https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #13 from Michel Dänzer ---
Does it still happen with the Radeon card with a 3.19-rc kernel?
--
You are receiving this mail because:
You are the assignee for the bug.
___
mesa-dev mailing l
Rendering with a GS and then using transform feedback with a program that does
not have a GS can crash in gen6. The reason for this is that
brw_begin_transform_feedback checks brw->geometry_program to decide if there
is a GS program, but this is not correct: brw->geometry_program is updated when
is
This is still awaiting for comments.
I'd rather hear what are the desirable modifications than try guessing.
On ma, 2014-11-10 at 15:18 +0200, Joonas Lahtinen wrote:
> Hi,
>
> On pe, 2014-11-07 at 17:40 -0800, Eric Anholt wrote:
> > Ian Romanick writes:
> >
> > > On 11/06/2014 06:16 PM, Michel
From: Marek Olšák
set_shader_resources is unused.
set_shader_buffers should support shader atomic counter buffers and shader
storage buffers from OpenGL.
The plan is to use slots 0..15 for atomic counters and slots 16..31
for storage buffers. Atomic counters are planned to be supported first.
How about the attached patch?
Marek
On Wed, Jan 7, 2015 at 1:23 AM, Tom Stellard wrote:
> On Wed, Jan 07, 2015 at 01:13:37AM +0100, Marek Olšák wrote:
>> Neither. It's because we use DX10_CLAMP, which converts NaNs to 0.
>>
>
> Ok, could we add a dx10_clamp bit to si_shader and make this attribu
On 07/01/2015 10:24, Joonas Lahtinenwrote :
This is still awaiting for comments.
I'd rather hear what are the desirable modifications than try guessing.
Well, ideally you would implement DRI3/Present support instead of
complementing DRI2 support with Present.
Why improve the old solution, i
https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #14 from Gustaw Smolarczyk ---
If you are using 3.18 kernel, you could also try the previous one (3.17.x). I
have a similar problem on radeon (though it's TAHITI, so radeonsi) and found
that it is a kernel regression.
In my case, I c
On Monday, December 15, 2014 12:30:27 PM Samuel Iglesias Gonsálvez wrote:
> On Thursday, December 11, 2014 11:55:10 AM Jason Ekstrand wrote:
> > On Tue, Dec 9, 2014 at 4:06 AM, Iago Toral Quiroga
> > wrote:
> >
> > [snip]
> >
> > > new file mode 100644
> > > index 000..5f6809e
> > > --- /dev
On Wed, Jan 7, 2015 at 5:56 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> set_shader_resources is unused.
>
> set_shader_buffers should support shader atomic counter buffers and shader
> storage buffers from OpenGL.
>
> The plan is to use slots 0..15 for atomic counters and slots 16..31
> for st
Hi,
Sounds great but, do you think a separate buffer pipe is required for this?
Changing Constant buffer to a generic buffer (with alu+load+store) can help.
What about for R600? Do we have to add
r600_init_atom(rctx, &rctx->shaderbuf_state[PIPE_SHADER_VERTEX].atom, id++,
r600_emit_vs_shader_buffe
On Wed, Jan 7, 2015 at 6:58 AM, Marek Olšák wrote:
> How about the attached patch?
Reviewed-by: Alex Deucher
>
> Marek
>
> On Wed, Jan 7, 2015 at 1:23 AM, Tom Stellard wrote:
>> On Wed, Jan 07, 2015 at 01:13:37AM +0100, Marek Olšák wrote:
>>> Neither. It's because we use DX10_CLAMP, which conv
On Wed, Jan 7, 2015 at 2:41 PM, Ilia Mirkin wrote:
> On Wed, Jan 7, 2015 at 5:56 AM, Marek Olšák wrote:
>> From: Marek Olšák
>>
>> set_shader_resources is unused.
>>
>> set_shader_buffers should support shader atomic counter buffers and shader
>> storage buffers from OpenGL.
>>
>> The plan is to
On Wed, Jan 7, 2015 at 4:56 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> set_shader_resources is unused.
>
> set_shader_buffers should support shader atomic counter buffers and shader
> storage buffers from OpenGL.
