OSDN Git Service
Rob Clark [Tue, 15 Sep 2015 22:55:48 +0000 (18:55 -0400)]
nir/print: print symbolic names from shader-enum
v2: split out moving of FILE *fp into state structure into it's own
(more complete patch) to reduce the noise in this one
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Rob Clark [Tue, 15 Sep 2015 22:50:41 +0000 (18:50 -0400)]
nir/print: bit of state refactoring
Rename print_var_state to print_state, and stuff FILE ptr into the state
object. This avoids passing around an extra parameter everywhere.
v2: even more extensive conversion.. use state *everywhere* instead of
FILE ptr, and convert nir_print_instr() to use state as well
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Rob Clark [Fri, 11 Sep 2015 16:48:05 +0000 (12:48 -0400)]
glsl: shader-enum to name debug fxns
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Rob Clark [Fri, 4 Sep 2015 15:35:33 +0000 (11:35 -0400)]
freedreno: one screen to rule them all
Similar to
fee0686c21c631d96d6042741267a3c218c23ffc, but in this case to
ensure that drm_gralloc and libGLES_mesa are sharing a single screen.
Bumps libdrm_freedreno version dependency, as it requires the new
fd_device_fd() API.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Mon, 14 Sep 2015 15:54:05 +0000 (11:54 -0400)]
freedreno/ir3: use NIR to lower ffract instead of tgsi_lowering
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Mon, 14 Sep 2015 15:13:19 +0000 (11:13 -0400)]
nir: add lowering for ffract
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Jordan Justen [Tue, 15 Sep 2015 21:01:17 +0000 (14:01 -0700)]
i965/fs: The barrier send uses only 1 payload register
When preparing the barrier payload, the instructions should operate in
simd8 mode since we only use 1 payload register.
fs_inst::regs_read is also updated to indicate that it only reads one
register for SHADER_OPCODE_BARRIER.
These issues were flagged by:
commit
cadd7dd384b33a779d46bd664f456bed4a21a5b7
Author: Jason Ekstrand <jason.ekstrand@intel.com>
Date: Thu Jul 2 15:41:02 2015 -0700
i965/fs: Add a very basic validation pass
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Jason Ekstrand [Tue, 15 Sep 2015 19:09:06 +0000 (12:09 -0700)]
nir/builder: Use a normal temporary array in nir_channel
C++ gets cranky if we take references of temporaries. This isn't a problem
yet in master because nir_builder is never used from C++. However, it will
be in the future so we should fix it now.
Reviewed-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Mon, 14 Sep 2015 19:15:06 +0000 (15:15 -0400)]
freedreno/a4xx: more texture formats
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 15 Sep 2015 21:25:47 +0000 (17:25 -0400)]
freedreno/a4xx: border-color support
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 15 Sep 2015 21:25:25 +0000 (17:25 -0400)]
freedreno/a4xx: wire up texture clamp lowering
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 15 Sep 2015 13:23:21 +0000 (09:23 -0400)]
freedreno: helper for a3xx/a4xx border-colors
Both use the same layout for the buffer containing border-color values,
so rather than duplicating the logic in a4xx, split it out into a
helper.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Mon, 14 Sep 2015 20:59:36 +0000 (16:59 -0400)]
freedreno: update generated headers
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Jason Ekstrand [Thu, 10 Sep 2015 00:18:55 +0000 (17:18 -0700)]
nir/lower_vec_to_movs: Coalesce into destinations of fdot instructions
Now that we have a replicating fdot instruction, we can actually coalesce
into the destinations of vec4 instructions. We couldn't really do this
before because, if the destination had to end up in .z, we couldn't
reswizzle the instruction. With a replicated destination, the result ends
up in all channels so we can just set the writemask and we're done.
Shader-db results for vec4 programs on Haswell:
total instructions in shared programs:
1747753 ->
1746280 (-0.08%)
instructions in affected programs: 143274 -> 141801 (-1.03%)
helped: 667
HURT: 0
It turns out that dot-products matter...
