I followed up with a v2 that includes the bugzilla reference.
Good point, I’ll look into following up with a patch to remove Clip().
Thanks for the quick review.
> On Feb 2, 2017, at 2:26 PM, Ilia Mirkin wrote:
>
> Reviewed-by: Ilia Mirkin
>
> I got confused by this code as well when I was t
Reviewed-by: Bruce Cherniak
> On Feb 6, 2017, at 6:40 PM, Tim Rowley wrote:
>
> Highlights include more avx512 work and fixes for llvm>=3.8.
>
> Tim Rowley (10):
> swr: [rasterizer core] Frontend SIMD16 WIP
> swr: [rasterizer jitter] Adjust jitter header includes
> swr: [rasterizer jitter]
Reviewed-by: Bruce Cherniak
> On Feb 17, 2017, at 2:30 PM, George Kyriazis
> wrote:
>
> For now, the cache key is all of FETCH_COMPILE_STATE.
>
> Use new/delete for swr_vertex_element_state, since we have to call the
> constructors/destructors of the struct elements.
> ---
> src/gallium/drive
Reviewed-by: Bruce Cherniak
> On Feb 17, 2017, at 2:30 PM, George Kyriazis
> wrote:
>
> Fix issue with index buffers that do not contain a 0 index. 0 index
> can be a non-valid index if the (copied) vertex buffers are a subset of the
> user's (which happens because we only copy the range betw
Reviewed-by: Bruce Cherniak
> On Sep 20, 2016, at 2:24 PM, Tim Rowley wrote:
>
> Replace old string comparison with a mapping table.
> ---
> src/gallium/drivers/swr/swr_screen.cpp | 344 -
> 1 file changed, 293 insertions(+), 51 deletions(-)
>
> diff --git a/src
Reviewed-by: Bruce Cherniak
> On Oct 26, 2016, at 7:08 PM, George Kyriazis
> wrote:
>
> This allows the post-processor to easily detect the API thread and to
> process frame information. The frame information is needed to
> optimized how data is processed from worker threads.
> ---
> src/gall
Reviewed-by: Bruce Cherniak
> On Oct 26, 2016, at 7:08 PM, George Kyriazis
> wrote:
>
> Used in abandoned all-or-nothing approach to converting to AVX512
> ---
> .../drivers/swr/rasterizer/common/simdintrin.h | 633 -
> .../drivers/swr/rasterizer/core/format_types.h
Reviewed-by: Bruce Cherniak
> On Oct 26, 2016, at 7:08 PM, George Kyriazis
> wrote:
>
> Add frontend dependency concept in the DRAW_CONTEXT, which
> allows serialization of frontend work if necessary.
> ---
> src/gallium/drivers/swr/rasterizer/core/api.cpp | 3 +++
> src/gallium/drivers/s
Reviewed-by: Bruce Cherniak
> On Oct 26, 2016, at 7:08 PM, George Kyriazis
> wrote:
>
> Used for common code reuse and simplification
> ---
> .../drivers/swr/rasterizer/core/backend.cpp| 561 -
> src/gallium/drivers/swr/rasterizer/core/backend.h | 150 +-
> 2 fi
Reviewed-by: Bruce Cherniak
> On Oct 26, 2016, at 7:08 PM, George Kyriazis
> wrote:
>
> ---
> .../rasterizer/scripts/templates/ar_eventhandlerfile_h.template| 7 ++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git
> a/src/gallium/drivers/swr/rasterizer/scripts/template
Reviewed-by: Bruce Cherniak
> On Nov 7, 2016, at 1:23 PM, Tim Rowley wrote:
>
> We need to compile a blend function when alphatest is enabled.
> ---
> src/gallium/drivers/swr/swr_state.cpp | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/src/gallium/drivers/swr/swr_st
Reviewed-by: Bruce Cherniak
> On Nov 9, 2016, at 9:18 PM, Tim Rowley wrote:
>
> Highlights include more 8x2 render work, fix for alphatest, and
> archrast updates.
