Marek Olšák [Tue, 1 Sep 2020 22:05:52 +0000 (18:05 -0400)]
radeonsi: stop using TGSI_PROPERTY_TES_VERTEX_ORDER_CW
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Marek Olšák [Tue, 1 Sep 2020 22:04:57 +0000 (18:04 -0400)]
radeonsi: stop using TGSI_PROPERTY_TES_SPACING
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Marek Olšák [Tue, 1 Sep 2020 21:54:47 +0000 (17:54 -0400)]
radeonsi: stop using TGSI_PROPERTY_TES_POINT_MODE / TES_PRIM_MODE
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Marek Olšák [Tue, 1 Sep 2020 21:53:10 +0000 (17:53 -0400)]
radeonsi: stop using TGSI_PROPERTY_TCS_VERTICES_OUT
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Marek Olšák [Tue, 1 Sep 2020 21:50:59 +0000 (17:50 -0400)]
radeonsi: stop using TGSI_PROPERTY_VS_WINDOW_SPACE_POSITION / VS_BLIT_SGPRS_AMD
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Marek Olšák [Tue, 1 Sep 2020 21:48:44 +0000 (17:48 -0400)]
radeonsi: stop using TGSI_PROPERTY_NEXT_SHADER
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6624>
Andreas Baierl [Wed, 2 Sep 2020 09:27:10 +0000 (11:27 +0200)]
lima/ppir: Skip instruction merge when having more than one successor
ppir_do_one_node_to_instr merges instructions and uses a pipeline reg
to save a reg. It tests if ppir_node_has_single_src_succ, but it should
check if ppir_node_has_single_succ.
The following deqp tests run into this issue because they have a node
with 2 successors in different blocks, where one was merged into the same
instruction and the second one is pointing to a missing predecessor then.
Fixes the following deqp tests:
dEQP-GLES2.functional.shaders.loops.do_while_dynamic_iterations.vector_counter_fragment
dEQP-GLES2.functional.shaders.loops.for_dynamic_iterations.vector_counter_fragment
dEQP-GLES2.functional.shaders.loops.while_dynamic_iterations.vector_counter_fragment
Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6555>
jzielins [Wed, 2 Sep 2020 17:55:29 +0000 (19:55 +0200)]
swr: Fix crashes on non-AVX hardware
Compilers may use vector instructions in calculating
hash values of std::string. This happens usualy when
high optimalization level is enabled. SWR had two
static std::map<std::string, T> variables which
lead to crashes on non-AVX systems during the initialization
of those variables. This commit makes those variables
dynamically allocated and fixes this AVX instruction
leak.
Closes: #3077
Closes: #198
Reviewed-by: Krzysztof Raszkowski <krzysztof.raszkowski@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6565>
Pierre-Eric Pelloux-Prayer [Mon, 7 Sep 2020 07:58:36 +0000 (09:58 +0200)]
amd/common: switch to 3-spaces style
Follow-up of !4319 using the same clang-format config.
Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5310>
Pierre-Eric Pelloux-Prayer [Mon, 7 Sep 2020 07:56:01 +0000 (09:56 +0200)]
amd/llvm: switch to 3-spaces style
Follow-up of !4319 using the same clang-format config.
Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5310>
Dave Airlie [Tue, 1 Sep 2020 23:08:52 +0000 (09:08 +1000)]
vulkan/device_select: don't pick a cpu driver as the default
This should stop apps using vallium by accident, when there
are better options.