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Thu, 10 Sep 2015 18:08:15 +0000 (11:08 -0700)]
i965/vec4: Use the replicated fdot instruction in NIR
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Thu, 10 Sep 2015 17:51:46 +0000 (10:51 -0700)]
nir: Add a fdot instruction that replicates the result to a vec4
Fortunately, nir_constant_expr already auto-splats if "dst" never shows up
in the constant expression field so we don't need to do anything there.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 21:40:06 +0000 (14:40 -0700)]
nir/lower_vec_to_movs: Coalesce movs on-the-fly when possible
The old pass blindly inserted a bunch of moves into the shader with no
concern for whether or not it was really needed. This adds code to try and
coalesce into the destination of the instruction providing the value.
Shader-db results for vec4 shaders on Haswell:
total instructions in shared programs:
1754420 ->
1747753 (-0.38%)
instructions in affected programs: 231230 -> 224563 (-2.88%)
helped: 1017
HURT: 2
This approach is heavily based on a different patch by Eduardo Lima Mitev
<elima@igalia.com>. Eduardo's patch did this in a separate pass as opposed
to integrating it into nir_lower_vec_to_movs.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 21:47:28 +0000 (14:47 -0700)]
nir/lower_vec_to_movs: Get rid of start_idx and swizzle compacting
Previously, we did this thing with keeping track of a separate start_idx
which was different from the iteration variable. I think this was a relic
of the way that GLSL IR implements writemasks. In NIR, if a given bit in
the writemask is unset then that channel is just "unused", not missing. In
particular, a vec4 operation with a writemask of 0xd will use sources 0, 2,
and 3 and leave source 1 alone. We can simplify things a good deal (and
make them correct) by removing this "compacting" step.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Jason Ekstrand [Wed, 9 Sep 2015 20:55:39 +0000 (13:55 -0700)]
i965/vec4_nir: Use partial SSA form rather than full non-SSA
We made this switch in the FS backend some time ago and it seems to make a
number of things a bit easier. In particular, supporting SSA values takes
very little work in the backend and allows us to take advantage of the
majority of the SSA information even after we've gotten rid of Phi nodes.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 20:42:14 +0000 (13:42 -0700)]
nir/lower_vec_to_movs: Handle partially SSA shaders
v2 (Jason Ekstrand):
- Use nir_instr_rewrite_dest
- Pass the impl directly into lower_vec_to_movs_block
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 19:58:58 +0000 (12:58 -0700)]
nir/lower_vec_to_movs: Pass the shader around directly
Previously, we were passing the shader around, we were just calling it
"mem_ctx". However, the nir_shader is (and must be for the purposes of
mark-and-sweep) the mem_ctx so we might as well pass it around explicitly.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Thu, 2 Jul 2015 22:41:02 +0000 (15:41 -0700)]
i965/fs: Add a very basic validation pass
Currently the validation pass only validates that regs_read and
regs_written are consistent with the sizes of VGRF's. We can add more as
we find it to be useful.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Mon, 14 Sep 2015 22:36:24 +0000 (15:36 -0700)]
i965/fs_surface_builder: Only apply predicate to components that exist
In certain conditions, we have to do bounds-checking in the shader for
image_load_store. The way this works for image loads is that we do a
predicated load and then emit a series of selects, one per component,
that gives us 0 or the loaded value depending on whether or not you're
in bounds. However, we were hard-coding 4 components which may not be
correct. Instead, we should be using size which is the number of
components read.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Jason Ekstrand [Mon, 14 Sep 2015 21:18:13 +0000 (14:18 -0700)]
i965/fs: Only read output_components many components when writing an output
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jason Ekstrand [Mon, 14 Sep 2015 22:09:00 +0000 (15:09 -0700)]
i965/fs: Set output_components for lowered clip distance outputs
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Nanley Chery [Thu, 27 Aug 2015 23:05:22 +0000 (16:05 -0700)]
mesa/teximage: restrict GL_ETC1_RGB8_OES support to GLES
According to the extensions table and our glext headers,
OES_compressed_ETC1_RGB8_texture is only supported in
GLES1 and GLES2. Since we may give users a GLES3 context
when a GLES2 context is requested, we also allow this
extension for GLES3 as well.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Signed-off-by: Nanley Chery <nanley.g.chery@intel.com>
Nanley Chery [Thu, 10 Sep 2015 17:48:46 +0000 (10:48 -0700)]
mesa/extensions: restrict GL_OES_EGL_image to GLES
Driver vendors do this as well. The extension specification
lists GLES 1.1 or 2.0 as requirements.