>
> Tim Rowley (14):
> swr: [rasterizer jitter] code style fix
> swr: [rasterizer core] 16-wide tile store nearly completed
>
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> This is the format used for the primary surface of a
> PIPE_FORMAT_Z32_FLOAT_S8X24_UINT resource.
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/rasterizer/memory/LoadTile.h | 1 +
> src/gallium
Reviewing/commenting on each patch individually. We’re also testing more
thoroughly in our CI.
In general, your help and interest in OpenSWR is very much appreciated!
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> This is a bit of a hodge-podge, but largely related to improving text
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Noticed by inspection.
>
> Signed-off-by: Ilia Mirkin
> ---
>
> Entirely untested. Just happened to jump out at me that the two tiling modes
> didn't match, whereas every other instance in this file matches.
>
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> In a BGR10X2 or BGR5X1 situation, there's no need to try to quantize the
> X channel - the default will have the proper quantization required.
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/raste
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_state.cpp | 20 +++-
> 1 file changed, 15 insertions(+), 5 deletions(-)
>
> diff --git a/src/gallium/drivers/swr/swr_state.cpp
> b/s
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Fixes the texsubimage piglit and lets the copyteximage one get further.
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_context.cpp | 5 +++--
> src/gallium/drivers/swr/swr_screen.cpp | 5 +++
Doesn’t look as though shader stencil export (GL_ARB_shader_stencil_export)
would be too hard to implement, until then best to disable it.
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_screen.c
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/.gitignore | 4
> 1 file changed, 4 insertions(+)
>
> diff --git a/src/gallium/drivers/swr/.gitignore
> b/src/gallium/drivers/swr/.gitignore
> inde
We need to run this through CI for all core users.
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> There's no guarantee that mip width/height will be a multiple of the
> compressed block size. Make sure to round up when dividing.
>
> Signed-off-by: Ilia Mirkin
> ---
>
> Note - I don't a
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> The support in swr requires shaders to output the components as UINTs.
> This is not how GL or Gallium work, and since this is not a
> required-renderable format, just leave it out.
>
> Signed-off-by: Ilia Mirkin
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> The EXT_texture_integer test says that blending and alphatest should
> all be disabled. st/mesa takes care of alphatest already.
>
> Fixes the ext_texture_integer-fbo-blending piglit test.
>
> Signed-off-by: Ili
This one is going to take a bit more testing and internal review before
adopting.
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> This is a bit of a mega-commit, but unfortunately there's no great way
> to break this up since a lot of different pieces have to match up. Here
> we do the fo
This table lists all supported formats (both renderable and texturable).
swr_texture_layout calls mesa_to_swr_format to convert the PIPE_FORMAT_* to the
appropriate SWR_FORMAT enum. Removing these entries would result in -1
(unsupported format). We might need to add a “is_renderable” field to
For book-keeping, to know which formats we have left to support/enable, we’d
like to leave the -1 entries. It’s a little clutter, but the sore thumb gets
the hammer.
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_screen.cp
This works for now, but totally agree with you, we need work attached to
fences. That’s been high on my list.
Reviewed-by: Bruce Cherniak
> On Nov 12, 2016, at 5:00 PM, Ilia Mirkin wrote:
>
> Rendering could still be ongoing (or have yet to start) when the shader
> is deleted. There's no ref
the texture width/height need not be a multiple of the
format's block size. However mip sizes are computed based on that unrounded
size.
On Nov 14, 2016 8:45 PM, "Cherniak, Bruce"
mailto:bruce.chern...@intel.com>> wrote:
We need to run this through CI for all core users.
>
hose in a comment?
Either way, the most important change is to remove all the l/la/i swr format
enums (I.e. make the function return -1) since they only have load, not store
implementations. If you still disagree with the above reasoning, I'll redo that
bit of the series.
On Nov 14, 2
16 8:45 PM, "Cherniak, Bruce"
mailto:bruce.chern...@intel.com>> wrote:
Reviewing/commenting on each patch individually. We’re also testing more
thoroughly in our CI.
In general, your help and interest in OpenSWR is very much appreciated!
