Re: [Mesa-dev] glxgears is faster but 3D render is so slow

2013-03-21 Thread jupiter
Hi Brian, On 3/21/13, Brian Paul wrote: > On 03/20/2013 04:07 AM, jupiter wrote: >> Hi Brian, >> >> On 3/19/13, Brian Paul wrote: It is fair to say, if running llvm driver in my local machine (a 32-bit CentOS 6.2 without VNC connection), it was indeed faster than the xlib driver.

[Mesa-dev] [PATCH 2/5] radeonsi: mark all loads as constant

2013-03-21 Thread Christian König
From: Christian König Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c | 32 ++-- 1 file changed, 25 insertions(+), 7 deletions(-) diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_shader.

[Mesa-dev] [PATCH 1/5] radeonsi: remove wqm intrinsic

2013-03-21 Thread Christian König
From: Christian König Now the backend handles that itself. Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c |9 - 1 file changed, 9 deletions(-) diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_sh

[Mesa-dev] [PATCH 3/5] radeonsi: mark most intrinsics as readnone/nounwind

2013-03-21 Thread Christian König
From: Christian König Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c | 18 ++ 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_shader.c inde

[Mesa-dev] [PATCH 4/5] radeonsi: add preloading of all constants

2013-03-21 Thread Christian König
From: Christian König Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c | 67 ++-- 1 file changed, 51 insertions(+), 16 deletions(-) diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_shader

[Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Christian König
From: Christian König Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c | 57 +++- 1 file changed, 45 insertions(+), 12 deletions(-) diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_shader

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Michel Dänzer
When I tried this earlier, something broke. Unfortunately, I can't seem to remember or dig up if it was a piglit test or e.g. one of the sampler demos in mesa/demos/src/glsl/. Did you test the latter with this change? Similar concern for patch 4, the rest of the series looks good to me. -- Ear

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Alex Deucher
On Thu, Mar 21, 2013 at 7:38 AM, Christian König wrote: > From: Christian König > > Signed-off-by: Christian König > --- > src/gallium/drivers/radeonsi/radeonsi_shader.c | 57 > +++- > 1 file changed, 45 insertions(+), 12 deletions(-) > > diff --git a/src/gallium/drivers/

Re: [Mesa-dev] glxgears is faster but 3D render is so slow

2013-03-21 Thread Brian Paul
On 03/21/2013 03:51 AM, jupiter wrote: Hi Brian, On 3/21/13, Brian Paul wrote: On 03/20/2013 04:07 AM, jupiter wrote: Hi Brian, On 3/19/13, Brian Paul wrote: It is fair to say, if running llvm driver in my local machine (a 32-bit CentOS 6.2 without VNC connection), it was indeed faster th

Re: [Mesa-dev] [PATCH 2/4] i965: Create a pointer in brw_context to the geometry output VUE map.

2013-03-21 Thread Eric Anholt
Paul Berry writes: > Currently, the GPU pipeline has one active VUE map in effect at any > given time--the one representing the layout of vertex data coming from > the vertex shader. However, when geometry shaders are added, they > will have their own independent VUE map. Later pipeline stages

Re: [Mesa-dev] [PATCH 3/4] i965: Use brw.vue_map_geom_out instead of VS output VUE map where appropriate.

2013-03-21 Thread Eric Anholt
Paul Berry writes: > This patch modifies post-GS pipeline stages (transform feedback, clip, > sf, fs) to refer to the VUE map through brw->vue_map_geom_out rather > than brw->vs.prog_data->vue_map. This ensures that when geometry > shader support is added, these pipeline stages will consult the

Re: [Mesa-dev] [PATCH 4/4] i965/fs: Rename vp_outputs_written to input_slots_valid.

2013-03-21 Thread Eric Anholt
Paul Berry writes: > With the introduction of geometry shaders, fragment inputs will no > longer come exclusively from the vertex shader; sometimes they come > from the geometry shader. So the name "vp_outputs_written" will > become a misnomer. This patch renames vp_outputs_written to > input_s

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Christian König
Am 21.03.2013 15:10, schrieb Alex Deucher: On Thu, Mar 21, 2013 at 7:38 AM, Christian König wrote: From: Christian König Signed-off-by: Christian König --- src/gallium/drivers/radeonsi/radeonsi_shader.c | 57 +++- 1 file changed, 45 insertions(+), 12 deletions(-) di

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Christian König
Am 21.03.2013 15:06, schrieb Michel Dänzer: When I tried this earlier, something broke. Unfortunately, I can't seem to remember or dig up if it was a piglit test or e.g. one of the sampler demos in mesa/demos/src/glsl/. Did you test the latter with this change? Similar concern for patch 4, the r

