Ian Romanick wrote:
>> On 07/18/2011 12:55 AM, Chad Versace wrote:
>>> Until now, the stencil buffer was allocated as a Y tiled buffer, because
>>> in several locations the PRM state
On 07/18/2011 08:57 AM, Eric Anholt wrote:
> On Mon, 18 Jul 2011 00:55:03 -0700, Chad Versace wrote:
>> diff --git a/src/mesa/drivers/dri/intel/intel_fbo.c
>> b/src/mesa/drivers/dri/intel/intel_fbo.c
>> index 1669af2..507cc33 100644
>> --- a/src/mesa/drivers/dri/intel/intel_fbo.c
>> +++ b/src/mes
As can be seen from the giant pile of tests recently sent to the
piglit mailing list for review, the
lower_variable_index_to_cond_assign pass had some issues. With this
patch series, *all* of the variable indexing tests posted to the
piglit list pass with (classic) swrast. 102 additional tests pa
From: Ian Romanick
There's no reason for it to be there, and another class that may not
have access to the visitor will need it soon.
---
src/glsl/lower_variable_index_to_cond_assign.cpp | 11 ++-
1 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/src/glsl/lower_variable_in
From: Ian Romanick
Other code will soon need to know if an array needs lowering based
exclusively on the storage mode.
---
src/glsl/lower_variable_index_to_cond_assign.cpp | 15 ++-
1 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/glsl/lower_variable_index_to_con
From: Ian Romanick
The previous implementation could easily get tricked if the LHS of an
assignment included a non-constant index that was "inside" another
dereference. For example:
mat4 m[2];
m[0][i] = vec4(0.0);
Due to the way it tracked whether the array was being assigned, it
would
From: Ian Romanick
If the non-constant index was in the LHS of an assignment, any
existing condititon on that assignment would be lost.
Fixes i965 piglit:
fs-temp-array-mat[234]-col-row-wr
fs-temp-array-mat[234]-index-col-row-wr
fs-temp-array-mat[234]-index-col-wr
fs-temp-array-
From: Ian Romanick
If the non-constant index was in the LHS of an assignment, any
existing condititon on that assignment would be lost.
---
src/glsl/lower_vec_index_to_cond_assign.cpp | 29 ++
1 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/src/glsl/lowe
From: Ian Romanick
This fixes many cases of accessing arrays of matrices using
non-constant indices at each level.
Fixes i965 piglit:
vs-temp-array-mat[234]-index-col-rd
vs-temp-array-mat[234]-index-col-row-rd
vs-temp-array-mat[234]-index-col-wr
vs-uniform-array-mat[234]-index-c
From: Ian Romanick
Fixes i965 piglit:
vs-temp-array-mat[234]-col-row-wr
vs-temp-array-mat[234]-index-col-row-wr
vs-temp-array-mat[234]-index-row-wr
vs-temp-mat[234]-col-row-wr
Fixes swrast piglit:
fs-temp-array-mat[234]-col-row-wr
fs-temp-array-mat[234]-index-col-row-wr
From: Ian Romanick
Itrating over all 3 possible source registers regardless of the opcode
was causing a variety of problems. The most noticable was an
occasional assertion failure:
brw_vs_emit.c:1381: get_src_reg: Assertion `c->regs[file][index].nr !=
0' failed.
This assertion seemed to be tri
On 07/18/2011 02:02 PM, Ian Romanick wrote:
> On 07/18/2011 01:54 PM, Chad Versace wrote:
>> On 07/18/2011 11:49 AM, Ian Romanick wrote:
>>> On 07/18/2011 12:55 AM, Chad Versace wrote:
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM st
On 18 July 2011 11:58, Brian Paul wrote:
> On 07/18/2011 12:37 PM, Paul Berry wrote:
>>
>> Several headers redundantly define the INLINE macro. Adding this
>> guard prevents the compiler from complaining about macro redefinition.
>
> Other Mesa headers? Or system headers? Or?
Other Mesa header
https://bugs.freedesktop.org/show_bug.cgi?id=39307
--- Comment #1 from Emeric Grange 2011-07-18 15:09:49
PDT ---
Indeed MPEG1 should be supported, but remains mostly untested.
That patch should do the trick, I'll try to get it commited later if I manage
to test it first.
http://cgit.freedesktop.
- Original Message -
> We can't do try-map + create + map + dereference internally in a
> driver. Creating a new buffer and replacing a pointer to the old one
> may lead to the following issue. If a buffer pointer is replaced, it
> doesn't necessarily update all the states the buffer is s
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 07/18/2011 02:24 PM, Chad Versace wrote:
> On 07/18/2011 02:02 PM, Ian Romanick wrote:
>> On 07/18/2011 01:54 PM, Chad Versace wrote:
>>> On 07/18/2011 11:49 AM, Ian Romanick wrote:
On 07/18/2011 12:55 AM, Chad Versace wrote:
> Until now, t
On Mon, Jul 18, 2011 at 6:21 PM, Jose Fonseca wrote:
>
>
> - Original Message -
>> We can't do try-map + create + map + dereference internally in a
>> driver. Creating a new buffer and replacing a pointer to the old one
>> may lead to the following issue. If a buffer pointer is replaced, i
On 07/18/2011 08:57 AM, Eric Anholt wrote:
> On Mon, 18 Jul 2011 00:55:03 -0700, Chad Versace wrote:
>> diff --git a/src/mesa/drivers/dri/intel/intel_fbo.c
>> b/src/mesa/drivers/dri/intel/intel_fbo.c
>> index 1669af2..507cc33 100644
>> --- a/src/mesa/drivers/dri/intel/intel_fbo.c
>> +++ b/src/mes
Patch 1 v2:
- Change buffer height from ALIGN(height / 2, 64) to
ALIGN((height + 1) / 2, 64).