>
> The plan is to use slots 0..15 for atomic counters and slots 16..31
> for s
On Jan 7, 2015 4:45 AM, "Samuel Iglesias Gonsálvez"
wrote:
>
> On Monday, December 15, 2014 12:30:27 PM Samuel Iglesias Gonsálvez wrote:
> > On Thursday, December 11, 2014 11:55:10 AM Jason Ekstrand wrote:
> > > On Tue, Dec 9, 2014 at 4:06 AM, Iago Toral Quiroga
> > > wrote:
> > >
> > > [snip]
>
On Wed, Jan 7, 2015 at 3:44 PM, Aditya Avinash wrote:
>
>
> On Wed, Jan 7, 2015 at 4:56 AM, Marek Olšák wrote:
>>
>> From: Marek Olšák
>>
>> set_shader_resources is unused.
>>
>> set_shader_buffers should support shader atomic counter buffers and shader
>> storage buffers from OpenGL.
>>
>> The
Oh. So, we get better performance if we use atomic counters as buffers
rather than textures (images) [manipulating views are expensive].
Am I right?
On Wed, Jan 7, 2015 at 8:52 AM, Marek Olšák wrote:
> On Wed, Jan 7, 2015 at 3:44 PM, Aditya Avinash
> wrote:
> >
> >
> > On Wed, Jan 7, 2015 at 4
This is still awaiting for comments.
Sam
On Thursday, December 11, 2014 11:34:12 PM Eduardo Lima Mitev wrote:
> From: Samuel Iglesias Gonsalvez
>
> When a vec has more elements than row components in a matrix, the
> code could end up failing an assert inside assign_to_matrix_column().
>
> This
On Wed, Jan 7, 2015 at 9:35 AM, Marek Olšák wrote:
> On Wed, Jan 7, 2015 at 2:41 PM, Ilia Mirkin wrote:
>> On Wed, Jan 7, 2015 at 5:56 AM, Marek Olšák wrote:
>>> From: Marek Olšák
>>>
>>> set_shader_resources is unused.
>>>
>>> set_shader_buffers should support shader atomic counter buffers and
On Tue, Jan 6, 2015 at 6:36 PM, Jason Ekstrand wrote:
>
>
> On Mon, Jan 5, 2015 at 10:45 PM, Connor Abbott wrote:
>>
>> I created nir_tex_src_sampler_index for exactly this purpose, which
>> fits in with the "stick all the sources in an array so we can easily
>> iterate over them" philosophy. If
On Wed, Jan 07, 2015 at 12:58:12PM +0100, Marek Olšák wrote:
> How about the attached patch?
>
Looks good, thanks.
Reviewed-by: Tom Stellard
> Marek
>
> On Wed, Jan 7, 2015 at 1:23 AM, Tom Stellard wrote:
> > On Wed, Jan 07, 2015 at 01:13:37AM +0100, Marek Olšák wrote:
> >> Neither. It's bec
On Wed, Jan 7, 2015 at 2:42 PM, Aditya Avinash wrote:
> Hi,
> Sounds great but, do you think a separate buffer pipe is required for this?
> Changing Constant buffer to a generic buffer (with alu+load+store) can help.
No, constant buffers should remain unchanged.
>
> What about for R600? Do we ha
On Wed, Jan 7, 2015 at 11:29 AM, Marek Olšák wrote:
> On Wed, Jan 7, 2015 at 2:42 PM, Aditya Avinash
> wrote:
>> Hi,
>> Sounds great but, do you think a separate buffer pipe is required for this?
>> Changing Constant buffer to a generic buffer (with alu+load+store) can help.
>
> No, constant buf
This->surfaces contains the surfaces associated to the levels
and faces. This->surfaces[6*Level] is what we want here,
since it gives us a face descriptor for the level 'Level'.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Signed-off-by: Xavier Bouchoux
Cc: "10.4"
---
src/gallium/s
Use same similar settings as u_sampler_view_default_template
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/basetexture9.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/gallium/state_trackers/nine/basetexture
The implicit swapchains are destroyed when the device instance is
destroyed. However for non-implicit swapchains, it is not the case,
and the application can have kept an reference on the swapchain
buffers to reuse them.
Fixes problems with battle.net launcher.
Cc: "10.4"
Tested-by: Nick Sarnie
From: David Heidelberg
Signed-off-by: David Heidelberg
Reviewed-by: Axel Davy
---
src/gallium/state_trackers/nine/query9.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/query9.c
b/src/gallium/state_trackers/nine/query9.c
index 6df4ead..466b4ba 100644
--- a
From: xavier
It's done by testing the existence of the point sprite output register *after*
parsing the vertex shader.