> On Nov 12, 2016, at 5:00 P
As per our IRC discussion…
Reviewed-by: Bruce Cherniak
> On Nov 14, 2016, at 7:46 PM, Cherniak, Bruce wrote:
>
> For book-keeping, to know which formats we have left to support/enable, we’d
> like to leave the -1 entries. It’s a little clutter, but the sore thumb gets
Looks good (and concise), with the exception of a few styling issues
(suggestions inlined)
Reviewed-by: Bruce Cherniak after minor styling
fixes.
> On Nov 15, 2016, at 8:26 PM, Tim Rowley wrote:
>
> Fixes fbo-mrt-alphatest
> ---
> src/gallium/drivers/swr/rasterizer/core/api.cpp| 10 ++---
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 7:49 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_state.cpp | 7 +++
> 1 file changed, 7 insertions(+)
>
> diff --git a/src/gallium/drivers/swr/swr_state.cpp
> b/src/gallium/drivers/swr/swr_sta
Reviewed-by: Bruce Cherniak
> On Nov 20, 2016, at 4:25 PM, Ilia Mirkin wrote:
>
> Multiple buffers may reference client arrays. When this happens, we
> might reach for scratch space multiple times, which could cause later
> arrays to invalidate the pointers allocated for the earlier ones.
>
>
Reviewed-by: Bruce Cherniak
> On Nov 19, 2016, at 11:09 PM, Ilia Mirkin wrote:
>
> Currently a sequence like draw/map/draw/map will cause the second map to
> not wait for the second draw. This is because the first map will clear
> the resource business bit, and the second draw won't reset it si
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 10:56 PM, Ilia Mirkin wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/drivers/swr/swr_screen.cpp | 126 -
> 1 file changed, 126 deletions(-)
>
> diff --git a/src/gallium/drivers/swr/swr_screen.cpp
> b/
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 10:56 PM, Ilia Mirkin wrote:
>
> This is a bit of a mega-commit, but unfortunately there's no great way
> to break this up since a lot of different pieces have to match up. Here
> we do the following:
> - change surface layout to match swr's Lo
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 10:56 PM, Ilia Mirkin wrote:
>
> I was getting a weird segfault from GCC 4.9.3:
>
> 0x754f27aa in strlen () from /lib64/libc.so.6
> (gdb) bt
> #0 0x754f27aa in strlen () from /lib64/libc.so.6
> #1 0x74f128e5 in get_cie
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 10:56 PM, Ilia Mirkin wrote:
>
> This table exists for the purpose of determining renderable formats.
> Without a StoreTile implementation, that can't happen.
>
> This basically removes rendering support to all L/LA/I formats. They can
> be re-
The entire set has now been r-v-b.
Reviewed-by: Bruce Cherniak
> On Nov 17, 2016, at 10:56 PM, Ilia Mirkin wrote:
>
> Some of these are new, others have had some changes done to them. Most
> notably,
> I've fixed the 1D layout in both the TilingFunctions logic as well as the
> frontend.
>
> A
oops.. obvious typo.
Reviewed-by: Bruce Cherniak
> On Nov 25, 2016, at 8:09 PM, Ilia Mirkin wrote:
>
> Among other things, blits would clear existing SO targets which would
> cause a bunch of updates from u_blitter to be missed.
>
> Fixes fbo-scissor-blit fbo, probably among many others.
>
Reviewed-by: Bruce Cherniak
> On Nov 22, 2016, at 8:03 PM, Rowley, Timothy O
> wrote:
>
> ---
> .../drivers/swr/rasterizer/common/swr_assert.cpp | 96 ++
> .../drivers/swr/rasterizer/common/swr_assert.h | 14
> 2 files changed, 95 insertions(+), 15 deletions(-)
>
Reviewed-by: Bruce Cherniak
> On Nov 29, 2016, at 8:18 PM, Ilia Mirkin wrote:
>
> We need to take the instance divisor and number of instances into
> account for instanced client-side arrays, rather than the vertex
> parameters.