Reviewed-by: Chad Versace <chad.versace@intel.com>
Signed-off-by: Nanley Chery <nanley.g.chery@intel.com>
Nanley Chery [Thu, 27 Aug 2015 23:05:22 +0000 (16:05 -0700)]
mesa/extensions: restrict luminance alpha formats to API_OPENGL_COMPAT
According the GL 3.1 spec, luminance alpha formats are deprecated.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Signed-off-by: Nanley Chery <nanley.g.chery@intel.com>
Thomas Hellstrom [Tue, 15 Sep 2015 06:40:07 +0000 (23:40 -0700)]
gallium/svga: Enable PIPE_FORMAT_L8_UNORM for vgpu10
It's extensively used by XA for a8- and planar yuv component surfaces.
This fixes broken XA yuv blits using vgpu10 contexts.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Emil Velikov [Thu, 10 Sep 2015 13:41:38 +0000 (14:41 +0100)]
egl/dri2: don't leak the fd on dri2_terminate
Currently the check was incorrect as it did not consider the (unlikely)
case of fd == 0. In order to fix this we should first correctly
initialize it to -1, as the swrast implementations leave it set to zero
(props to calloc()).
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Boyan Ding <boyan.j.ding@gmail.com>
Emil Velikov [Mon, 7 Sep 2015 08:53:53 +0000 (09:53 +0100)]
egl/dri2/drm: compact existing device mgmt
Move the fcntl(dupfd_cloexec) to the else branch where it belongs.
Otherwise it's not immediately obvious that the code is hit, only when
an existing device is used.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Boyan Ding <boyan.j.ding@gmail.com>
Matt Turner [Wed, 15 Jul 2015 16:00:41 +0000 (09:00 -0700)]
egl/dri2: Close file descriptor on error.
v2: [Emil Velikov]
Rework the error path to a common goto, close only if we own the fd.
v3; [Emil Velikov]
Always close the fd (we either opened the device or dup'd) (Boyan, Ian)
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Boyan Ding <boyan.j.ding@gmail.com>
Ray Strode [Fri, 28 Aug 2015 18:50:21 +0000 (14:50 -0400)]
gbm: convert gbm bo format to fourcc format on dma-buf import
At the moment if a gbm buffer is imported and the gbm buffer
has an old-style GBM_BO_FORMAT format, the import will crash,
since it's passed directly to DRI functions that expect
a fourcc format (as provided by the newer GBM_FORMAT
definitions)
This commit addresses the problem in two ways:
1) it prevents invalid formats from leading to a crash by
returning EINVAL if the image couldn't be created
2) it translates GBM_BO_FORMAT formats into the comparable
GBM_FORMAT formats.
Reference: https://bugzilla.gnome.org/show_bug.cgi?id=753531
CC: "10.6 11.0" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Alejandro Piñeiro [Mon, 14 Sep 2015 18:16:25 +0000 (20:16 +0200)]
docs: document INTEL_DEBUG 'optimizer' envvar
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kristian Høgsberg Kristensen [Sat, 5 Sep 2015 00:09:40 +0000 (17:09 -0700)]
i965: Move perf_debug code to brw_codegen_*_prog()
We're trying to avoid a libdrm dependency in the core compiler, so let's
move the perf_debug code one level up from the brw_*_emit() helpers to
the brw_codegen_*_prog() helpers.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Kristian Høgsberg Kristensen <krh@bitplanet.net>
Kristian Høgsberg Kristensen [Fri, 4 Sep 2015 23:55:03 +0000 (16:55 -0700)]
i965: Move brw_fs_precompile() to brw_wm.c
All other precompile functions live in the brw_<stage>.c files, make fs
follow the convention.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Kristian Høgsberg Kristensen <krh@bitplanet.net>
Kristian Høgsberg Kristensen [Fri, 4 Sep 2015 23:35:34 +0000 (16:35 -0700)]
i965: Move compute shader code around
This moves the compute shader code around in order to make the way the
code is split up more consistent. There should be no functional changes.
