fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2025-12-24 04:30:10 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1
100044 commits 64 branches 914 tags 1 GiB
Commit graph

465 commits

Author SHA1 Message Date
Samuel Pitoiset
cd6dfacda9 ac/nir: drop nir_to_llvm_context from visit_load_push_constant() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:25 +01:00
Samuel Pitoiset
5c9e398c83 ac/nir: drop nir_to_llvm_context from cast_ptr() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:23 +01:00
Samuel Pitoiset
5ef5944848 ac/nir: drop nir_to_llvm_context from visit_load_local_invocation_index() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:21 +01:00
Samuel Pitoiset
da8b0b8264 ac/nir: drop nir_to_llvm_context from emit_f2f16() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:19 +01:00
Samuel Pitoiset
e32f374944 ac: remove unused parameters in abi::load_tess_coord() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:17 +01:00
Samuel Pitoiset
1e69db003d ac/nir: remove useless bitcast in load_tess_coord() 
nir_intrinsic_load_tess_coord always returns a v3i32.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:15 +01:00
Samuel Pitoiset
ed179fbdf3 ac: add load_resource() to the ABI 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:13 +01:00
Samuel Pitoiset
ecf229706f ac: add load_sample_mask_in() to the ABI 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:11 +01:00
Samuel Pitoiset
0f48eeea05 ac: move view_index to the ABI 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:09 +01:00
Samuel Pitoiset
0efbede949 ac: move push_constants to the ABI 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:07 +01:00
Samuel Pitoiset
460d3ce726 ac: move tg_size to the ABI 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:04 +01:00
Samuel Pitoiset
054c92190c ac/nir: remove unused nir_to_llvm_context:{defs,phis} 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-12 11:54:02 +01:00
Timothy Arceri
ef8082baf8 ac: convert nir_op_f2f32 src to a float 
Fixes the following piglit test:

./bin/arb_vertex_attrib_64bit-check-explicit-location -auto -fbo

Where we would end up with the nir such as:

	vec1 64 ssa_11 = pack_64_2x32_split ssa_9, ssa_10
	vec1 32 ssa_12 = f2f32 ssa_2

And our pack_64_2x32_split nir to llvm code always produces
a 64bit integer as output.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-02-10 10:46:28 +11:00
Timothy Arceri
1b1e5f8edf ac: fix some 64bit unpack asserts 
Previously the asserts did not take swizzles into account.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-02-10 10:46:28 +11:00
Samuel Pitoiset
3a2bb4db23 ac/nir: compute correct number of user SGPRs on GFX9 
For merged shaders.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
 2018-02-09 10:16:04 +01:00
Timothy Arceri
c77078c942 ac: pass struct ac_llvm_context to emit_membar() 
Fixes segfault in piglit test:

./bin/arb_shader_image_load_store-shader-mem-barrier --quick -auto -fbo

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-09 12:51:27 +11:00
Timothy Arceri
a9f6b392c7 ac: move get_elem_bits() to ac_llvm_build.c 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-09 09:42:59 +11:00
Timothy Arceri
19f9839f0b ac: add 64bit bitCount support 
v2: use LLVMBuildTrunc()

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-09 09:42:59 +11:00
Samuel Pitoiset
bb750d265c ac/nir: clean up handle_fs_outputs_post() 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:33 +01:00
Samuel Pitoiset
528bc14fa5 ac/nir: add radv_load_output() helper 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:30 +01:00
Samuel Pitoiset
834d9845ca ac/shader: scan info about output PS declarations 
NIR->LLVM should only be a translation pass, and all scan stuff
should be done before.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:27 +01:00
Samuel Pitoiset
a8e04e91de ac/nir: add radv_export_param() helper 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:26 +01:00
Samuel Pitoiset
e3cfd6b805 ac/nir: remove set but unused export_mask 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:24 +01:00
Samuel Pitoiset
724136d590 ac/nir: remove dead code in handle_vs_outputs_post() 
The memcpy can't be reached because the condition is always false.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:22 +01:00
Samuel Pitoiset
c63d8d0284 ac/nir: remove useless check in si_llvm_init_export_args() 
values can't be NULL because we use ac_build_export_null() now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:14:20 +01:00
Samuel Pitoiset
26ab5a4269 ac/nir: use ac_build_export_null() 
The number of enabled channels should be 0 when exporting null.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-08 22:11:44 +01:00
Samuel Pitoiset
757d36ee70 ac/nir: use new pknorm_i16/u16 and pk_i16/u16 LLVM intrinsics 
Ported from RadeonSI.