Patch 2 v2:
- Change buffer height from ALIGN(height / 2, 64) to
ALIGN((height + 1) / 2, 64).
- Change return type of intel_offset_S8 changed to intptr_t.
- Improve performance of Y_
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM states that it is. However, it is actually
W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
4.5.2.1 W-Major Format:
W-Major Tile Format is used for separate stencil.
The GTT
Until now, the stencil buffer was allocated as a Y tiled buffer, because
in several locations the PRM states that it is. However, it is actually
W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
4.5.2.1 W-Major Format:
W-Major Tile Format is used for separate stencil.
The GTT
On 07/18/2011 05:08 PM, Chad Versace wrote:
> Until now, the stencil buffer was allocated as a Y tiled buffer, because
> in several locations the PRM states that it is. However, it is actually
> W tiled. From the PRM, 2011 Sandy Bridge, Volume 1, Part 2, Section
> 4.5.2.1 W-Major Format:
> W-Ma
On Tue, Jul 19, 2011 at 12:21 AM, Jose Fonseca wrote:
>
>
> - Original Message -
>> We can't do try-map + create + map + dereference internally in a
>> driver. Creating a new buffer and replacing a pointer to the old one
>> may lead to the following issue. If a buffer pointer is replaced,
- Original Message -
> On Tue, Jul 19, 2011 at 12:21 AM, Jose Fonseca
> wrote:
> >
> >
> > - Original Message -
> >> We can't do try-map + create + map + dereference internally in a
> >> driver. Creating a new buffer and replacing a pointer to the old
> >> one
> >> may lead to th
ES 2.0.25 page 127 says:
If the value of FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE is NONE, then
querying any other pname will generate INVALID_ENUM.
See also:
b9e9df78a03edb35472c2e231aef4747e09db792
NOTE: This is a candidate for the 7.10 and 7.11 branches.
---
src/mesa/main/fbobject.c | 22 +++
On 07/18/2011 11:37 AM, Paul Berry wrote:
> +++ b/src/glsl/glsl_parser_extras.cpp
> @@ -926,6 +926,39 @@ do_common_optimization(exec_list *ir, bool linked,
> unsigned max_unroll_iteration
> return progress;
> }
>
> +void initialize_context_to_defaults(struct gl_context *ctx, gl_api api)
> +
On 07/18/2011 11:37 AM, Paul Berry wrote:
> Several headers redundantly define the INLINE macro. Adding this
> guard prevents the compiler from complaining about macro redefinition.
> ---
> src/mesa/main/compiler.h | 42 ++
> 1 files changed, 22 insertion
Pushed, thanks.
Marek
On Fri, Jul 15, 2011 at 12:07 AM, Emil Velikov wrote:
> In a rare case of building gallium only, we need to
> check if the required packages are available
>
> libdrm_[intel|nouveau] - gallium[i915 i965|nouveau]
>
> v2: r300g and r600g do not need libdrm_radeon
>
> Signed-of
On Mon, Jul 18, 2011 at 8:32 PM, Marek Olšák wrote:
> On Tue, Jul 19, 2011 at 12:21 AM, Jose Fonseca wrote:
>>
>>
>> - Original Message -
>>> We can't do try-map + create + map + dereference internally in a
>>> driver. Creating a new buffer and replacing a pointer to the old one
>>> may l
On 07/18/2011 09:29 AM, Vadim Girlin wrote:
On Mon, 2011-07-18 at 08:34 -0600, Brian Paul wrote:
On Mon, Jul 18, 2011 at 8:11 AM, Vadim Girlin wrote:
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=39286
---
src/mesa/state_tracker/st_cb_texture.c |8
1 files changed, 8 inser
This looks good, I'll test it when I get a chance.
-Tom
On Mon, Jul 18, 2011 at 07:14:06AM +0200, Tobias Droste wrote:
> Signed-off-by: Tobias Droste
> ---
> .../drivers/dri/r300/compiler/radeon_optimize.c| 35 ++-
> 1 files changed, 18 insertions(+), 17 deletions(-)
>
>
On Mon, Jul 18, 2011 at 11:19:17AM -0700, Eric Anholt wrote:
> On Sun, 17 Jul 2011 16:25:42 -0700, Ben Widawsky wrote:
> > Upload the system routine as part of the invariant state if debugging.
> >
> > Remove SIP setting if not debugging to make it more friendly for others
> > that may be debuggi
https://bugs.freedesktop.org/show_bug.cgi?id=38716
--- Comment #12 from Jos van Wolput 2011-07-18
21:56:55 PDT ---
(In reply to comment #1)
> The command for linking wrongly put the local library search path after the
> system's. It should be fixed with 24137af. Please test.
I am using Debian
72 matches
Mail list logo