>
> Loosely based on the comparable nvc0 logic.
>
> Signed-off-
Reviewed-by: Bruce Cherniak
> On Dec 2, 2016, at 10:53 AM, Rowley, Timothy O
> wrote:
>
> Fixes build problem with llvm-svn.
> ---
> src/gallium/drivers/swr/rasterizer/jitter/builder_misc.cpp | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/src/gallium/drivers/swr/ra
Reviewed-by: Bruce Cherniak
> On Dec 2, 2016, at 1:28 PM, Rowley, Timothy O
> wrote:
>
> Fixes build problem with llvm-svn.
>
> v2: use cstdarg instead of stdarg.h
> ---
> src/gallium/drivers/swr/rasterizer/jitter/builder_misc.cpp | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
>
Reviewed-by: Bruce Cherniak
> On Dec 5, 2016, at 11:40 AM, Tim Rowley wrote:
>
> ---
> src/gallium/drivers/swr/swr_screen.cpp | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/src/gallium/drivers/swr/swr_screen.cpp
> b/src/gallium/drivers/swr/swr_screen.cpp
> index 539acf1..b5c2cd3 100
Reviewed-by: Bruce Cherniak
> On Dec 5, 2016, at 11:39 AM, Rowley, Timothy O
> wrote:
>
> Uses llvmpipe's string formating.
> ---
> src/gallium/drivers/swr/swr_screen.cpp | 12 +++-
> 1 file changed, 11 insertions(+), 1 deletion(-)
>
> diff --git a/src/gallium/drivers/swr/swr_screen.cp
Reviewed-by: Bruce Cherniak
> On Dec 7, 2016, at 7:16 PM, Tim Rowley wrote:
>
> v2: use fmul(1/65536) instead of fdiv(65535)
> ---
> .../drivers/swr/rasterizer/common/formats.cpp | 104 ++---
> .../drivers/swr/rasterizer/common/formats.h| 7 +-
> .../drivers/swr/ra
Reviewed-by: Bruce Cherniak
> On Dec 7, 2016, at 5:55 PM, Rowley, Timothy O
> wrote:
>
> This one-off specialization was missed.
> ---
> src/gallium/drivers/swr/rasterizer/core/backend.h | 4 +-
> .../drivers/swr/rasterizer/memory/StoreTile.h | 46 +++---
> 2 files change
Reviewed-by: Bruce Cherniak
> On Dec 12, 2016, at 1:30 PM, Rowley, Timothy O
> wrote:
>
> Fixes to 128-bit formats.
> ---
> src/gallium/drivers/swr/rasterizer/core/utils.h| 8 +-
> .../drivers/swr/rasterizer/memory/StoreTile.h | 352 -
> 2 files changed, 138 inser
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> Fixes relocation errors with LLVM 7.0.0
> ---
> .../swr/rasterizer/jitter/JitManager.cpp | 79 +++
> .../swr/rasterizer/jitter/JitManager.h| 28 +--
> 2 files changed, 65 insertions(+
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> - updated sample from TRTT surfaces correctly
> - implemented mapped status return for TRTT surfaces
> - implemented per-sample instruction minLod clamp
> - updated bilinear filter weight calculation to be closer t
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> This improves cache filesystem performance, especially during CI tests
> Also updated jitcache magic number due to codegen parameter changes
> Removed 2 `if constexpr` to prevent C++17 requirement
> ---
> .../swr/r
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> Use gfxptr_t type value for stream pointer uses in gather and similar
> calls
> ---
> .../swr/rasterizer/jitter/fetch_jit.cpp | 34 +--
> 1 file changed, 16 insertions(+), 18 deletions(-)
>
>
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> - added graphics address translation in odd gathers
> - added support for unaligned gathers in fetch shader
> - changed how 2+ GB offsets are handled to make them compatible with
> unaligned offsets
> ---
> .../swr
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> Avoids confusion with other defaulted integer parameters
>
> - fixed some unspecified usages
> - removed unnecessary includes
> - removed unecessary protected access specifier in buckets framework
> ---
> .../driv
Reviewed-by: Bruce Cherniak
> On Dec 17, 2018, at 8:36 AM, Alok Hota wrote:
>
> To make debugging simpler
> ---
> .../swr/rasterizer/jitter/JitManager.cpp | 27 +--
> .../swr/rasterizer/jitter/JitManager.h| 12 -
> 2 files changed, 36 insertions(+), 3 deletio
Reviewed-by: Bruce Cherniak
On 4/13/16, 10:48 AM, "mesa-dev on behalf of George Kyriazis"
wrote:
>- Incorporate flatshade flag into the shader generation
>- Use provoking vertex (vc) in shader when flat shading.