Typically we have a few files per stage:
brw_vs.c, brw_wm.c brw_gs.c:
code to drive code generation and implement precompiling and
cache search.
genX_<stage>_state.c
gen specific implementation of the state emission for the shader
stage.
The brw_*_emit() functions are all in the same files as the visitor
classes they use (with the exception of VS, which may use either vec4 or
fs).
To make compute follow this convention, we move the brw_cs_emit()
function into brw_fs.cpp. We can then rename brw_cs.cpp to brw_cs.c and
do this in C like the other similar files. Finally, move state setup
and atoms to gen7_cs_state.c.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Kristian Høgsberg Kristensen <krh@bitplanet.net>
Anuj Phogat [Fri, 24 Jul 2015 22:53:58 +0000 (15:53 -0700)]
meta: Abort meta pbo path if TexSubImage need signed unsigned conversion
See similar fix for Readpixels in mesa commit
0d20790. Jason suggested
we need that for TexSubImage as well.
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Ilia Mirkin [Fri, 11 Sep 2015 03:58:17 +0000 (23:58 -0400)]
nvc0/ir: start offset at texBindBase for txq, like regular texturing
Curiously this has no actual effect. I think it's because the first 8
textures are bound in multiple slots for some reason. However seems
prudent to use these the same way as regular texturing, esp in the case
where there are more than 8 textures bound.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Eric Anholt [Mon, 14 Sep 2015 15:21:07 +0000 (11:21 -0400)]
vc4: Fix build from recent NIR cleanups.
Antia Puentes [Mon, 14 Sep 2015 07:50:59 +0000 (09:50 +0200)]
i965/vec4_nir: Load constants as integers
Loads constants using integer as their register type, like it is
done in FS backend.
No shader-db changes in HSW.
Cc: "10.6 11.0" <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91716
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Antia Puentes [Wed, 5 Aug 2015 13:57:33 +0000 (15:57 +0200)]
i965/vec4: Fix saturation errors when coalescing registers
If the register types do not match and the instruction
that contains the final destination is saturated, register
coalescing generated non-equivalent code.
This did not happen when using IR because types usually
matched, but it is visible in nir-vec4.
For example,
mov vgrf7:D vgrf2:D
mov.sat m4:F vgrf7:F
is coalesced to:
mov.sat m4:D vgrf2:D
The patch prevents coalescing in such scenario, unless the
instruction we want to coalesce into is a MOV (without type
conversion implied). In that case, the patch sets the register
types to the type of the final destination.
Shader-db results in HSW (only vec4 instructions shown):
total instructions in shared programs:
1754415 ->
1754416 (0.00%)
instructions in affected programs: 74 -> 75 (1.35%)
helped: 0
HURT: 1
GAINED: 0
LOST: 0
Only one extra instruction in one of the shaders, that comes from
eliminating a saturation error by preventing register coalesce.
Cc: "10.6 11.0" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Tapani Pälli [Mon, 14 Sep 2015 05:50:51 +0000 (08:50 +0300)]
docs: cleanups + mark some work as done
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Mon, 14 Sep 2015 05:07:05 +0000 (01:07 -0400)]
docs: only astc ldr required for ES3.2, not hdr
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Ilia Mirkin [Fri, 11 Sep 2015 01:44:45 +0000 (21:44 -0400)]
st/mesa: emit TXQS, support ARB_shader_texture_image_samples
The image component of the ext is a no-op since there is no image support
in gallium (yet).