Reviewed-by: David Heidelberg
Reviewed-by: Axel Davy
Signed-off-by: Xavier Bouchoux
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 6 +++---
1 file change
From: Stanislaw Halik
Buffers in the MANAGED pool are supposed to have the content in a ram buffer,
a copy in VRAM if there is enough memory (driver manages memory and decide when
to delete the buffer in VRAM).
This is not implemented properly in nine, and a VRAM copy is going to be created
when
Convert them to shader booleans at earlier stage
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/device9.c| 35 +---
src/gallium/state_trackers/nine/device9.h| 6 ++---
src/gallium/state_trackers/nine/nine_state.c | 13 +++
3 fil
The cap means D3DFVF_XYZRHW vertices will see clipping.
This is not the case when
PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION is supported, since
it'll disable clipping.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/adapter9.c | 4 +++-
1 file changed, 3 insertions(+), 1 d
From: xavier
Reviewed-by: David Heidelberg
Reviewed-by: Axel Davy
Signed-off-by: Xavier Bouchoux
Cc: "10.4"
---
include/D3D9/d3d9types.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/include/D3D9/d3d9types.h b/include/D3D9/d3d9types.h
index 0a8f9e5..e53e389 100644
--- a/in
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/cubetexture9.c | 8
src/gallium/state_trackers/nine/texture9.c | 9 -
src/gallium/state_trackers/nine/volumetexture9.c | 10 +-
3 files changed, 25 insertions(+), 2 delet
Most of the patches of this serie are fixes,
that's why most of them are CC 10.4
The patches can be retrived here:
https://github.com/iXit/Mesa-3D/tree/submit_mesa
Some of them may not appear directly as fixes, so little
explanation is required:
"st/nine: Add ATI1 and ATI2 support"
These two co
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
include/D3D9/d3d9types.h| 3 +++
src/gallium/state_trackers/nine/nine_pipe.h | 3 +++
2 files changed, 6 insertions(+)
diff --git a/include/D3D9/d3d9types.h b/include/D3D9/d3d9types.h
index e53e389..456a
From: xavier
Reviewed-by: David Heidelberg
Reviewed-by: Axel Davy
Signed-off-by: Xavier Bouchoux
Cc: "10.4"
---
include/D3D9/d3d9.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/include/D3D9/d3d9.h b/include/D3D9/d3d9.h
index f872be7..e8b5214 100644
--- a/include/D3D9/d3d9
The clip state was reset everytime, incurring an overhead.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/nine_state.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_state.c
b/src/gallium/s
While previous code was having the correct behaviour in general,
this new code is more readable (without checking all gallium formats
manually) and has a more defined behaviour for depth stencil resources.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_
Previous implementation was fine,
just instead of having increasing counter,
have a decreasing counter.
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/nine_shader.c | 41 +++
1 file changed, 23 insertions(+), 18 deletions(-)
diff --git a/src/gallium/state_t
Not filling them correctly results in bad padding and later crash.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 46 +--
1 file changed, 23 insertions(+), 23 deletions(-)
diff --git a/src/gallium/
Relative addressing needs the constant buffer to get all
the correct constants, even those defined by the shader.
The code to copy the shader constants to the constant buffer
was enabled only for debug build. Enable it always.
Cc: "10.4"
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
-
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 12 +++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index c484e7c..02fb69e 100644
---
According to msdn, we must act as if user didn't ask srgb if we don't
support it.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/basetexture9.c | 11 ++-
src/gallium/state_trackers/nine/surface9.c | 10 +-
2 files chan
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_ff.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_ff.c
b/src/gallium/state_trackers/nine/nine_ff.c
index a6bd360..d2b30f8 100644
--- a/src/gallium/stat
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/device9.c | 10 ++
src/gallium/state_trackers/nine/device9.h | 2 ++
src/gallium/state_trackers/nine/stateblock9.c | 4 ++--
3 files changed, 10 insertions(+), 6 deletions(-)
diff --git a/src/gallium/state
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 27 ++-
1 file changed, 26 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 440f6f7..ac86
The shader code is already behaving as if they are floats when the the card
doesn't support integers
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/device9.c | 65 ---
1 file changed, 52 insertions(+), 13 deletions(-)
diff --git a/src/gall
POW doesn't match directly TGSI, since we should
take the absolute value of src0.