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6552>
Mauro Rossi [Sat, 5 Sep 2020 21:35:28 +0000 (23:35 +0200)]
android: nv50/ir: Add nv50_ir_prog_info_out serialize and deserialize
Fixes the following building errors:
ld.lld: error: undefined symbol: nv50_ir_prog_info_serialize
>>> referenced by nvc0_program.c:647 (external/mesa/src/gallium/drivers/nouveau/nvc0/nvc0_program.c:647)
>>> nvc0_program.o:(nvc0_program_translate) in archive out/target/product/x86_64/obj/STATIC_LIBRARIES/libmesa_pipe_nouveau_intermediates/libmesa_pipe_nouveau.a
ld.lld: error: undefined symbol: nv50_ir_prog_info_out_deserialize
>>> referenced by nvc0_program.c:656 (external/mesa/src/gallium/drivers/nouveau/nvc0/nvc0_program.c:656)
>>> nvc0_program.o:(nvc0_program_translate) in archive out/target/product/x86_64/obj/STATIC_LIBRARIES/libmesa_pipe_nouveau_intermediates/libmesa_pipe_nouveau.a
ld.lld: error: undefined symbol: nv50_ir_prog_info_out_serialize
>>> referenced by nvc0_program.c:674 (external/mesa/src/gallium/drivers/nouveau/nvc0/nvc0_program.c:674)
>>> nvc0_program.o:(nvc0_program_translate) in archive out/target/product/x86_64/obj/STATIC_LIBRARIES/libmesa_pipe_nouveau_intermediates/libmesa_pipe_nouveau.a
Fixes:
f20a210dc ("nv50/ir: Add nv50_ir_prog_info_out serialize and deserialize")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6627>
Tomeu Vizoso [Fri, 21 Aug 2020 08:49:51 +0000 (10:49 +0200)]
ci: Run deqp-gles3 and deqp-gles31 on RadeonSI
Only a fourth of the GLES3 tests and a fifth of the GLES31 tests for
now, as the machine is quite slow and there's a problem with dEQP
calling fesetround all the time and that being very expensive on x86_64.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6580>
Qiang Yu [Thu, 3 Sep 2020 03:30:28 +0000 (11:30 +0800)]
radeonsi: fix user fence space when MCBP is enabled
When MCBP is enabled, IB maybe preempted which will also update
the preempted fence field of the user fence. So we need to reserve
enough space for each user fence.
Fixes:
89d2dac5548 "radeonsi: enable preemption if the kernel enabled it"
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6577>
Qiang Yu [Thu, 3 Sep 2020 09:39:41 +0000 (17:39 +0800)]
radeonsi: fix syncobj wait timeout
syncobj wait takes absolute timeout value.
Fixes:
162502370c7 "winsys/amdgpu: implement sync_file import/export"
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6579>
Mauro Rossi [Sat, 5 Sep 2020 16:49:19 +0000 (18:49 +0200)]
android: util: fix missing include path
Fixes the following building error:
external/mesa/src/util/format/u_format_bptc.c:28:10:
fatal error: 'u_format_pack.h' file not found
^~~~~~~~~~~~~~~~~
1 error generated.
Fixes:
8d38b2578 ("util: Explicitly call the unpack functions from inside bptc pack/unpack.")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6619>
Mauro Rossi [Sat, 5 Sep 2020 15:59:48 +0000 (17:59 +0200)]
android: amd/registers: switch to new generated register definitions
Android building rules are aligned to meson ones
Fixes the following building error:
FAILED: ninja: 'external/mesa/src/amd/registers/amdgfxregs.json',
needed by 'out/target/product/x86_64/gen/STATIC_LIBRARIES/libmesa_amd_common_intermediates/common/sid_tables.h',
missing and no known rule to make it
Fixes:
b7a6333ee ("amd/registers: switch to new generated register definitions")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6618>
Marek Olšák [Sun, 6 Sep 2020 07:46:59 +0000 (03:46 -0400)]
ac/llvm: fix b2f for v2f16
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 3 Sep 2020 09:31:36 +0000 (05:31 -0400)]
ac/llvm: add better code for fsign
There are 2 improvements:
- better code for 16, 32, and 64 bits
- vector support for 16 and 32 bits
Totals:
SGPRS: 2639738 -> 2625882 (-0.52 %)
VGPRS: 1534120 -> 1533916 (-0.01 %)
Spilled SGPRs: 3541 -> 3557 (0.45 %)
Spilled VGPRs: 33 -> 33 (0.00 %)
Private memory VGPRs: 256 -> 256 (0.00 %)
Scratch size: 292 -> 292 (0.00 %) dwords per thread
Code Size:
55640332 ->
55384892 (-0.46 %) bytes
Max Waves: 964785 -> 964857 (0.01 %)
Totals from affected shaders:
SGPRS: 377352 -> 363496 (-3.67 %)
VGPRS: 209800 -> 209596 (-0.10 %)
Spilled SGPRs: 1979 -> 1995 (0.81 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 256 -> 256 (0.00 %)
Scratch size: 256 -> 256 (0.00 %) dwords per thread
Code Size:
12549300 ->
12293860 (-2.04 %) bytes
Max Waves: 105762 -> 105834 (0.07 %)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 3 Sep 2020 09:51:17 +0000 (05:51 -0400)]
ac/llvm: add better code for isign
There are 2 improvements:
- select v_med3_i32
- support vectors
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 23 Jul 2020 02:05:45 +0000 (22:05 -0400)]
ac/llvm: remove dead code handling for fmod
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Sat, 4 Jul 2020 20:49:52 +0000 (16:49 -0400)]
ac/llvm: fix bcsel for v2*16
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 23 Jul 2020 03:03:39 +0000 (23:03 -0400)]
ac/llvm: fix amdgcn.rsq for v2f16
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Sat, 4 Jul 2020 20:50:06 +0000 (16:50 -0400)]
ac/llvm: fix amdgcn.fract for v2f16
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Sat, 4 Jul 2020 20:17:28 +0000 (16:17 -0400)]
ac/llvm: fix amdgcn.rcp for v2f16
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 3 Sep 2020 10:18:18 +0000 (06:18 -0400)]
ac/llvm: remove stub prototype for fmed3
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Tue, 1 Sep 2020 03:37:37 +0000 (23:37 -0400)]
Revert "ac: generate FMA for inexact instructions for radeonsi"
This reverts commit
4b9370cb0f3a2d9030e827f847f66bdefeaf08fd.