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Michel Dänzer
On Don, 2013-03-21 at 17:12 +0100, Christian König wrote: > Am 21.03.2013 15:06, schrieb Michel Dänzer: > > When I tried this earlier, something broke. Unfortunately, I can't seem > > to remember or dig up if it was a piglit test or e.g. one of the sampler > > demos in mesa/demos/src/glsl/. Did yo

Re: [Mesa-dev] [PATCH 5/5] radeonsi: add preloading for all samplers

2013-03-21 Thread Alex Deucher
On Thu, Mar 21, 2013 at 12:09 PM, Christian König wrote: > Am 21.03.2013 15:10, schrieb Alex Deucher: > >> On Thu, Mar 21, 2013 at 7:38 AM, Christian König >> wrote: >>> >>> From: Christian König >>> >>> Signed-off-by: Christian König >>> --- >>> src/gallium/drivers/radeonsi/radeonsi_shader.c

[Mesa-dev] [PATCH] R600: Use legacy (0 * anything = 0) MUL instructions for pow intrinsics

2013-03-21 Thread Michel Dänzer
From: Michel Dänzer Fixes wrong lighting in some corner cases with r600g and radeonsi, e.g. manifested by failure of two piglit/glean tests and intermittent black patches in many apps. Tested on SI and RS880. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=62012 [radeonsi] Bugzilla: http

Re: [Mesa-dev] [PATCH] R600: Use legacy (0 * anything = 0) MUL instructions for pow intrinsics

2013-03-21 Thread Christian König
Am 21.03.2013 17:32, schrieb Michel Dänzer: From: Michel Dänzer Fixes wrong lighting in some corner cases with r600g and radeonsi, e.g. manifested by failure of two piglit/glean tests and intermittent black patches in many apps. Tested on SI and RS880. Bugzilla: https://bugs.freedesktop.org/s

[Mesa-dev] [PATCH] r600g: Honour legacy debugging environment variables

2013-03-21 Thread Michel Dänzer
From: Michel Dänzer This helps minimize confusion / effort when moving between branches or helping others. Signed-off-by: Michel Dänzer --- src/gallium/drivers/r600/r600_pipe.c | 10 ++ 1 file changed, 10 insertions(+) diff --git a/src/gallium/drivers/r600/r600_pipe.c b/src/gallium

Re: [Mesa-dev] [PATCH] r600g: Honour legacy debugging environment variables

2013-03-21 Thread Alex Deucher
On Thu, Mar 21, 2013 at 12:59 PM, Michel Dänzer wrote: > From: Michel Dänzer > > This helps minimize confusion / effort when moving between branches or > helping others. Reviewed-by: Alex Deucher > > Signed-off-by: Michel Dänzer > --- > src/gallium/drivers/r600/r600_pipe.c | 10 ++

[Mesa-dev] [PATCH] mesa: Delete VERT_ATTRIB_GENERIC_NV and VERT_BIT_GENERIC_NV macros.

2013-03-21 Thread Kenneth Graunke
These haven't been used since we deleted NV_vertex_program support. Signed-off-by: Kenneth Graunke --- src/mesa/main/mtypes.h | 10 -- 1 file changed, 10 deletions(-) diff --git a/src/mesa/main/mtypes.h b/src/mesa/main/mtypes.h index 7900897..1a35e63 100644 --- a/src/mesa/main/mtypes.h

[Mesa-dev] [PATCH] mesa: Rename gl_context::swtnl_im to vbo_context.

2013-03-21 Thread Kenneth Graunke
The main GL context's swtnl_im field is the VBO module's vbo_context structure. Using the name "swtnl" in the name is confusing since some drivers use hardware texturing and lighting, but still rely on the VBO module for drawing. Signed-off-by: Kenneth Graunke --- src/mesa/main/mtypes.h | 2

Re: [Mesa-dev] [PATCH] mesa: Rename gl_context::swtnl_im to vbo_context.

2013-03-21 Thread Brian Paul
On 03/21/2013 12:19 PM, Kenneth Graunke wrote: The main GL context's swtnl_im field is the VBO module's vbo_context structure. Using the name "swtnl" in the name is confusing since some drivers use hardware texturing and lighting, but still rely on the VBO module for drawing. Signed-off-by: Ken

Re: [Mesa-dev] [PATCH] mesa: Delete VERT_ATTRIB_GENERIC_NV and VERT_BIT_GENERIC_NV macros.