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Ilia Mirkin [Fri, 11 Sep 2015 02:33:34 +0000 (22:33 -0400)]
r600g: add support for TXQS tgsi opcode
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Glenn Kennard <glenn.kennard@gmail.com>
Ilia Mirkin [Fri, 11 Sep 2015 02:07:27 +0000 (22:07 -0400)]
nv50/ir: add support for TXQS tgsi opcode
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Fri, 11 Sep 2015 21:29:49 +0000 (17:29 -0400)]
gallium: add PIPE_CAP_TGSI_TXQS to let st know if TXQS is supported
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Glenn Kennard <glenn.kennard@gmail.com>
Ilia Mirkin [Fri, 11 Sep 2015 01:37:23 +0000 (21:37 -0400)]
tgsi: add a TXQS opcode to retrieve the number of texture samples
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Jordan Justen [Mon, 17 Aug 2015 23:32:42 +0000 (16:32 -0700)]
glsl/cs: Initialize gl_LocalInvocationIndex in main()
We initialize gl_LocalInvocationIndex based on the extension spec
formula:
gl_LocalInvocationIndex =
gl_LocalInvocationID.z * gl_WorkGroupSize.x * gl_WorkGroupSize.y +
gl_LocalInvocationID.y * gl_WorkGroupSize.x +
gl_LocalInvocationID.x;
https://www.opengl.org/registry/specs/ARB/compute_shader.txt
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Jordan Justen [Mon, 17 Aug 2015 22:49:44 +0000 (15:49 -0700)]
glsl/cs: Exclude gl_LocalInvocationIndex from builtin variable stripping
We lower gl_LocalInvocationIndex based on the extension spec formula:
gl_LocalInvocationIndex =
gl_LocalInvocationID.z * gl_WorkGroupSize.x * gl_WorkGroupSize.y +
gl_LocalInvocationID.y * gl_WorkGroupSize.x +
gl_LocalInvocationID.x;
https://www.opengl.org/registry/specs/ARB/compute_shader.txt
We need to set this variable in main(), even if gl_LocalInvocationIndex
is not referenced by the shader. (It may be used by a linked shader.)
Therefore, we can't eliminate it as a dead variable.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Jordan Justen [Mon, 17 Aug 2015 21:35:44 +0000 (14:35 -0700)]
glsl/cs: Initialize gl_GlobalInvocationID in main()
We initialize gl_GlobalInvocationID based on the extension spec
formula:
gl_GlobalInvocationID =
gl_WorkGroupID * gl_WorkGroupSize + gl_LocalInvocationID
https://www.opengl.org/registry/specs/ARB/compute_shader.txt
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Jordan Justen [Mon, 17 Aug 2015 19:22:34 +0000 (12:22 -0700)]
glsl: Move link_get_main_function_signature to a common location
Also rename to _mesa_get_main_function_signature.
We will call it near the end of compilation to insert some code into
main for initializing some compute shader global variables.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jordan Justen [Mon, 17 Aug 2015 19:30:25 +0000 (12:30 -0700)]
glsl/cs: Don't strip gl_GlobalInvocationID and dependencies
We lower gl_GlobalInvocationID based on the extension spec formula:
gl_GlobalInvocationID =
gl_WorkGroupID * gl_WorkGroupSize + gl_LocalInvocationID
https://www.opengl.org/registry/specs/ARB/compute_shader.txt
We need to set this variable in main(), even if gl_GlobalInvocationID
is not referenced by the shader. (It may be used by a linked shader.)
Therefore, we can't eliminate these as dead variables.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Jordan Justen [Fri, 13 Mar 2015 18:39:53 +0000 (11:39 -0700)]
i965/nir: Support gl_WorkGroupID variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 10 Oct 2014 15:28:24 +0000 (08:28 -0700)]
i965/cs: Initialize gl_WorkGroupID variable from payload
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Mar 2015 18:37:03 +0000 (11:37 -0700)]
nir: Add gl_WorkGroupID system variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 10 Oct 2014 15:28:24 +0000 (08:28 -0700)]
glsl/cs: Add gl_WorkGroupID variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 13 Mar 2015 18:34:48 +0000 (11:34 -0700)]
i965/nir: Support gl_LocalInvocationID variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 22 Nov 2014 03:14:41 +0000 (19:14 -0800)]
i965/cs: Initialize gl_LocalInvocationID from payload
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 10 Oct 2014 15:33:23 +0000 (08:33 -0700)]
i965/cs: Initialize gl_LocalInvocationID in push constant data
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Sat, 22 Nov 2014 02:47:49 +0000 (18:47 -0800)]
i965/cs: Reserve local invocation id in payload regs
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Jordan Justen [Fri, 13 Mar 2015 18:32:43 +0000 (11:32 -0700)]
nir: Add gl_LocalInvocationID variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Jordan Justen [Fri, 10 Oct 2014 15:28:24 +0000 (08:28 -0700)]
glsl/cs: Add gl_LocalInvocationID variable
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Krzesimir Nowak [Sat, 12 Sep 2015 14:17:00 +0000 (08:17 -0600)]
softpipe: Change faces type to uint
This is to avoid needless float<->int conversions, since all
face-related computations are made on integers. Spotted by Emil
Velikov.