Fixes:
4b9370cb0f3a2d9030e827f847f66bdefeaf08fd
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3429
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Sun, 3 May 2020 23:38:51 +0000 (19:38 -0400)]
radeonsi: move nir_shader_compiler_options into si_screen
so that they can be different depending on the GPU (for 16-bit support)
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Marek Olšák [Thu, 3 Sep 2020 08:42:38 +0000 (04:42 -0400)]
radeonsi: remove redundant no-signed-zero-fp-math LLVM attribute
Already set by AC_FLOAT_MODE_DEFAULT_OPENGL.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6284>
Andrey Vostrikov [Thu, 27 Aug 2020 06:32:48 +0000 (09:32 +0300)]
egl/x11: Free memory allocated for reply structures on error
This patch fixes memory leaks when reply is allocated and is not freed
on error execution path.
Found by enabling address sanitizer on simple EGL app.
```c
int main()
{
EGLDisplay display = eglGetDisplay(EGL_DEFAULT_DISPLAY);
EGLint major;
EGLint minor;
if (!eglInitialize(display, &major, &minor))
{
return 1;
}
eglTerminate(display);
return 0;
}
```
Compiled with: `gcc testme.c -o testme -fsanitize=address -lasan -lEGL`
Execution environment:
- Windows 10, VMWare Player 15.5.2 build-
15785246 without 3D accelaration
- Guest OS: OpenSUSE Leap 15.2
- Mesa 19.3.4
Program output:
```sh
ASAN_OPTIONS=fast_unwind_on_malloc=0 ./testme
libEGL warning: DRI2: failed to authenticate
==52510==ERROR: LeakSanitizer: detected memory leaks
Direct leak of 32 byte(s) in 1 object(s) allocated from:
#0 0x7fa62315f500 in malloc (/usr/lib64/libasan.so.4+0xdc500)
#1 0x7fa61e12d86b (/usr/lib64/libxcb.so.1+0xf86b)
#2 0x7fa61e12b5c7 (/usr/lib64/libxcb.so.1+0xd5c7)
#3 0x7fa61e12cc3e (/usr/lib64/libxcb.so.1+0xec3e)
#4 0x7fa61e12cd4f in xcb_wait_for_reply (/usr/lib64/libxcb.so.1+0xed4f)
#5 0x7fa61ebe02a5 (/usr/lib64/libEGL_mesa.so.0+0x202a5)
#6 0x7fa61ebdb5ca (/usr/lib64/libEGL_mesa.so.0+0x1b5ca)
#7 0x7fa61ebd750c (/usr/lib64/libEGL_mesa.so.0+0x1750c)
#8 0x7fa61ebd7554 (/usr/lib64/libEGL_mesa.so.0+0x17554)
#9 0x7fa61ebd1107 (/usr/lib64/libEGL_mesa.so.0+0x11107)
#10 0x400856 in main (/home/user/testme+0x400856)
#11 0x7fa622ad8349 in __libc_start_main (/lib64/libc.so.6+0x24349)
#12 0x4006e9 in _start (/home/user/testme+0x4006e9)
SUMMARY: AddressSanitizer: 32 byte(s) leaked in 1 allocation(s).
