Commit graph

8755 commits

Author SHA1 Message Date
Karol Herbst
d6a48ff402 vtn/opencl: always lower to libclc fmod
The nir/spirv variant is simply not precise enough and almost everybody
lowers it anyway.

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25837>
2023-10-27 10:52:54 +00:00
Yonggang Luo
ce5475366e compiler,vulkan,drm-shim: Remove unused include directories from meson.build
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24462>
2023-10-27 01:35:10 +00:00
Yonggang Luo
73b639ec5c nir: #include "util/macros.h" for BITFIELD64_MASK in nir.c
There is no neeed #include "main/menums.h" in nir.c,
as it's belongs to gallium code

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24462>
2023-10-27 01:35:10 +00:00
Faith Ekstrand
49b3118302 nir/lower_bit_size: Use b2b for boolean subgroup ops
Without this, we replace vote_ieq(b) with vote_ieq(u2u32(b)) which is
wonky because we're doing a u2u on a 1-bit type. With this, we now
replace it with vote_ieq(b2b32(b)).  For other subgroup ops, we replace
things like *scan[op](b) with *scan[op](b2b32(b)).  For scan ops, this
assumes that b2b1(op(b1b32(x), b2b32(y))) = op(x, y) for all of the ops
iand, ior, and ixor.  This is true on all the back-ends I'm aware of.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
5014759133 nir: Return b2b ops from nir_type_conversion_op()
Without this, nir_type_conversion_op(bool, bool32, RND) will return
u2u32 instead of b2b32 which is pretty unexpected behavior.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
d5c310899a nir: Split nir_lower_subgroup_options::lower_vote_eq into two bits
On NVIDIA, we can do a vote_ieq on bool in one hardware op so we don't
want that lowered.  We do want to lower vote_feq and other vote_ieq,
though.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
f10d768a88 nir/lower_bit_size: Use u_intN_min/max()
May as well clean it up while we're here.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
5465e5b157 nir/lower_bit_size: Handle vote_feq/ieq separately
They're different enough from all the other subgroup ops so it's best to
handle them as their own case.

Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
5979e74177 nir/lower_bit_size: Fix subgroup lowering for floats
Using u2u is always correct for integers, including signed integers,
because we're doing a down-cast.  It's wrong for floats, though.

Fixes: f95665cfeb ("nir/lower_bit_size: Add support for lowering subgroup ops")
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
16664b74a2 nir: Add a lower_read_first_invocation option to lower_subgroups
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Faith Ekstrand
3d027cca1e nir: Add a lower_first_invocation_to_ballot option to lower_subgroups
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25894>
2023-10-26 23:05:44 +00:00
Ian Romanick
18d8a96a00 nir/split_vars: Don't split arrays of cooperative matrix types
glsl_type_is_vector_or_scalar would more accruately be called "can be an
r-value that isn't an array, structure, or matrix. This optimization
pass really shouldn't do anything to cooperative matrices. These
matrices will eventually be lowered to something else (dependent on the
backend), and that thing may (or may not) be handled by this or another
pass.

Fixes: 2d0f4f2c17 ("compiler/types: Add support for Cooperative Matrix types")
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25871>
2023-10-26 18:45:29 +00:00
Jani Nikula
eabd93bba8 docs/nir: use hawkmoth instead of doxygen
Use the hawkmoth c:auto* directives to incorporate nir documentation.

Convert @param style parameter descriptions to rst info field lists.
Add static stubs for generated headers. Fix a lot of references, in
particular the symbols are now in the Sphinx C domain, not C++
domain. Tweak syntax here and there.

Based on the earlier work by Erik Faye-Lund <kusmabite@gmail.com>

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24507>
2023-10-26 16:13:26 +00:00
Jani Nikula
d407cd8216 nir: drop **< style documentation comments
Prepare for using Hawkmoth.

Hawkmoth does not support trailing comments using /**< ... */
syntax. Replace with regular documentation comments.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24507>
2023-10-26 16:13:25 +00:00
Jani Nikula
4024d0c196 nir: add names to some typedef'd structs/enums
In order to document "typedef struct { ... } T;" as a struct in
hawkmoth, the structs need to have names. Similar for enums.