2013-03-21 Thread Brian Paul
On 03/21/2013 12:10 PM, Kenneth Graunke wrote: These haven't been used since we deleted NV_vertex_program support. Signed-off-by: Kenneth Graunke --- src/mesa/main/mtypes.h | 10 -- 1 file changed, 10 deletions(-) diff --git a/src/mesa/main/mtypes.h b/src/mesa/main/mtypes.h index 790

[Mesa-dev] [PATCH] tgsi: fix regression introduced in 16caeff

2013-03-21 Thread Rob Clark
The recently added tgsi_declaration::Array field was not being initialized to zero, resulting in nonsense shaders like: FRAG PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1 DCL IN[0], ARRAY(48), GENERIC[0], CONSTANT -> DCL NULL[1..0]. 0: MOV OUT[0], IN[0] 1: END Signed-off-by: Rob Clark

Re: [Mesa-dev] [PATCH] tgsi: fix regression introduced in 16caeff

2013-03-21 Thread Brian Paul
I already fixed this in commit 460ae8a11a33340a979c7e0721f76e63a990 -Brian On 03/21/2013 12:39 PM, Rob Clark wrote: The recently added tgsi_declaration::Array field was not being initialized to zero, resulting in nonsense shaders like: FRAG PROPERTY FS_COLOR0_WRITES_ALL_CBUFS 1

Re: [Mesa-dev] [PATCH] tgsi: fix regression introduced in 16caeff

2013-03-21 Thread Rob Clark
ahh, so you did.. looks like I had rebased just before your patch BR, -R On Thu, Mar 21, 2013 at 2:47 PM, Brian Paul wrote: > I already fixed this in commit 460ae8a11a33340a979c7e0721f76e63a990 > > -Brian > > > On 03/21/2013 12:39 PM, Rob Clark wrote: >> >> The recently added tgsi_declaratio

Re: [Mesa-dev] [PATCH] mesa: Delete VERT_ATTRIB_GENERIC_NV and VERT_BIT_GENERIC_NV macros.

2013-03-21 Thread Eric Anholt
Kenneth Graunke writes: > These haven't been used since we deleted NV_vertex_program support. > > Signed-off-by: Kenneth Graunke Reviewed-by: Eric Anholt pgpvMo2WYmc2C.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesk

Re: [Mesa-dev] [PATCH] mesa: Rename gl_context::swtnl_im to vbo_context.

2013-03-21 Thread Eric Anholt
Kenneth Graunke writes: > The main GL context's swtnl_im field is the VBO module's vbo_context > structure. Using the name "swtnl" in the name is confusing since > some drivers use hardware texturing and lighting, but still rely on the > VBO module for drawing. > > Signed-off-by: Kenneth Graunke

Re: [Mesa-dev] [PATCH] r600g: Honour legacy debugging environment variables

2013-03-21 Thread Marek Olšák
Reviewed-by: Marek Olšák Marek On Thu, Mar 21, 2013 at 5:59 PM, Michel Dänzer wrote: > From: Michel Dänzer > > This helps minimize confusion / effort when moving between branches or > helping others. > > Signed-off-by: Michel Dänzer > --- > src/gallium/drivers/r600/r600_pipe.c | 10 +++

[Mesa-dev] [PATCH 1/4] cso: add constant buffer save/restore feature for postprocessing

2013-03-21 Thread Marek Olšák
Postprocessing is an internal meta op and should restore the states it changes. --- src/gallium/auxiliary/cso_cache/cso_context.c | 59 + src/gallium/auxiliary/cso_cache/cso_context.h | 13 ++ src/gallium/auxiliary/postprocess/pp_mlaa.c |4 +- src/gallium/auxi

[Mesa-dev] [PATCH 2/4] r300g: fix crash while binding a NULL constant buffer

2013-03-21 Thread Marek Olšák
--- src/gallium/drivers/r300/r300_state.c |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/gallium/drivers/r300/r300_state.c b/src/gallium/drivers/r300/r300_state.c index ad93510..2de0fd6 100644 --- a/src/gallium/drivers/r300/r300_state.c +++ b/src/gallium/drivers/r300/r

[Mesa-dev] [PATCH 3/4] r600g: fix crash while binding a NULL constant buffer

2013-03-21 Thread Marek Olšák
--- src/gallium/drivers/r600/r600_state_common.c |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/gallium/drivers/r600/r600_state_common.c b/src/gallium/drivers/r600/r600_state_common.c index b0e66ac..34c70ed 100644 --- a/src/gallium/drivers/r600/r600_state_common.c +++

[Mesa-dev] [PATCH 4/4] radeonsi: fix crash while binding a NULL constant buffer

2013-03-21 Thread Marek Olšák
--- src/gallium/drivers/radeonsi/si_state.c |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/gallium/drivers/radeonsi/si_state.c b/src/gallium/drivers/radeonsi/si_state.c index a395ec4..fee1b7f 100644 --- a/src/gallium/drivers/radeonsi/si_state.c +++ b/src/gallium/driver

[Mesa-dev] [Bug 62612] New: dispatch_sanity test lumps GL4.0+ functions into 4.3

2013-03-21 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=62612 Priority: medium Bug ID: 62612 Assignee: mesa-dev@lists.freedesktop.org Summary: dispatch_sanity test lumps GL4.0+ functions into 4.3 Severity: normal Classification: Unclassified

[Mesa-dev] [Bug 62612] dispatch_sanity test lumps GL4.0+ functions into 4.3

2013-03-21 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=62612 Matt Turner changed: What|Removed |Added CC||i...@freedesktop.org, |

Re: [Mesa-dev] [PATCH 2/4] i965: Create a pointer in brw_context to the geometry output VUE map.