```
Signed-off-by: Andrey Vostrikov <av.linux.dev@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6611>
Marek Olšák [Thu, 3 Sep 2020 13:27:58 +0000 (09:27 -0400)]
glsl_to_nir: fix crashes with int16 shifts
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6613>
Samuel Pitoiset [Thu, 3 Sep 2020 20:02:01 +0000 (22:02 +0200)]
spirv: fix emitting switch cases that directly jump to the merge block
As shown in the valid SPIR-V below, if one switch case statement
directly jumps to the merge block, it has no branches at all and
we have to reset the fall variable. Otherwise, it creates an
unintentional fallthrough.
OpSelectionMerge %97 None
OpSwitch %96 %97 1 %99 2 %100
%100 = OpLabel
%102 = OpAccessChain %_ptr_StorageBuffer_v4float %86 %uint_0 %uint_37
%103 = OpLoad %v4float %102
%104 = OpBitcast %v4uint %103
%105 = OpCompositeExtract %uint %104 0
%106 = OpShiftLeftLogical %uint %105 %uint_1
OpBranch %97
%99 = OpLabel
OpBranch %97
%97 = OpLabel
%107 = OpPhi %uint %uint_4 %75 %uint_5 %99 %106 %100
This fixes serious corruption in Horizon Zero Dawn.
v2: Changed the code to skip the entire if-block instead of resetting
the fallthrough variable.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3460
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6590>
Jason Ekstrand [Fri, 4 Sep 2020 19:06:29 +0000 (14:06 -0500)]
compiler/types: Fix deserializing structs with >= 15 members
This was a typo in
a0b82c24b6d08c where we used the wrong struct member
to decide whether or not to read the explicit_alignment.
Fixes:
a0b82c24b6d08c "nir/glsl: Add an explicit_alignment field to glsl_type"
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3487
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6605>
Jonathan Marek [Fri, 4 Sep 2020 17:36:22 +0000 (13:36 -0400)]
freedreno/regs: add 7nm DSI PHY/PLL regs
This is for the kernel driver.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6603>
Roland Scheidegger [Wed, 2 Sep 2020 02:04:50 +0000 (04:04 +0200)]
gallivm: add InstSimplify pass
This is the recommended replacement for the removed ConstantPropagation
pass, and llvm now added c binding for it.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6585>
Marcin Ślusarz [Thu, 2 Jul 2020 16:57:08 +0000 (18:57 +0200)]
intel/compiler: fix typo in a comment
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6602>
Marcin Ślusarz [Thu, 2 Jul 2020 11:37:10 +0000 (13:37 +0200)]
intel/compiler: print dispatch width when shader fails to compile
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6602>
Marcin Ślusarz [Fri, 26 Jun 2020 18:15:35 +0000 (20:15 +0200)]
intel/compiler: move extern C functions out of namespace brw
brw_compile_gs and brw_compile_tcs are extern C functions, but are
defined inside of brw namespace, which somehow works but confuses
Eclipse CDT's code analysis.
Move these functions out of brw namespace and fix references to
objects from brw namespace.
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6602>
Marcin Ślusarz [Fri, 4 Sep 2020 16:43:35 +0000 (18:43 +0200)]
intel/compiler: use the same name for nir shaders in brw_compile_* functions
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6602>
Marcin Ślusarz [Fri, 26 Jun 2020 17:54:29 +0000 (19:54 +0200)]
intel/compiler: match brw_compile_* declarations with their definitions
Current state confuses Eclipse CDT's code analysis.
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6602>
Marek Olšák [Tue, 1 Sep 2020 06:30:13 +0000 (02:30 -0400)]
nir/algebraic: add 16-bit versions of a few 32-bit patterns
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6599>
Marek Olšák [Tue, 11 Aug 2020 22:49:37 +0000 (18:49 -0400)]
nir/algebraic: trivially enable existing 32-bit patterns for all bit sizes
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6599>
Marek Olšák [Thu, 23 Jul 2020 02:13:16 +0000 (22:13 -0400)]
nir: get ffma support from NIR options for nir_lower_flrp
This also fixes the inverted last parameter of nir_lower_flrp in most drivers.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6599>
Nanley Chery [Tue, 23 Jun 2020 17:23:12 +0000 (10:23 -0700)]
iris: Better determine map_would_stall for Z/S
Use iris_has_invalid_primary to determine if mapping a depth or stencil
surface would lead to a stall due to resolving.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6192>
Nanley Chery [Tue, 23 Jun 2020 16:46:38 +0000 (09:46 -0700)]
iris: Drop a use of the need_resolve boolean
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6192>
Nanley Chery [Wed, 24 Jun 2020 16:41:16 +0000 (09:41 -0700)]
iris: Avoid resolving Z/S reads in transfer_map
Resolves are destructive for depth and stencil surfaces, so avoid
resolving them for reads.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6192>
Nanley Chery [Tue, 23 Jun 2020 16:52:47 +0000 (09:52 -0700)]
iris: Make iris_has_color_unresolved more generic
Replace iris_has_color_unresolved with iris_has_invalid_primary.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6192>
Nanley Chery [Wed, 24 Jun 2020 14:45:49 +0000 (07:45 -0700)]
iris: Fold a condition into no_gpu for consistency
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6192>
Samuel Pitoiset [Fri, 4 Sep 2020 10:02:50 +0000 (03:02 -0700)]
aco: handle unaligned loads on GFX10.3
Same as GFX10.