Reviewed-by: Brian Paul <brianp@vmware.com>
Rob Clark [Sun, 13 Sep 2015 15:22:51 +0000 (11:22 -0400)]
freedreno/ir3: fix compile warn after
1807a08e
New enum to add to switch so compiler doesn't complain.
commit
1807a08e4f35b014f2a80d1e88dd74a9f096d7a5
Author: Ilia Mirkin <imirkin@alum.mit.edu>
AuthorDate: Thu Aug 27 23:05:03 2015 -0400
Commit: Ilia Mirkin <imirkin@alum.mit.edu>
CommitDate: Thu Sep 10 17:38:33 2015 -0400
nir: add nir_texop_texture_samples and convert from glsl
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Sun, 13 Sep 2015 15:21:28 +0000 (11:21 -0400)]
freedreno/ir3: fix compile break after
a4aa25be
Following commit dropped the unused memctx arg:
commit
a4aa25be1e0a27b1a6a6b0bcf576beb9dfe1ea7a
Author: Jason Ekstrand <jason.ekstrand@intel.com>
AuthorDate: Wed Sep 9 13:24:35 2015 -0700
Commit: Jason Ekstrand <jason.ekstrand@intel.com>
CommitDate: Fri Sep 11 09:21:20 2015 -0700
nir: Remove the mem_ctx parameter from ssa_def_rewrite_uses
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Thu, 10 Sep 2015 20:06:05 +0000 (16:06 -0400)]
nir: add nir_channel() to get at single components of vec's
Rather than make yet another copy of channel(), let's move it into nir.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Rob Clark [Wed, 9 Sep 2015 22:28:55 +0000 (18:28 -0400)]
tgsi/scan: add support to figure out max nesting depth
Sometimes a useful thing for compilers (or, for example, tgsi_to_nir) to
know. And pretty trivial for scan to figure this out for us.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
Kai Wasserbäch [Sat, 12 Sep 2015 08:39:50 +0000 (10:39 +0200)]
r600: Fix llvm build since const buffer changes
In commit
f9caabe8f1bff86d19b53d9ecba5c72b238d9e23:
One place in r600_llvm.c was forgotten when replacing
R600_UCP_CONST_BUFFER with R600_BUFFER_INFO_CONST_BUFFER.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91985
Signed-off-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Signed-off-by: Dave Airlie <airlied@gmail.com>
Jason Ekstrand [Thu, 10 Sep 2015 23:19:42 +0000 (16:19 -0700)]
i965/vec4: Don't reswizzle hardware registers
Cc: "11.0 10.6" <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91719
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 10 Sep 2015 23:19:22 +0000 (16:19 -0700)]
i965/emit: Add assertions for accumulator restrictions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Emil Velikov [Sat, 12 Sep 2015 12:50:33 +0000 (13:50 +0100)]
docs: add news item and link release notes for 11.0.0
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Emil Velikov [Sat, 12 Sep 2015 12:32:56 +0000 (13:32 +0100)]
docs: add sha256 checksums for 11.0.0
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
(cherry picked from commit
c4bae5792bb5515da42e23f166f5ba5d68f79615)
Emil Velikov [Sat, 12 Sep 2015 09:33:49 +0000 (10:33 +0100)]
docs: Update 11.0.0 release notes
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
(cherry picked from commit
4f1e500150be2e82a2d7eb954f7198cc0c5cbec1)
Glenn Kennard [Fri, 11 Sep 2015 10:42:23 +0000 (12:42 +0200)]
r600: Enable fp64 on chips with native support
Cypress/Cayman/Aruba, earlier r6xx/r7xx chips only support a subset
of the needed fp64 ops, and don't do GL4 anyway.
Signed-off-by: Glenn Kennard <glenn.kennard@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Glenn Kennard [Fri, 11 Sep 2015 10:42:22 +0000 (12:42 +0200)]
r600g: Support I2D/U2D/D2I/D2U
Only for Cypress/Cayman/Aruba, older chips have only partial fp64 support.