>---
> src/gallium/drivers/swr/swr_shader.cpp | 4
> src/gallium/drivers/s
Reviewed-by: Bruce Cherniak
On 4/19/16, 12:25 PM, "mesa-dev on behalf of Tim Rowley"
wrote:
>Code was using an incorrect address for the base pointer.
>
>v2: use swr_resource_data() utility function.
>
>Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94979
>---
> src/gallium/drivers/
>
> On Apr 21, 2016, at 11:20 AM, Ilia Mirkin wrote:
>
> On Thu, Apr 21, 2016 at 12:17 PM, Tim Rowley
> wrote:
>> ---
>> src/gallium/drivers/swr/swr_query.cpp | 2 +-
>> src/gallium/drivers/swr/swr_shader.h | 8
>> 2 files changed, 5 insertions(+), 5 deletions(-)
>>
>> diff --git a/sr
Tested-by: Bruce Cherniak
On 4/22/16, 8:33 AM, "mesa-dev on behalf of srol...@vmware.com"
wrote:
>From: Roland Scheidegger
>
>Some cases (especially these using fract for coord wrapping) did not handle
>NaNs (or Infs) correctly - the following code assumed the fract result
>could not be o
Reviewed-by: Bruce Cherniak
mailto:bruce.chern...@intel.com>>
On Apr 21, 2016, at 11:50 AM, Tim Rowley
mailto:timothy.o.row...@intel.com>> wrote:
v2: use alternate logic version in swr_check_render_cond
---
src/gallium/drivers/swr/swr_query.cpp | 2 +-
src/gallium/drivers/swr/swr_shader.h | 8 +
Reviewed-by: Bruce Cherniak On Apr 14, 2016, at 2:53 PM, Tim Rowley wrote:
>
> Another batch of updates to the swr rasterizer. No huge new features,
> but should fix at least one of the coverity flagged issues.
>
> Tim Rowley (11):
> swr: [rasterizer core] Use CS spill/fill size in core
> sw
Reviewed-by: Bruce Cherniak , with addition of
.gitignore on commit.
-Original Message-
From: mesa-dev [mailto:mesa-dev-boun...@lists.freedesktop.org] On Behalf Of Tim
Rowley
Sent: Tuesday, April 26, 2016 11:55 AM
To: mesa-dev@lists.freedesktop.org
Subject: [Mesa-dev] [PATCH] swr: autoge
Reviewed-by: Bruce Cherniak
On 3/8/16, 11:50 AM, "mesa-dev on behalf of Tim Rowley"
wrote:
>Remove use of a win32-style type leaked from the swr rasterizer.
>---
> src/gallium/drivers/swr/swr_memory.h| 8
> src/gallium/drivers/swr/swr_scratch.cpp | 8
> src/gallium/drive
Reviewed-by: Bruce Cherniak
On 3/14/16, 5:40 PM, "mesa-dev on behalf of George Kyriazis"
wrote:
>Removed bound_to_context. We now pick up the context from the screen
>instead of the resource itself. The resource could be out-of-date
>and point to a pipe that is already freed.
>
>Fixes m
Reviewed-by: Bruce Cherniak
> On Jun 3, 2016, at 12:39 PM, Rowley, Timothy O
> wrote:
>
> Use rasterizer provoking vertex API.
>
> Fix rasterizer provoking vertex for tristrips and quad list/strips.