Fixes black textures in some games
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 13 -
1 file changed, 12 insertions(+), 1 del
The fix is that this line:
"src[s] = tx->regs.vT[s];" is wrong if s doesn't start from 0.
Instead access tx->regs.vT directly when needed.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 53 ++-
1 file changed, 36 insertions(+), 17
This is wine (and windows) behaviour.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 3fefce4..f
Nine code was clamping color outputs for vs < 3,
but msdn docs says it is done in the ps.
Wine seems to clamp them at the vs level.
It makes more sense to clamp at vs level for performance,
but according to doc, ps 2.x shouldn't see clamping.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Dav
Signed-off-by: Axel Davy
Signed-off-by: Tiziano Bacocco
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 16 +++-
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader
ps3 allows definitions of the inputs like:
DCL_TEXCOORD0 v0.xy;
DCL_NORMAL2 v0.z;
DCL_NORMAL3 v0.w;
Nine wouldn't have handled this situation properly.
Apparently very few applications use this feature.
Still remain an issue with this new implementation:
It is allowed to do indirect addressing o
Let's say we have c1 and c2 declared in the shader and c0 given by the app
Then here we would have read c0, c1 and c2 given by the app, instead
of the correct c0, c1, c2.
This correction fixes several issues in some games.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nin
relative adressing for constants is possible only for vs float
constants.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 14 --
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 23 ++-
1 file changed, 22 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_s
Since constant indirect adressing is not allowed for ps,
we can remove our code to handle that.
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_state.c | 43 +++---
src/gallium/state_trackers/nine/pixelshader9.c | 10 +++---
src/gallium/state_t
We should use the absolute value of the input as input to ureg_RSQ.
Moreover, an input of 0.0 should return FLT_MAX.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 13 -
1 file changed, 12 insertions(+), 1 del
Previous code semantic was:
. if ff ps will not run a ff stage, then do not output texture coords for this
stage
for vs
. if XYZRHW is used (position_t), use only the mode where input coordinates are
copied
to the outputs.
Problem is when apps don't give texture inputs. When apps precise PASSTH
Cc: "10.4"
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/nine_shader.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 0b378d5..6
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 20 +++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shad
Previous implementation didn't work well with nested loops.
Instead of using several address registers, put a0 and aL
into normal registers, and copy them to one address register when
we need to use them.
Wine tests loop_index_test() and nested_loop_test() now pass correctly.
Fixes r600g crash w
This patch raises nine requirements and disables nine for old
hw that don't match them.
It would be possible to make a lot of things work with these hw,
though not everything, but it needs special care for them in the
code, and since they are very old, it's better to drop explicitly
support for th
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 39 ++-
1 file changed, 38 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 20a8e8a..3b29
We should take the absolute value of the input.
Also return -FLT_MAX instead of -Inf for an input of 0.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 15 +--
1 file changed, 13 insertions(+), 2 deletions(-)
When the shader does indirect addressing on the constants,
we allocate a temporary constant buffer to which we copy
the constants from the app given user constants and
the constants filled in the shader.
This patch makes this buffer be allocated once.
Signed-off-by: Axel Davy
Signed-off-by: Tizi
texcoord for ps < 1_4 should clamp between 0 and 1 the values.
texcrd (texcoord ps 1_4) does not clamp and can be used with
two modifiers _dw and _dz that means the channels are divided
by w or z.
Implement those in shared code, since the same modifiers can be used
for texld ps 1_4.
Signed-off-by
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/nine_state.c | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_state.c
b/src/gallium/state_trackers/nine/nine_state.c
index 09b4401..99173fa 100644
--- a/src/gallium/st
Nothing seems to indicates the negation modifier would be stored in the
instruction flags instead of the source modifier. tx_src_param has
already handled it if it is in the source modifier.
In addition,
when the card supports native integers, the boolean
are stored in 32 bits int and are equal to
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
---
src/gallium/state_trackers/nine/nine_shader.c | 12
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 8bcf67b..
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 39 ---
1 file changed, 36 insertions(+), 3 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 794a1db..9d4
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 23 +--
1 file changed, 1 insertion(+), 22 deletions(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index fcc1c68..8b96673
According to docs and Wine, these two vs outputs have
to be saturated.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/gallium/state_trackers/nine/ni
Adds ATI1 and ATI2 support to nine.