Cc: 20.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6594>
Eric Anholt [Fri, 21 Aug 2020 23:59:33 +0000 (16:59 -0700)]
nir/load_store_vectorizer: Add unit tests for alignment handling.
The alignment upgrading logic is pretty tricky, but this gives me good
confidence that it actually works.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4710>
Eric Anholt [Fri, 21 Aug 2020 23:49:13 +0000 (16:49 -0700)]
nir/load_store_vectorizer: Use more imm helpers in the tests.
It's a lot more readable this way, I think.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4710>
Rhys Perry [Thu, 23 Apr 2020 15:13:10 +0000 (16:13 +0100)]
nir/load_store_vectorize: rework alignment calculation
It now also updates align_offset and creates better alignment information
with a constant 0 offset.
shader-db (Navi):
Totals from 63 (0.05% of 127638) affected shaders:
SGPRs: 3072 -> 3064 (-0.26%)
VGPRs: 2736 -> 2740 (+0.15%)
CodeSize: 325180 -> 324336 (-0.26%); split: -0.27%, +0.01%
Instrs: 63555 -> 63413 (-0.22%); split: -0.24%, +0.02%
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4710>
Rhys Perry [Mon, 24 Aug 2020 19:00:10 +0000 (20:00 +0100)]
aco: fix byte_align_scalar for 3 dword vectors
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Fixes:
fe08f0ccf94a7315bded5868b4f6a8bae744de79
('aco: add byte_align_scalar() & trim_subdword_vector() helper functions')
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4710>
Marcin Ślusarz [Wed, 2 Sep 2020 12:26:41 +0000 (14:26 +0200)]
intel/fs: add hint how to get more info when shader validation fails
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6559>
Marcin Ślusarz [Wed, 2 Sep 2020 12:25:54 +0000 (14:25 +0200)]
intel: add INTEL_DEBUG=shaders
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6559>
Marek Olšák [Fri, 4 Sep 2020 10:38:15 +0000 (06:38 -0400)]
ac/llvm: fix unaligned VS input loads on gfx10.3
Fixes:
a23802bcb9a
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6595>
Erik Faye-Lund [Thu, 3 Sep 2020 13:21:23 +0000 (15:21 +0200)]
nir: fix const-cast warning on MSVC
We're casting pointers to const memory to const pointers. MSVC complains
about this with the following warning:
warning C4090: 'initializing': different 'const' qualifiers
In this case, we can easily use both constnesses, because all we do is
read here. So let's avoid the warning by adding another const-keyword.
Fixes:
193765e26ba ("nir/lower_goto_if: Sort blocks in select_fork")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6582>
Eric Engestrom [Tue, 1 Sep 2020 20:09:43 +0000 (22:09 +0200)]
meson: drop leftover PTHREAD_SETAFFINITY_IN_NP_HEADER
55765f80 replaced this with a check for the header itself.
Fixes:
55765f80b9ce7cce4ec6 ("util/u_thread: include pthread_np.h if found")
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Jonathan Gray <jsg@jsg.id.au>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6548>
Samuel Pitoiset [Fri, 28 Aug 2020 06:41:20 +0000 (08:41 +0200)]
radv,aco: disable opts if VK_PIPELINE_CREATE_DISABLE_OPTIMIZATION_BIT
Sounds useful to determine if ACO breaks a specific pipeline
because of various optimizations.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6487>
Vinson Lee [Tue, 1 Sep 2020 23:21:47 +0000 (16:21 -0700)]
panfrost: Fix gnu-empty-initializer errors.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3473
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6551>
Nanley Chery [Wed, 17 Jul 2019 20:19:43 +0000 (13:19 -0700)]
iris: Fix aux assertion in resource_get_handle
iris_resource_get_handle currently asserts that the resource has an aux
state that is suitable for sharing. However, the caller of this function
can pass a flag to specify that it will handle flushing/resolving the
resource as needed for sharing. Take this flag into account when
asserting the state of the aux buffer.