>
> v2: make provoking vertex tables static const
> ---
> .../drivers/swr/rasterizer/core/fr
Reviewed-by: Bruce Cherniak
> On Jun 8, 2016, at 9:59 AM, Tim Rowley wrote:
>
> v2: only load the clip vertex once
>
> v3: fix clip enable logic, add cullDistance
>
> v4: remove duplicate fields in vs jit key, fix test of clip fixup needed
>
> v5: fix clipdistance linkage for slot!=0,4
>
>
> On Jun 17, 2016, at 8:56 AM, Rowley, Timothy O
> wrote:
>
>
>> On Jun 17, 2016, at 8:58 AM, Bruce Cherniak wrote:
>>
>> A pipe pointer in the screen allows for access to current device context
>> in flush_frontbuffer and resource_destroy. This wasn't tracking current
>> context in multi-c
Reviewed-by: Bruce Cherniak
> On Jun 20, 2016, at 4:36 PM, Tim Rowley wrote:
>
> v2:
> add conservativeRast.h to Makefile.sources
> minimize changes in llvm support cleanup
> remove tabs that were added by the v1 patches
>
> Tim Rowley (14):
> swr: [rasterizer common] workar
Reviewed-by: Bruce Cherniak
> On Jun 17, 2016, at 11:01 AM, Tim Rowley wrote:
>
> llvm redefines DEBUG; adding push/pop prevents a undefined reference
> to debug_refcnt_state in llvm-3.7+.
>
> v2: add undef DEBUG
> ---
> src/gallium/drivers/swr/swr_shader.cpp | 10 +++---
> src/gallium/driv
On May 26, 2016, at 3:06 PM, Tim Rowley
mailto:timothy.o.row...@intel.com>> wrote:
v2: style code, add avx512 to cpu dump
---
src/gallium/auxiliary/util/u_cpu_detect.c | 26 ++
src/gallium/auxiliary/util/u_cpu_detect.h | 10 ++
2 files changed, 36 insertions(+)
di
Reviewed-by: Bruce Cherniak
> On May 27, 2017, at 4:24 PM, Tim Rowley wrote:
>
> Highlights include lots of simd16 progress and code cleanups.
>
> No regressions on piglit or vtk ctest.
>
> Tim Rowley (24):
> swr/rast: remove extra pixel center adjustment in BinPostSetupPoints
> swr/rast: a
Reviewed-by: Bruce Cherniak
> On Jun 8, 2017, at 5:42 PM, Tim Rowley wrote:
>
> Remove c++14 generic lambda to keep compiler requirement at c++11.
>
> No regressions on piglit or vtk test suites.
>
> Tested-by: Chuck Atkins
>
> CC: mesa-sta...@lists.freedesktop.org
> ---
> configure.ac
Reviewed-by: Bruce Cherniak
> On Oct 19, 2017, at 8:12 AM, Tim Rowley wrote:
>
> Highlights are code cleanups, some more simd16 work (disabled by default),
> and tuning for the Intel Xeon Phi architecture.
>
> Tim Rowley (7):
> swr/rast: Minor changes for os-x
> swr/rast: Miscellaneous vie
Reviewed-by: Bruce Cherniak
> On Oct 19, 2017, at 4:40 PM, George Kyriazis
> wrote:
>
> Remove allocation of > 2kbyte buffers into context memory in
> swr_copy_to_scatch_space() (which is used to copy small vertex/index buffers
> and shader constants to a scratch space to be used by the upcom
Reviewed-by: Bruce Cherniak
> On Jul 14, 2017, at 3:01 PM, Tim Rowley wrote:
>
> Fixes performance regression from f50aa21456d - was forcing internal
> code generation to target AVX (no gather, etc).
> ---
> src/gallium/drivers/swr/swr_screen.cpp | 3 ++-
> 1 file changed, 2 insertions(+), 1 de
Reviewed-by: Bruce Cherniak
> On Jul 20, 2017, at 5:09 PM, Tim Rowley wrote:
>
> Prevents unalignment crashes with avx512 code on gcc/clang.