Uses float intermediate values so only accurate for int24 range, which
matches what the blob does.
Signed-off-by: Glenn Kennard <glenn.kennard@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Fri, 11 Sep 2015 03:43:53 +0000 (04:43 +0100)]
r600g: lower number of driver const buffers
I'm going to want a driver constant buffer for tess to coordinate
LDS storage, so before I go tackling that I decided to merge the
clip/samplepos and texture info buffers into one. So I can steal
the spare one.
This creates a single constant buffer between the two, with
clip/samplepos taking up a reserved 128 bytes at the start.
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Glenn Kennard <glenn.kennard@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Fri, 11 Sep 2015 02:11:43 +0000 (03:11 +0100)]
r600: define some values for the fetch constant offsets.
This just puts these in one place and #defines them.
Reviewed-by: Glenn Kennard <glenn.kennard@gmail.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Thomas Helland [Wed, 12 Aug 2015 13:07:57 +0000 (15:07 +0200)]
docs: Update with GLES3.2 entries and status
V2: -Change to "not started" for most entries
-Add status for multisample_2d_array
-Change shader_multisample_interpolation to "not_stared"
V3 (idr): Move the GLES 3.2 section after the "Additional functions"
section from GLES 3.1. Note that GL_KHR_texture_compression_astc_hdr is
done for i965 on gen9+ hardware. Note that GL_OES_shader_io_blocks is
based on some features from GLSL 1.50.
Signed-off-by: Thomas Helland <thomashelland90@gmail.com>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com> [v2]
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:42 +0000 (20:07 +0200)]
softpipe: Constify variables
This commit makes a lot of variables constant - this is basically done
by moving the computation to variable definition. Some of them are
moved into lower scopes (like in img_filter_2d_ewa).
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:41 +0000 (20:07 +0200)]
softpipe: Constify sp_tgsi_sampler
Add a small inline function doing the casting - this is to make sure
we don't do a cast from some completely unrelated type. This commit
does not make tgsi_sampler parameters const in vfuncs themselves for
now - probably llvmpipe would need looking at before making such a
change.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:40 +0000 (20:07 +0200)]
softpipe: Constify sampler and view parameters in mip filters
Those functions actually could always take them as constants.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:39 +0000 (20:07 +0200)]
softpipe: Constify sampler and view parameters in img filters
Those functions actually could always take them as constants.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:38 +0000 (20:07 +0200)]
tgsi, softpipe: Constify tgsi_sampler in query_lod vfunc
A followup from previous commit - since all functions called by
query_lod take pointers to const sp_sampler_view and const sp_sampler,
which are taken from tgsi_sampler subclass, we can the tgsi_sampler as
const itself now.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:37 +0000 (20:07 +0200)]
softpipe: Constify some sampler and view parameters
This is to prepare for making tgsi_sampler parameter in query_lod a
const too. These functions do not modify anything in either sampler or
view anymore.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Krzesimir Nowak [Fri, 11 Sep 2015 18:07:36 +0000 (20:07 +0200)]
softpipe: Move the faces array from view to filter_args
With that, sp_sampler_view instances are not abused anymore as a local
storage, so we can later make them constant.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Jason Ekstrand [Wed, 9 Sep 2015 23:03:10 +0000 (16:03 -0700)]
nir/from_ssa: Use instr_rewrite_dest
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 22:58:25 +0000 (15:58 -0700)]
nir: Add a function for rewriting instruction destinations
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Jason Ekstrand [Wed, 9 Sep 2015 22:58:08 +0000 (15:58 -0700)]
nir: Only unlink sources that are actually valid
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Jason Ekstrand [Wed, 9 Sep 2015 20:24:35 +0000 (13:24 -0700)]
nir: Remove the mem_ctx parameter from ssa_def_rewrite_uses
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Jason Ekstrand [Wed, 9 Sep 2015 20:18:29 +0000 (13:18 -0700)]
nir: Fix a bunch of ralloc parenting errors
As of
a10d4937, we would really like things associated with an instruction
to be allocated out of that instruction and not out of the shader. In
particular, you should be passing the instruction that will ultimately be
holding the source into nir_src_copy rather than an arbitrary memory
context.