Fixes:
e81392868e6 ("iris/resource: Drop redundant checks for aux support")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/128
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/1371>
Eric Anholt [Mon, 31 Aug 2020 20:00:14 +0000 (13:00 -0700)]
ci/bare-metal: Use python for handling fastboot booting and parsing
Modeling after what I did for cros_servo_run.py, this gives us easy
support for restarting the test run a530 when we detect a spontaneous
reboot. I had to touch up serial_buffer.py to handle buffering in from a
file instead of a serial device, to support the upcoming etnaviv CI
(tested by running it against a serial log from db410c and seeing it step
to calling "fastboot")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Eric Anholt [Mon, 31 Aug 2020 20:04:28 +0000 (13:04 -0700)]
ci/bare-metal: Fix capturing of serial output as job artifacts.
I tried to put them in the wrong directory -- everything needs to go in
results/, which we want clean and ready before we start our job.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Eric Anholt [Mon, 31 Aug 2020 20:41:57 +0000 (13:41 -0700)]
ci/bare-metal: Log why our run restarts when it does.
It would be confusing to see a job quietly restart itself in the middle.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Eric Anholt [Mon, 31 Aug 2020 19:41:57 +0000 (12:41 -0700)]
ci/bare-metal: Include a timestamp in our serial reads.
gitlab CI doesn't include timestamps in its logs by default, but it's
really useful for finding delays in our CI so stuff one in on the lines
coming in from serial and being output to the gitlab log. The artifacts
file is still the raw serial output.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Eric Anholt [Mon, 31 Aug 2020 19:08:34 +0000 (12:08 -0700)]
ci/bare-metal: Fix detection of "POWER_GOOD not seen in time" fails
We were only reading from the CPU serial, not EC, so we'd never notice
these sources of job timeouts. I couldn't find a cleaner solution, so I
spawned two threads to do the blocking reads from our serial line fifos
and merge them together in a single queue to read.
Closes: #3470
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Eric Anholt [Mon, 31 Aug 2020 20:38:15 +0000 (13:38 -0700)]
ci/bare-metal: Use re.search() instead re.match() for our line matching.
match() looks for the start of the line to match our regex, while search
just looks for the regex anywhere in the line. I messed this up when
converting our greps in shell to python, which was part of breaking the
POWER_GOOD flake detection. Most of our matches worked, but let's
consistently use this one so we don't mess this up in the future.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6529>
Tony Wasserka [Wed, 2 Sep 2020 17:15:01 +0000 (19:15 +0200)]
amd/common: Fix various non-critical integer overflows
The result of 0xf << 28 is a signed integer and hence overflows into the sign
bit. In practice compilers did the right thing here, since the intent of the
code was unsigned arithmetic anyway.
Cc: mesa-stable
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6568>
Tony Wasserka [Wed, 2 Sep 2020 16:28:36 +0000 (18:28 +0200)]
aco: Fix integer overflows when emitting parallel copies during RA
32-bit shifts were accidentally used before this change despite the intended
output being 64 bits.
This was observed when compiling Dolphin's ubershaders.
Cc: mesa-stable
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6568>
Tony Wasserka [Wed, 2 Sep 2020 16:13:57 +0000 (18:13 +0200)]
radv: Fix various non-critical integer overflows
The result of 0xf << 28 is a signed integer and hence overflows into the sign
bit. In practice compilers did the right thing here, since the intent of the
code was unsigned arithmetic anyway.
These conditions were observed in:
* dEQP-VK.pipeline.image.suballocation.sampling_type.combined.view_type.1d.format.r4g4b4a4_unorm_pack16.count_8.size.512x1
* dEQP-VK.binding_model.descriptorset_random.sets32.noarray.ubolimitlow.sbolimitlow.sampledimglow.outimgonly.noiub.nouab.frag.ialimithigh.0
Cc: mesa-stable
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6568>
Rhys Perry [Wed, 12 Aug 2020 14:58:32 +0000 (15:58 +0100)]
aco: remove omod_success/clamp_success
This simplifies the optimizer and should make SDWA optimizations easier.