They map to PIPE_FORMAT_RGTC1_UNORM and PIPE_FORMAT_RGTC2_UNORM,
but need special handling.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Signed-off-by: Xavier Bouchoux
Cc: "10.4"
---
src/gallium/state_trackers/nine/adapter9.c | 3 +++
src
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_shader.c | 20 +++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/nine/nine_shader.c
b/src/gallium/state_trackers/nine/nine_shader.c
index 02fb69e..794a1db 100
When the input's xyz are 0.0, the output
should be 0.0. This is due to the fact that
Inf * 0 = 0 for dx9. To handle this case,
cap the result of RSQ to FLT_MAX. We have
FLT_MAX * 0 = 0.
Reviewed-by: David Heidelberg
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/nine_
Signed-off-by: Axel Davy
Cc: "10.4"
---
src/gallium/state_trackers/nine/device9.c| 6 +++---
src/gallium/state_trackers/nine/nine_state.c | 7 ---
src/gallium/state_trackers/nine/nine_state.h | 2 +-
3 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/src/gallium/state_tracker
Yes. I expect lower overhead on SI. This interface should have no
impact on r600g.
Marek
On Wed, Jan 7, 2015 at 3:59 PM, Aditya Avinash wrote:
> Oh. So, we get better performance if we use atomic counters as buffers
> rather than textures (images) [manipulating views are expensive].
>
> Am I rig
Thank you!
On Wednesday, January 7, 2015, Marek Olšák wrote:
> On Wed, Jan 7, 2015 at 2:42 PM, Aditya Avinash > wrote:
> > Hi,
> > Sounds great but, do you think a separate buffer pipe is required for
> this?
> > Changing Constant buffer to a generic buffer (with alu+load+store) can
> help.
>
>
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> From: xavier
Would be nice to fix up the from name... you can do this with 'git
commit --amend --author asdf' or by reimporting the patch.
>
> Reviewed-by: David Heidelberg
> Reviewed-by: Axel Davy
> Signed-off-by: Xavier Bouchoux
>
> Cc: "
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> The cap means D3DFVF_XYZRHW vertices will see clipping.
> This is not the case when
> PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION is supported, since
> it'll disable clipping.
>
> Signed-off-by: Axel Davy
>
> Cc: "10.4"
> ---
> src/gallium/state_tr
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> Reviewed-by: David Heidelberg
> Signed-off-by: Axel Davy
> ---
> src/gallium/state_trackers/nine/cubetexture9.c | 8
> src/gallium/state_trackers/nine/texture9.c | 9 -
> src/gallium/state_trackers/nine/volumetexture
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> While previous code was having the correct behaviour in general,
> this new code is more readable (without checking all gallium formats
> manually) and has a more defined behaviour for depth stencil resources.
>
> Reviewed-by: David Heidelberg
>
On Wed, 7 Jan 2015, Ilia Mirkin wrote:
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
The cap means D3DFVF_XYZRHW vertices will see clipping.
This is not the case when
PIPE_CAP_TGSI_VS_WINDOW_SPACE_POSITION is supported, since
it'll disable clipping.
Signed-off-by: Axel Davy
Cc: "10.4"
On Wed, Jan 7, 2015 at 4:39 PM, Ilia Mirkin wrote:
> On Wed, Jan 7, 2015 at 9:35 AM, Marek Olšák wrote:
>> On Wed, Jan 7, 2015 at 2:41 PM, Ilia Mirkin wrote:
>>> On Wed, Jan 7, 2015 at 5:56 AM, Marek Olšák wrote:
From: Marek Olšák
set_shader_resources is unused.
set_sh
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> Convert them to shader booleans at earlier stage
Why? What's wrong with the conversion as it is now?
>
> Signed-off-by: Axel Davy
>
> Cc: "10.4"
> ---
> src/gallium/state_trackers/nine/device9.c| 35
> +---
> src
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> Signed-off-by: Axel Davy
> Cc: "10.4"
> ---
> src/gallium/state_trackers/nine/nine_shader.c | 23 +--
> 1 file changed, 1 insertion(+), 22 deletions(-)
>
> diff --git a/src/gallium/state_trackers/nine/nine_shader.c
> b/src
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> Adds ATI1 and ATI2 support to nine.
>
> They map to PIPE_FORMAT_RGTC1_UNORM and PIPE_FORMAT_RGTC2_UNORM,
> but need special handling.