Only one F1 2017 shader is affected, code size decreased
from 532 to 488 on both Polaris10 and Vega10.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-07 12:42:13 +01:00
Samuel Pitoiset
2f54d7382d ac/nir: avoid loading unused VS input components 
Polaris10:
Totals from affected shaders:
SGPRS: 122840 -> 120984 (-1.51 %)
VGPRS: 78812 -> 78440 (-0.47 %)
Spilled SGPRs: 177 -> 129 (-27.12 %)
Code Size: 2950028 -> 2941276 (-0.30 %) bytes
Max Waves: 17899 -> 17976 (0.43 %)

Vega10:
Totals from affected shaders:
SGPRS: 117144 -> 115776 (-1.17 %)
VGPRS: 77580 -> 77532 (-0.06 %)
Spilled SGPRs: 0 -> 152 (0.00 %)
Code Size: 3352656 -> 3347860 (-0.14 %) bytes
Max Waves: 19756 -> 19866 (0.56 %)

This increases SGPRs spilling a bit with Talos, but I have
some other ideas that might reduce it.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-07 12:42:09 +01:00
Timothy Arceri
9c52902c76 ac/radeonsi: add num_work_groups to the abi 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Timothy Arceri
f12e2f9c12 ac: implement nir_intrinsic_shader_clock 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Timothy Arceri
d116af383f ac/radeonsi: add load_local_group_size() to the abi 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Timothy Arceri
e3ebffdbb0 ac: don't call emit_outputs() for compute 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Timothy Arceri
c8066cdfa7 ac/radeonsi: add local_invocation_ids to the abi 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Timothy Arceri
fa5239c153 ac/radeonsi: add workgroup_ids to the abi 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-02-07 08:43:08 +11:00
Bas Nieuwenhuizen
c7d640fbbf ac/nir: fix GS load input type. 
Fixes: df1d5174fc "ac/nir: replace SI.buffer.load.dword with amdgcn.buffer.load"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-02-06 21:52:38 +01:00
Samuel Pitoiset
0170ae1e23 ac/nir: remove emission of nir_op_fdiv 
RadeonSI and RADV lower fdiv.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-05 23:09:34 +01:00
Samuel Pitoiset
a1d568c830 ac/nir: fix a crash in load_gs_input() on pre-GFX9 chips 
Fixes: df1d5174fc ("ac/nir: replace SI.buffer.load.dword with amdgcn.buffer.load")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-05 11:05:52 +01:00
Marek Olšák
3bf1e036e8 amd: remove support for LLVM 3.9 
Only these are supported:
- LLVM 4.0
- LLVM 5.0
- LLVM 6.0
- master (7.0)

Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-02 23:47:40 +01:00
Samuel Pitoiset
df1d5174fc ac/nir: replace SI.buffer.load.dword with amdgcn.buffer.load 
The old one generates useless instructions in there, found while
comparing geometry shaders between RadeonSI and RADV.

This improves all Vulkan demos that use geometry shaders, +4%
for deferredshadows, +9% for viewportarray, +7% for
geometryshader on Polaris10.

This seems to also improve DOW3 a little bit (+1%).

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by:  Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-02-02 12:32:21 +01:00
Marek Olšák
b0a6053a99 ac/nir: use ac_build_buffer_load_format for image buffer loads 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-02-01 16:20:19 +01:00
Marek Olšák
bac9fa9f17 ac: add glc parameter to ac_build_buffer_load_format 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-02-01 16:20:19 +01:00
Samuel Pitoiset
26cc3e74b9 ac/nir: fix emission of ffract for 64-bit 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
 2018-01-31 14:10:24 +01:00
Timothy Arceri
d185190222 ac/radeonsi: add lookup_interp_param and load_sample_position to the abi 
This will enable the interpolateAt builtins to work on the radeonsi
nir backend.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-01-31 09:14:07 +11:00
Timothy Arceri
97058168a4 radeonsi/nir: add prim_mask to the abi 
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-01-31 09:14:07 +11:00
Timothy Arceri
b8808848ce ac/nir_to_llvm: move some interp defines to the header 
These will be used in the following patch.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
 2018-01-31 09:14:07 +11:00
Timothy Arceri
9afc38c799 ac: fix indentation 
Reviewed-by: Dave Airlie <airlied@redhat.com>
 2018-01-29 11:14:23 +11:00
Timothy Arceri
03086f86ae ac: remove unused nir2llvmtype() 
The last use of this was removed in the previous patch.

Reviewed-by: Dave Airlie <airlied@redhat.com>
 2018-01-29 11:14:23 +11:00
Timothy Arceri
fa29a9625e ac: fix gs load inputs type 
This fixes the scenario where the input is a struct. With this
the Unreal engines Elemental demo now works on radeonsi.

Reviewed-by: Dave Airlie <airlied@redhat.com>
 2018-01-29 11:14:23 +11:00
Kai Wasserbäch
0aba967328 ac/nir: call glsl_get_sampler_dim() only once where possible 
Changes since v1:
  * Rebased on top of e68150de26 and
    82adf53308.

Signed-off-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
 2018-01-29 10:47:31 +11:00
Marek Olšák
b633999a4e ac: rename and move si_const_array into common code 
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
 2018-01-27 02:09:09 +01:00