No fossil-db changes on Navi.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6293>
Rhys Perry [Tue, 17 Dec 2019 14:55:24 +0000 (14:55 +0000)]
aco: fix mad splitting after applying output modifiers
Previously, this wasn't done because the mad label wasn't passed to the
new definition.
fossil-db (Navi):
Totals from 5770 (4.24% of 135946) affected shaders:
SGPRs: 391920 -> 391872 (-0.01%)
VGPRs: 349084 -> 348424 (-0.19%); split: -0.20%, +0.01%
CodeSize:
34639636 ->
34637496 (-0.01%); split: -0.02%, +0.01%
MaxWaves: 58828 -> 58862 (+0.06%)
Instrs: 6723436 -> 6723297 (-0.00%); split: -0.02%, +0.02%
Cycles:
197594168 ->
197591968 (-0.00%); split: -0.02%, +0.02%
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6293>
Rhys Perry [Thu, 3 Sep 2020 18:06:39 +0000 (19:06 +0100)]
radv: remove descriptor_indexing fails from expected fails
I think these should be fixed since
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6451. They also
pass for me on polaris and navi.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6586>
Jason Ekstrand [Fri, 28 Aug 2020 20:42:45 +0000 (15:42 -0500)]
anv: Set alignments on UBO/SSBO root derefs
This doesn't really do anything for us today. One day, I suppose we
could use it to do something with wide loads with non-uniform offsets.
The big reason to do this is to get better testing to make sure that NIR
doesn't blow up on the deref paths.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 21:00:26 +0000 (16:00 -0500)]
spirv: Drop the OpenCL type layout code
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 20:56:38 +0000 (15:56 -0500)]
clover/nir: Use lower_vars_to_explicit for uniform and global
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Tue, 1 Sep 2020 01:09:42 +0000 (20:09 -0500)]
spirv: Stop counting inputs in entry_point_wrapper
nir_shader::num_inputs isn't supposed to be a count of how many input
variables we have. It's a size of the lowered input space.
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Tue, 1 Sep 2020 15:30:17 +0000 (10:30 -0500)]
clover: Use args.size() to compute new var locations
This is better than using num_uniforms as it guarantees what we want: a
mapping from nir_variable to the args vector.
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 21:00:00 +0000 (16:00 -0500)]
nir: Allow uniform in nir_lower_vars_to_explicit_types
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 21:12:57 +0000 (16:12 -0500)]
nir: Allow var_mem_global in nir_lower_vars_to_explicit_types
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 23:34:50 +0000 (18:34 -0500)]
spirv: Propagate alignments to deref chains via casts
This commit propagates the alignment information provided either through
the Alignment decoration on pointers or via the alignment mem operands
to OpLoad, OpStore, and OpCopyMemory to the NIR deref chain. It does so
by wrapping the deref in a cast. NIR should be able to clean up most
unnecessary casts only leaving us with the useful alignment information.
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 23:33:50 +0000 (18:33 -0500)]
spirv: Add pointer helper vars to OpCopyMemory
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Mon, 24 Aug 2020 16:27:31 +0000 (11:27 -0500)]
nir/opt_deref: Remove restrictive alignment information from casts
If we have a cast deref with alignment information and we can get equal
or better alignment information from something further up the deref
chain, we can strip the alignment information from the cast and allow
other optimizations to potentially eliminate the cast.
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Mon, 24 Aug 2020 16:26:18 +0000 (11:26 -0500)]
nir/opt_deref: Don't remove casts with alignment information
Generally, if a cast has alignment information, that information is
useful and we don't want to loose it.
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Mon, 24 Aug 2020 15:57:57 +0000 (10:57 -0500)]
nir/lower_io: Apply alignments from derefs when available
If the deref has no explicit alignment in the chain, we assume component
alignment which is what we currently assume for all derefs today. This
should be correct for all APIs in the sense that we can usually assume
at least component alignment. However, for some APIs such as OpenCL, we
could potentially make larger alignment assumptions. The intention is
that those will be handled via alignment-increasing casts.
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Mon, 24 Aug 2020 15:48:51 +0000 (10:48 -0500)]
nir: Add a helper for getting the alignment of a deref
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Thu, 27 Aug 2020 16:59:54 +0000 (11:59 -0500)]
nir: Handle all array stride cases in nir_deref_instr_array_stride
This renames it to drop the ptr_as and makes it handle all of the stride
cases. There's a bit of a tricky bit in here around Booleans but we
currently use 32-bit for those always.