> ---
> src/gallium/drivers/swr/rasterizer/core/tilemgr.cpp | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/src/gallium/dr
Reviewed-by: Bruce Cherniak
> On Jul 20, 2017, at 5:09 PM, Tim Rowley wrote:
>
> Source/destination will not be AVX512 aligned, use the
> unaligned load/store intrinsics.
> ---
> .../drivers/swr/rasterizer/common/simdlib_128_avx512.inl | 10 +-
> .../drivers/swr/rasterizer/common/
Reviewed-by: Bruce Cherniak
> On Jul 20, 2017, at 5:09 PM, Tim Rowley wrote:
>
> Tested with clang-4.0 and gcc-6.3.
> ---
> .../swr/rasterizer/common/simdlib_512_avx512.inl | 43 +-
> .../swr/rasterizer/common/simdlib_types.hpp| 2 +-
> 2 files changed, 35 insertio
Reviewed-by: Bruce Cherniak
> On Jul 21, 2017, at 11:46 AM, Tim Rowley wrote:
>
> Linux-specific gettid() syscall shouldn't be used in portable code.
> Fix does assume a 1:1 thread:LWP architecture, but works for our
> current target platforms and can be revisited later if needed.
>
> Fixes u
Reviewed-by: Bruce Cherniak
> On Jul 21, 2017, at 2:17 PM, Tim Rowley wrote:
>
> Linux-specific gettid() syscall shouldn't be used in portable code.
> Fix does assume a 1:1 thread:LWP architecture, but works for our
> current target platforms and can be revisited later if needed.
>
> Fixes unr
> On Jul 24, 2017, at 10:22 AM, Roland Scheidegger wrote:
>
> Am 24.07.2017 um 15:26 schrieb Jose Fonseca:
>> In particular:
>>
>> 1) For buffers the box expresses bytes, not pixels.
> This shouldn't be an issue, since buffers are required to have a 1-byte
> format (usually R8_UINT, but R8_UNOR
Reviewed-by: Bruce Cherniak
> On Jul 26, 2017, at 4:38 PM, Tim Rowley wrote:
>
> gcc prior to 4.9 didn't implement , causing a startup crash
> in the swr knob parameter reading code.
>
> CC:
> ---
> .../swr/rasterizer/codegen/templates/gen_knobs.cpp | 21 +
> 1 file chang
Reviewed-by: Bruce Cherniak
> On Jul 26, 2017, at 5:27 PM, George Kyriazis
> wrote:
>
> The shader that is used to copy vertex data out of the vs/gs shaders to
> the user-specified buffer (streamout or SO shader) was not using the
> correct offsets.
>
> Adjust the offsets that are used just
Entire set
Reviewed-by: Bruce Cherniak
> On Aug 1, 2017, at 2:44 PM, Tim Rowley wrote:
>
> Highlights include enabling the simd16 frontend and code cleanups.
>
> v2: split previous gen_knobs patch into five chunks: scons fix,
> relicense to match rest of tree, code style, template split,
>
>
> On Aug 2, 2017, at 10:30 PM, Brian Paul wrote:
>
> I'll let Charmaine review, but the subject line prefix should probably be
> "st/osmesa".
Thanks, I'll clarify the subject line on commit.
> -Brian
>
> On 08/02/2017 05:28 PM, Bruce Cherniak wrote:
>> Commit bbc29393d3 didn't include osme
Reviewed-by: Bruce Cherniak
> On Aug 9, 2017, at 5:22 PM, Tim Rowley wrote:
>
> CID: 1416243, 1416244, 1416255
> CC: mesa-sta...@lists.freedesktop.org
> ---
> src/gallium/drivers/swr/rasterizer/core/api.cpp | 2 +-
> src/gallium/drivers/swr/rasterizer/core/context.h | 8
> src/gall
>
> On Apr 24, 2017, at 10:16 AM, Brian Paul wrote:
>
> On 04/21/2017 12:25 AM, green we wrote:
>> I tried to use OSMesa to do offscreen rendering, refer to the osdemo and
>> the tutorial here: http://www.alecjacobson.com/weblog/?p=2827
>> There are obvious serrations on the edge of the model.
