https://bugs.freedesktop.org/show_bug.cgi?id=46082
Bug #: 46082
Summary: piglit ext_texture_integer-api-teximage regression
Classification: Unclassified
Product: Mesa
Version: git
Platform: x86-64 (AMD64)
OS/Version: Linux (All)
Both patches seem good to me.
Jose
- Original Message -
> st_equal_formats() is no longer used now.
> ---
> src/mesa/state_tracker/st_cb_texture.c |4 +++-
> 1 files changed, 3 insertions(+), 1 deletions(-)
>
> diff --git a/src/mesa/state_tracker/st_cb_texture.c
> b/src/mesa/state_t
This reduces recomputation of state based on non-clipping-related
transform changes, and is a step toward removing VUE map
recomputation.
---
src/mesa/drivers/dri/i965/brw_clip.c |6 +++---
src/mesa/drivers/dri/i965/brw_context.h|4 ++--
src/mesa/drivers/dri/i965/brw_gs.c
With this and the previous patch, 640x480 nexuiz is running 0.169118%
+/- 0.0863696% faster (n=121). On a VS state change microbenchmark,
performance is increased 8.28645% +/- 0.460478% (n=52).
---
To those who saw me throwing my hands in the air saying "why does this
patch not show up in the mi
On 02/13/2012 04:09 PM, Brian Paul wrote:
As suggested by José.
Reviewed-by: Ian Romanick
---
src/mesa/main/format_pack.c | 31 +++
src/mesa/main/format_unpack.c | 20 +++-
2 files changed, 30 insertions(+), 21 deletions(-)
diff --git a/
On 02/13/2012 04:57 PM, Eric Anholt wrote:
In the table of of push/pop attributes, this one doesn't fall under
the enable group.
For the series,
Reviewed-by: Ian Romanick
---
src/mesa/main/attrib.c |5 -
1 files changed, 0 insertions(+), 5 deletions(-)
diff --git a/src/mesa/main
On Tue, Feb 14, 2012 at 1:14 PM, Patrick Baggett
wrote:
> Ack, I realized I didn't CC that to the list, thanks for bringing that up.
Sorry, I CC'd to wrong mailing list.
> On Tue, Feb 14, 2012 at 2:45 PM, Anuj Phogat wrote:
>
>> On Mon, Feb 13, 2012 at 11:40 AM, Patrick Baggett <
>> baggett.pat
On Tue, Feb 14, 2012 at 2:53 AM, wrote:
> From: Zhigang Gong
>
> Added two color formats RGB565 and A8 to gbm layer.
> Added A8 to dri layer.
I was planning something like this:
http://people.freedesktop.org/~krh/gbm-get-format.patch
where we reuse the format codes introduced in drm. Also
Commit dc7f449d1ac53a66e6efb56ccf2a5953418a26ca introduced a new method
for avoiding MOVs: try to rewrite the destination of the instruction
that produced the RHS so it writes into the LHS.
Unfortunately, this is not safe for swizzled texturing operations, as
they return a set of four contiguous r
Certain instructions write more than one register. Texturing, for
example, returns 4 registers. (We set rlen to 4 even for TXS and float
shadow sampling.) Some math functions return 2. Most return 1.
The next commit introduces a use of this function.
Signed-off-by: Kenneth Graunke
---
src/m
On Mon, Feb 13, 2012 at 4:33 PM, Brian Paul wrote:
> On 02/13/2012 12:34 PM, Anuj Phogat wrote:
>
>> Reallocate/resize decompress FBO only if texture image width/height is
>> greater than existing decompress FBO width/height.
>>
>> Signed-off-by: Anuj Phogat
>> ---
>> src/mesa/drivers/common/met
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
The following changes since commit 65526d54aa2599b069bd443d3e6e9762e613042d:
docs: remove link to the GLSL compiler page (2012-02-09 18:08:02 -0700)
are available in the git repository at:
git://people.freedesktop.org/~chadversary/mesa.git 8.0-hi
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 02/14/2012 10:51 AM, Ian Romanick wrote:
> From: Ian Romanick
>
> If the generated shader for _TexEnvProgram is empty, force the use of
> the fixed-function code. Otherwise, go ahead and use the shader.
> This works around a mysterious issue on i
From: Ian Romanick
Before calling _tnl_run_pipeline, the i915 driver maps all textures
used in vertex shaders (via intel_tex_map_images). Later run_vp tries
to map the textures again (via _swrast_map_texture). However, this
just replaces the actual mappings with NULL pointers. When the
pointer
From: Ian Romanick
If the generated shader for _TexEnvProgram is empty, force the use of
the fixed-function code. Otherwise, go ahead and use the shader.
This works around a mysterious issue on i915 where fixed-function
software fallbacks are not working correctly.
This isn't really the fix we
Christian König wrote:
Benchmark wise sharpen seems cheap, but denoise at full strength is
expensive, and comparing full on before and after the patch on 1080
material shows that it's doing more work.
Correct, before anything under 0.4 would have given you a zero or one
pixel window and 1.0 wou
It was incomplete and didn't take byte swapping into account either.
---
src/mesa/state_tracker/st_format.c | 18 --
src/mesa/state_tracker/st_format.h |3 ---
2 files changed, 0 insertions(+), 21 deletions(-)
diff --git a/src/mesa/state_tracker/st_format.c
b/src/mesa/state
st_equal_formats() is no longer used now.