We also change the prototypes of nir_dest_copy and nir_alu_src/dest_copy to
explicitly take an instruction so we catch this earlier in the future.
Cc: "11.0" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Jason Ekstrand [Thu, 10 Sep 2015 20:56:08 +0000 (13:56 -0700)]
nir/lower_outputs_to_temporaries: Reparent the output name
We copy the output, make the old output the temporary, and give the
temporary a new name. The copy keeps the pointer to the old name. This
works just fine up until the point where we lower things to SSA and delete
the old variable and, with it, the name. Instead, we should re-parent to
the copy.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Alejandro Piñeiro [Fri, 11 Sep 2015 10:21:13 +0000 (12:21 +0200)]
i965/vec4: check writemask when bailing out at register coalesce
opt_register_coalesce stopped to check previous instructions to
coalesce with if somebody else was writing on the same
destination. This can be optimized to check if somebody else was
writing to the same channels of the same destination using the
writemask.
Shader DB results (taking into account only vec4):
total instructions in shared programs:
1781593 ->
1734957 (-2.62%)
instructions in affected programs:
1238390 ->
1191754 (-3.77%)
helped: 12782
HURT: 0
GAINED: 0
LOST: 0
v2: removed some parenthesis, fixed indentation, as suggested by
Matt Turner
v3: added brackets, for consistency, as suggested by Eduardo Lima
Reviewed-by: Matt Turner <mattst88@gmail.com>
Brian Paul [Thu, 10 Sep 2015 18:26:18 +0000 (12:26 -0600)]
tgsi,softpipe: capitalize the tgsi_sampler_control enum values
We use capitalized enum values everywhere else.
This improves understanding a bit too.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Kenneth Graunke [Tue, 18 Aug 2015 09:07:47 +0000 (02:07 -0700)]
nir: Store some geometry shader data in nir_shader.
This makes it possible for NIR shaders to know the number of output
vertices and the number of invocations. Drivers could also access
these directly without going through gl_program.
We should probably add InputType and OutputType here too, but currently
those are stored as GL_* enums, and I wanted to avoid using those in
NIR, as I suspect Vulkan/SPIR-V will use different enums. (We should
probably make our own.)
We could add VerticesIn, but it's easily computable from the input
topology, so I'm not sure whether it's worth it. It's also currently
not stored in gl_shader (only gl_shader_program), which would require
changes to the glsl_to_nir interface or require us to store it there.
This is a bit of duplication of data...ideally, we would factor these
substructs out of gl_program, gl_shader_program, and nir_shader, creating
a gl_geometry_info class...but it would need to go in a new place (in
src/glsl?) that isn't mtypes.h nor nir.h.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Kenneth Graunke [Tue, 12 May 2015 08:53:24 +0000 (01:53 -0700)]
nir/builder: Add nir_load_var() and nir_store_var() helpers.
These provide a convenient way to do simple variable loads and stores.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Kenneth Graunke [Wed, 12 Nov 2014 07:16:13 +0000 (23:16 -0800)]
glsl: Use hash tables for opt_constant_propagation() kill sets.
Cuts compile/link time of the fragment shader in #91857 by 19%
(16.28 -> 13.05).
I didn't bother with the acp sets because they're smaller, but it
might be worth doing as well.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91857
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
Tested-by: Tapani Pälli <tapani.palli@intel.com>
Kenneth Graunke [Sat, 5 Sep 2015 07:51:33 +0000 (00:51 -0700)]
i965: Use hash tables for brw_fs_vector_splitting().
Cuts compile/link time of the fragment shader in #91857 by 25%
(21.64 -> 16.28).
v2: Drop unnecessary _mesa_hash_table_destroy call, and use
refs.ht->entries == 0 rather than ad-hoc checking (suggested by
Timothy Arceri).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91857
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
Tested-by: Tapani Pälli <tapani.palli@intel.com>
Kenneth Graunke [Sat, 5 Sep 2015 07:22:57 +0000 (00:22 -0700)]
glsl: Use hash tables in opt_constant_variable().
Cuts compile/link time of the fragment shader in bug #91857 by 31%
(31.79 -> 21.64). It has over 8,000 variables so linked lists are
terrible.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91857
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
Tested-by: Tapani Pälli <tapani.palli@intel.com>