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Mon, 24 Aug 2020 14:51:04 +0000 (09:51 -0500)]
nir: Add alignment information to cast derefs
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Sat, 29 Aug 2020 23:01:46 +0000 (18:01 -0500)]
nir/glsl: Add an explicit_alignment field to glsl_type
When creating explicit type, the alignment information is lost, thus
forcing explicit type users to recalculate the alignment using the same
size_align() function. Let's add a new field to cache this information.
Only structs, matrices, and vectors have and explicit alignment. Arrays
alignment is implicitly set to its element alignment and matrices are
required to have an alignment that matches that of its vector columns.
the concept of alignment simply doesn't apply to other types.
We make the strategic choice to not allow explicit alignments on
scalars. This is for a couple of reasons:
1. There are no cases today where we use explicit types where we want
any other alignment for scalars than natural alignment.
2. Vectors don't have a component alignment that's separate from the
explicit_alignment so it's impossible to get an explicitly aligned
scalar type which is the component of the explicitly aligned vector
type.
This choice may cause problems if we ever want to use explicitly laid
out types for things like varyings where we sometimes want vec4
alignment of scalars. We can deal with that when the time comes.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Tue, 1 Sep 2020 21:24:19 +0000 (16:24 -0500)]
compiler/types: Make booleans 32-bit for cl_size/align
OpenCL doesn't mandate a size and this is consistent with the rest of
the glsl_type system. While we're here, we also clean ::cl_size() up a
bit and use a new explicit_type_scalar_byte_size() helper.
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Boris Brezillon [Mon, 29 Jun 2020 07:29:09 +0000 (09:29 +0200)]
nir: Expose the packed attribute attached to glsl_type objects
This should help code calculating field offsets to get it right when
the structure is marked packed.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Boris Brezillon [Tue, 30 Jun 2020 10:03:00 +0000 (12:03 +0200)]
nir/glsl: Consider block interfaces as structs when it comes to size/align calculation
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Boris Brezillon [Fri, 26 Jun 2020 08:27:16 +0000 (10:27 +0200)]
glsl: Propagate packed info in get_explicit_type_for_size_align()
Right now, when calling get_explicit_type_for_size_align() on a packed
struct, the packed attribute is lost and field offsets are wrong.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Boris Brezillon [Tue, 23 Jun 2020 10:51:49 +0000 (12:51 +0200)]
spirv: Propagate packed information to glsl_type
We need to parse the CPacked decoration early enough to apply it when
calculating field offsets and creating the struct type.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Boris Brezillon [Tue, 23 Jun 2020 10:49:23 +0000 (12:49 +0200)]
spirv: Don't accept CPacked decoration on struct members
CPacked decoration is only allowed on struct definitions, not struct
members.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Wed, 2 Sep 2020 22:43:07 +0000 (17:43 -0500)]
clover: Call nir_lower_mem_constant_vars
Fixes:
26a4c8f375e "clover/nir: Use nir_var_mem_constant for..."
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Wed, 2 Sep 2020 22:38:23 +0000 (17:38 -0500)]
nir: Don't bail too early in lower_mem_constant_vars
If there were no constant variables, we would bail out entirely.
However, we may still have constant input pointers coming in from the
client.
Fixes:
4360a8a2b3fce "nir/lower_io: Add support for nir_var_mem_constant"
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6472>
Jason Ekstrand [Tue, 14 Jul 2020 21:01:18 +0000 (16:01 -0500)]
intel/nir: Stop using nir_lower_vars_to_scratch
Instead, we do a limited indirect deref lowering and then use
nir_lower_vars_to_explicit_types and nir_lower_explicit_io to lower it
as if it were SSBO or global memory access. Among other things, this
should enable pointer arithmetic on local variables. Fun!
The only shader-db change from this change on ICL was a few tiny cycle
count changes in 7 Aztec Ruins compute shaders.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5909>
Jason Ekstrand [Tue, 14 Jul 2020 18:55:19 +0000 (13:55 -0500)]
nir/lower_indirect_derefs: Add a threshold
Instead of always lowering everything, we add a threshold such that if
the total indirected array size (AoA size) is above that threshold, it
won't lower. It's assumed that the driver will sort things out somehow
by, for instance, lowering to scratch.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5909>
Jason Ekstrand [Tue, 14 Jul 2020 20:19:59 +0000 (15:19 -0500)]
intel/compiler: Handle all indirect lowering choices in brw_nir.c
Since everything flows through NIR and we're doing all of our indirect
deref lowering there now, there's no reason to keep making those
decisions in brw_compiler and stuffing them in the GLSL compiler
structs.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5909>
projects / platform / upstream / mesa.git / log