2013-03-21 Thread Paul Berry
On 21 March 2013 08:56, Eric Anholt wrote: > Paul Berry writes: > > > Currently, the GPU pipeline has one active VUE map in effect at any > > given time--the one representing the layout of vertex data coming from > > the vertex shader. However, when geometry shaders are added, they > > will hav

[Mesa-dev] [PATCH] llvmpipe: add EXT_packed_float render target format support

2013-03-21 Thread sroland
From: Roland Scheidegger New conversion code to handle conversion from/to r11g11b10 AoS to/from SoA floats, and also add code for conversion from rgb9e5 AoS to float SoA (which works pretty much the same as r11g11b10 except for the packing). (This code should also be used for texture sampling ins

Re: [Mesa-dev] [PATCH] llvmpipe: add EXT_packed_float render target format support

2013-03-21 Thread Roland Scheidegger
Ok so before someone else notices that, ignore the rgb9e5 part. The format isn't quite what I thought it was... Roland Am 21.03.2013 23:28, schrieb srol...@vmware.com: > From: Roland Scheidegger > > New conversion code to handle conversion from/to r11g11b10 AoS to/from > SoA floats, and also ad

[Mesa-dev] [PATCH v2 1/5] i965: Move brw_vs_prog_data::outputs_written into VUE map.

2013-03-21 Thread Paul Berry
Future patches will allow for there to be separate VUE maps when both a geometry shader and a vertex shader are in use. When this happens, we will want to have correspondingly separate outputs_written bitfields. Moving outputs_written into the VUE map will make this easy. For consistency with th

[Mesa-dev] [PATCH v2 2/5] i965: Store the geometry output VUE map in brw_context.

2013-03-21 Thread Paul Berry
Currently, the GPU pipeline has one active VUE map in effect at any given time--the one representing the layout of vertex data coming from the vertex shader. However, when geometry shaders are added, they will have their own independent VUE map. Later pipeline stages (clip, sf, fs) will need to c

[Mesa-dev] [PATCH v2 3/5] i965: Use brw.vue_map_geom_out instead of VS output VUE map where appropriate.

2013-03-21 Thread Paul Berry
This patch modifies post-GS pipeline stages (transform feedback, clip, sf, fs) to refer to the VUE map through brw->vue_map_geom_out rather than brw->vs.prog_data->vue_map. This ensures that when geometry shader support is added, these pipeline stages will consult the geometry shader output VUE ma

[Mesa-dev] [PATCH v2 4/5] i965/fs: Rename vp_outputs_written to input_slots_valid.

2013-03-21 Thread Paul Berry
With the introduction of geometry shaders, fragment inputs will no longer come exclusively from the vertex shader; sometimes they come from the geometry shader. So the name "vp_outputs_written" will become a misnomer. This patch renames vp_outputs_written to input_slots_valid, to reflect the true

[Mesa-dev] [PATCH v2 5/5] i965: Shrink brw_vue_map struct.

2013-03-21 Thread Paul Berry
This patch changes the arrays in brw_vue_map (which only ever contain values from -1 to 58) from ints to signed chars. This reduces the size of the struct from 488 bytes to 136 bytes. --- src/mesa/drivers/dri/i965/brw_context.h | 4 ++-- src/mesa/drivers/dri/i965/brw_vs.c | 6 ++ 2 files

[Mesa-dev] [PATCH] llvmpipe: add EXT_packed_float render target format support

2013-03-21 Thread sroland
From: Roland Scheidegger New conversion code to handle conversion from/to r11g11b10 AoS to/from SoA floats, and also add code for conversion from rgb9e5 AoS to float SoA (which works pretty much the same as r11g11b10 except for the packing). (This code should also be used for texture sampling ins

Re: [Mesa-dev] [PATCH 4/4] radeonsi: fix crash while binding a NULL constant buffer

2013-03-21 Thread Michel Dänzer
On Don, 2013-03-21 at 20:27 +0100, Marek Olšák wrote: > --- > src/gallium/drivers/radeonsi/si_state.c |2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/src/gallium/drivers/radeonsi/si_state.c > b/src/gallium/drivers/radeonsi/si_state.c > index a395ec4..fee1b7f 100644 >