>
Hi Emil,
Thanks for the comments, I’ll work on improving my commit messages.
In this case, the single change to fix crashes yields all 3 improvements —
tests were crashing because our format types were limited. The additional
benefit of supporting these types is much increased performance. T
Entire set reviewed-by: Bruce Cherniak
> On Apr 26, 2017, at 1:33 PM, Tim Rowley wrote:
>
> v2:
> Rewrite commit messages
> Split some commits
> No code/functional changes
>
> Tim Rowley (17):
> swr/rast: more flexible max attribute slots
> swr/rast: add additional jit utility functions
> s
> On Apr 27, 2017, at 7:38 PM, Ilia Mirkin wrote:
>
> Erm, so ... what happens if I render to FB1, then render to FB2, then
> render to FB1 again (and I have blending enabled)? Doesn't the resolve
> lose the per-sample information? Or does the resolve merely precompute
> the resolved version on
> On Apr 27, 2017, at 7:50 PM, Ilia Mirkin wrote:
>
> On Thu, Apr 27, 2017 at 8:45 PM, Cherniak, Bruce
> wrote:
>>
>>> On Apr 27, 2017, at 7:38 PM, Ilia Mirkin wrote:
>>>
>>> Erm, so ... what happens if I render to FB1, then render to FB2, then
> On Apr 28, 2017, at 3:20 PM, Ilia Mirkin wrote:
>
> On Fri, Apr 28, 2017 at 3:58 PM, Cherniak, Bruce
> wrote:
>>
>>> On Apr 27, 2017, at 7:50 PM, Ilia Mirkin wrote:
>>>
>>> On Thu, Apr 27, 2017 at 8:45 PM, Cherniak, Bruce
>>> wrote
> PART 2: Call for testing
>
> These drivers have been tested:
> - ddebug
> - llvmpipe
> - r300 (also with SWTCL)
> - r600
> - radeonsi
> - softpipe
> - trace
>
> These drivers need testing:
> - etnaviv
> - freedreno
> - nv30
> - nv50
> - nvc0
> - svga
> - swr
> - vc4
> - virgl
Tested on swr. N
> On May 2, 2017, at 3:41 PM, Cherniak, Bruce wrote:
>
>> PART 2: Call for testing
>>
>> These drivers have been tested:
>> - ddebug
>> - llvmpipe
>> - r300 (also with SWTCL)
>> - r600
>> - radeonsi
>> - softpipe
>> - tr
On May 2, 2017, at 4:11 PM, Cherniak, Bruce
mailto:bruce.chern...@intel.com>> wrote:
On May 2, 2017, at 3:41 PM, Cherniak, Bruce
mailto:bruce.chern...@intel.com>> wrote:
PART 2: Call for testing
These drivers have been tested:
- ddebug
- llvmpipe
- r300 (also with SWTCL)
- r60
> On May 2, 2017, at 5:33 PM, Marek Olšák wrote:
>
> On Wed, May 3, 2017 at 12:18 AM, Cherniak, Bruce
> wrote:
>>
>> On May 2, 2017, at 4:11 PM, Cherniak, Bruce
>> wrote:
>>
>>
>> On May 2, 2017, at 3:41 PM, Cherniak, Bruce
>> wrote
Reviewed-by: Bruce Cherniak
> On May 4, 2017, at 10:35 AM, George Kyriazis
> wrote:
>
> Rasterizer core only supports polygonmode front==back. Add logic for
> populating fillMode for the rasterizer only for that case correctly.
> Provide enum conversion between mesa enums and core enums.
>
Reviewed-by: Bruce Cherniak
> On May 4, 2017, at 10:35 AM, George Kyriazis
> wrote:
>
> Add support for polygonmode point in the binner. This is done by
> splitting BinPostSetupPoints from BinPoints, so the earlier call can be
> called from BinTriangles. Setup has already been done at the ti
1 - 100 of 215 matches
Mail list logo