>
> Reviewed-by: David Heidelberg
> Signed-off-by: Axel Davy
> Signed-off-by: Xavier Bouchoux
>
> Cc: "10.4"
Hmm I'm not quite sure what to think of it. Apparently
set_shader_resources was a closer match to what d3d does (seems to use
UAVs for everything as far as I can tell, plus they are global and not
per stage - I guess another reason why they were set together with RTs).
I guess set_shader_buffers wo
On Wed, Jan 7, 2015 at 7:55 AM, Connor Abbott wrote:
> On Tue, Jan 6, 2015 at 6:36 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Mon, Jan 5, 2015 at 10:45 PM, Connor Abbott
> wrote:
> >>
> >> I created nir_tex_src_sampler_index for exactly this purpose, which
> >> fits in with the "stick all the s
On Wed, Jan 7, 2015 at 9:58 AM, Jason Ekstrand wrote:
>
>
> On Wed, Jan 7, 2015 at 7:55 AM, Connor Abbott wrote:
>
>> On Tue, Jan 6, 2015 at 6:36 PM, Jason Ekstrand
>> wrote:
>> >
>> >
>> > On Mon, Jan 5, 2015 at 10:45 PM, Connor Abbott
>> wrote:
>> >>
>> >> I created nir_tex_src_sampler_index
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> We should use the absolute value of the input as input to ureg_RSQ.
>
> Moreover, an input of 0.0 should return FLT_MAX.
>
> Reviewed-by: David Heidelberg
> Signed-off-by: Axel Davy
>
> Cc: "10.4"
> ---
> src/gallium/state_trackers/nine/nine_
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> When the input's xyz are 0.0, the output
> should be 0.0. This is due to the fact that
> Inf * 0 = 0 for dx9. To handle this case,
> cap the result of RSQ to FLT_MAX. We have
> FLT_MAX * 0 = 0.
>
> Reviewed-by: David Heidelberg
> Signed-off-by:
On Wed, Jan 7, 2015 at 1:11 PM, Ilia Mirkin wrote:
> On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
>> When the input's xyz are 0.0, the output
>> should be 0.0. This is due to the fact that
>> Inf * 0 = 0 for dx9. To handle this case,
>> cap the result of RSQ to FLT_MAX. We have
>> FLT_MAX *
If we capture transform feedback from n stream in (n-1) buffers we face a
NULL buffer, use the buffer (n-1) to capture the output of stream n.
This fixes one piglit test with nvc0:
arb_gpu_shader5-xfb-streams-without-invocations
Signed-off-by: Tobias Klausmann
---
src/mesa/state_tracker/st_c
On Wed, Jan 07 2015, Jose Fonseca wrote:
> I lost bit track of email over the Christmas period. Just noticed I had
> flagged this one for replay. Sorry.
No worries. Thanks for following up now. :-)
> Do you still need me to test anything on Windows? If so are the patches
> in some pull-able gi
We need to dispose the cached LLVMTargetMachine before calling
r600_destory_common_screen(), because this function frees the si_screen
object which invalidates LLVMTargetMachine pointer.
https://bugs.freedesktop.org/show_bug.cgi?id=88170
---
src/gallium/drivers/radeonsi/si_pipe.c | 6 --
1 fi
https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #15 from almos ---
I tried l4d2 again with mesa 10.5-dev (git-1829f9c), and still nothing. Kernel
is the same as before (3.17.7). Do I need to underclock my CPU to see the lag
spikes?
--
You are receiving this mail because:
You are
https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #16 from Stéphane Travostino ---
Is it possible there's a weird interaction with PRIME? @almos, does your system
have a muxless setup?
My CPU isn't underclocked nor undervolted, and using the performance governor
doesn't help in any
On Mon, Jan 05, 2015 at 12:18:40AM +0100, Marek Olšák wrote:
> From: Marek Olšák
>
> - the relocs array is unused, remove it
> - ndw is at most 115 (init), set 140 as the maximum
> - compute needs 4 buffers per state, graphics only needs 1; set 4 as the
> maximum
The number of buffers per state
On Wed, Jan 7, 2015 at 11:36 AM, Axel Davy wrote:
> Previous implementation was fine,
> just instead of having increasing counter,
> have a decreasing counter.
>
> Signed-off-by: Axel Davy
> ---
> src/gallium/state_trackers/nine/nine_shader.c | 41
> +++
> 1 file changed
1 - 100 of 193 matches
Mail list logo