---
src/mesa/state_tracker/st_cb_texture.c |4 +++-
1 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/src/mesa/state_tracker/st_cb_texture.c
b/src/mesa/state_tracker/st_cb_texture.c
index d66791e..74ae2d1 100644
--- a/src/mesa/state_track
On 02/14/2012 08:19 AM, Jose Fonseca wrote:
Looks right.
We should probably bail out if the src_view is null too.
Yeah I can add that, but there's already several other places where we
don't check for null pointers in that function.
-Brian
___
mes
Looks right.
We should probably bail out if the src_view is null too.
Jose
- Original Message -
> Similar to the previous commit. Also fix incorrect setting of the
> sampler view's state after it's created. We need to specify the
> first/last_level fields in the template instead.
>
>
Similar to the previous commit. Also fix incorrect setting of the
sampler view's state after it's created. We need to specify the
first/last_level fields in the template instead.
NOTE: This is a candidate for the 8.0 branch.
---
src/mesa/state_tracker/st_cb_texture.c | 18 ++
On 02/14/2012 07:55 AM, Jose Fonseca wrote:
- Original Message -
Similar to the previous commit.
---
src/mesa/state_tracker/st_cb_texture.c |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/src/mesa/state_tracker/st_cb_texture.c
b/src/mesa/state_tracker/
Looks good.
Jose
- Original Message -
> Implement in terms of st_create_texture_sampler_view_format().
> ---
> src/mesa/state_tracker/st_texture.h | 21 +
> 1 files changed, 9 insertions(+), 12 deletions(-)
>
> diff --git a/src/mesa/state_tracker/st_texture.h
> b/s
Seems sensible thing to do. views are per context while texture objects are
shared, so stObj->sampler_view is definetely evil.
Jose
- Original Message -
> ---
> src/mesa/state_tracker/st_texture.h | 16
> 1 files changed, 0 insertions(+), 16 deletions(-)
>
> diff --g
- Original Message -
> Similar to the previous commit.
> ---
> src/mesa/state_tracker/st_cb_texture.c |7 +--
> 1 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/src/mesa/state_tracker/st_cb_texture.c
> b/src/mesa/state_tracker/st_cb_texture.c
> index c2011da..b0
Looks good to me.
Jose
- Original Message -
> Rather than the one in st_texture_object. This sampler view really
> has
> no connection to the one used for rendering.
> ---
> src/mesa/state_tracker/st_gen_mipmap.c |6 +-
> 1 files changed, 5 insertions(+), 1 deletions(-)
>
> dif
Looks a nice cleanup. Thanks. Just one suggestion inline.
- Original Message -
> As suggested by José.
> ---
> src/mesa/main/format_pack.c | 31 +++
> src/mesa/main/format_unpack.c | 20 +++-
> 2 files changed, 30 insertions(+), 21 deletio
On 02/13/2012 05:57 PM, Eric Anholt wrote:
Fixes (with the previous commit) piglit GL_ARB_multisample/pushpop.
---
src/mesa/main/attrib.c | 19 +--
1 files changed, 17 insertions(+), 2 deletions(-)
For both:
Reviewed-by: Brian Paul
r300g is able to sleep until a fence completes rather than busywait because
it creates a special buffer object and relocation that stays busy until the
CS containing the fence is finished.
Copy the idea into r600g, and use it to sleep if the user asked for an
infinite wait, falling back to busywai
On 13.02.2012 16:09, Maarten Lankhorst wrote:
Op 13-02-12 14:36, Christian König schreef:
@@ -242,8 +243,24 @@ VdpStatus vlVdpVideoMixerRender(VdpVideoMixer
mixer,
}
vl_compositor_clear_layers(&vmixer->compositor);
+
+ switch (current_picture_structure) {
+ case VDP_VIDEO_MIXER_P
On 13.02.2012 16:05, Maarten Lankhorst wrote:
Op 13-02-12 14:36, Christian König schreef:
Tracing function entry/exits is a bit pointless
when VDPAU_TRACE=1 does the same thing.
Signed-off-by: Christian König
---
src/gallium/state_trackers/vdpau/decode.c | 10 --
src/gallium/
On 13.02.2012 15:57, Maarten Lankhorst wrote:
Hey Christian,
Op 13-02-12 13:09, Christian König schreef:
No, it's not really needed, but I had it this way in the first place
and you need a lot of "if (something_is_not_null) cleanup(something)"
in the destructor functions, and with this impleme
On 13.02.2012 17:31, Andy Furniss wrote:
Christian König wrote:
A filter strength of zero or one doesn't make any
sense.
Low strengths work OK now, but this patch seems to have made denoise
stronger. Maybe I am looking at the wrong material, but more than 0.2
for sharpen or denoise looks too
https://bugs.freedesktop.org/show_bug.cgi?id=45292
Mathieu Zhang changed:
What|Removed |Added
CC||lingfe...@yahoo.com
--
Configure bugmai
From: Zhigang Gong
Added two color formats RGB565 and A8 to gbm layer.
Added A8 to dri layer.
Signed-off-by: Zhigang Gong
---
include/GL/internal/dri_interface.h |1 +
src/gbm/backends/dri/gbm_dri.c|6 ++
src/gbm/main/gbm.h|6 +-
sr
From: Zhigang Gong
To extent gbm to support more formats and usages, it's better
to do the format and usage matching in a seprated function.
Also fixed a memory leak at gbm_dri_bo_create, if failed to
create a DRI image, before return NULL, we need to free
bo.
Signed-off-by: Zhigang Gong
---
s
36 matches
Mail list logo