Note: This is no longer required with Hawkmoth 0.16.0+ and Clang 16 and
later. With the next Hawkmoth release, this should be fixed also for
Clang 15 and earlier.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24507>
2023-10-26 16:13:25 +00:00
Yonggang Luo
ffa458ee8f nir: remove redundant include of gallium headers
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25880>
2023-10-26 09:35:04 +00:00
Yonggang Luo
43715516fc treewide: Merge num_mesh_vertices_per_primitive and u_vertices_per_prim into mesa_vertices_per_prim
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25880>
2023-10-26 09:35:04 +00:00
Yonggang Luo
be431e0dc7 compiler: Implement num_mesh_vertices_per_primitive to match u_vertices_per_prim
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25880>
2023-10-26 09:35:04 +00:00
Rhys Perry
1afd0878e9 nir/lower_shader_calls: skip zero-sized qsort
Fixes UBSan:
src/compiler/nir/nir_lower_shader_calls.c:1681:7: runtime error: null pointer passed as argument 1, which is declared to never be null

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25853>
2023-10-25 17:27:47 +00:00
Rhys Perry
f9289dfd02 nir/serialize: fix signed integer overflow
Fixes UBSan error:
src/compiler/nir/nir_serialize.c:1277:70: runtime error: left shift of 524287 by 13 places cannot be represented in type 'int'

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25853>
2023-10-25 17:27:47 +00:00
Alyssa Rosenzweig
9a6c20e64f nir/trivialize_registers: Handle obscure load hazard
Somebody less tired than me would add a unit test for this. Offending snippet:

        32    %58 = @load_reg (%55) (base=0, legacy_fabs=0, legacy_fneg=0)
        32    %57 = @load_reg (%55) (base=0, legacy_fabs=0, legacy_fneg=0)
        32    %21 = iadd %57, %15 (0x1)
                    @store_reg (%21, %55) (base=0, wrmask=x, legacy_fsat=0)
        32    %56 = @load_reg (%55) (base=0, legacy_fabs=0, legacy_fneg=0)
        32    %22 = i2f32 %56
        32    %23 = load_const (0x41000000 = 8.000000)
        32    %24 = fdiv %22, %23 (8.000000)
        32    %90 = mov %24
                    @store_reg_indirect (%90, %78, %58) (base=0, wrmask=x, legacy_fsat=0)

Closes: #10031
Fixes: d313eba94e ("nir: Add pass for trivializing register access")
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reported-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25865>
2023-10-25 16:34:47 +00:00
Caio Oliveira
67450674c0 compiler/types: Move comments and reorganize declarations
Move comments from C++ member functions to the C functions.  In
some cases just delete comments or consolidate them together.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
dfcca13800 compiler/types: Remove warnings about potential fallthrough
None of those cases are expected to fallthrough, but should be unreachable.
Just break them so they get to the unreachable entry at the end.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
9e32cc3d0b compiler/types: Rename glsl_types.cpp to glsl_types.c
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
c45276c665 compiler/types: Annotate extern "C" only once in glsl_types.cpp
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
cecdc686e4 compiler/types: Remove usages of C++ members in glsl_types.cpp
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
94bdf351dc compiler/types: Use C instead of C++ constants for builtin types
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
07ee4bd69f compiler/types: Add remaining type extraction functions and use them in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
ada6183d60 compiler/types: Add glsl_simple_explicit_type() and simplify glsl_simple_type()
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
e17adf51db compiler/types: Implement glsl_type::field_type() in terms of existing functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
9e514b89a0 compiler/types: Add glsl_get_explicit_*() functions and use them in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
d2a804a25b compiler/types: Add glsl_get_std430_array_stride() and use it in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
7b42fe62a1 compiler/types: Add glsl_type_uniform_locations() and use it in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
e98ba3b53f compiler/types: Add glsl_type_compare_no_precision() and use it in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
3ce4d5e033 compiler/types: Add glsl_get_mul_type() and use it in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
795bf4244c compiler/types: Add more glsl_contains_*() functions and use them in C++
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
68f80e6fc1 compiler/types: Move remaining code from nir_types to glsl_types
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
b2407d7859 compiler/types: Flip wrapping of numeric type conversion functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
8bebd40d5c compiler/types: Flip wrapping of remaining small data getters
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
922fe24739 compiler/types: Flip wrapping of remaining non-trivial type getters
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
a5e6e5b6d3 compiler/types: Flip wrapping of get row/column type helpers
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
beaac525e8 compiler/types: Flip wrapping of various get instance functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
5c91cf9288 compiler/types: Flip wrapping of texture/sampler/image get instance functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
87b3812f10 compiler/types: Flip wrapping of get_instance()
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
2117158619 compiler/types: Flip wrapping of record_compare
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
e486384540 compiler/types: Flip wrapping of layout related functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
a4cfeea850 compiler/types: Flip wrapping of interface related functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
ad092fcab5 compiler/types: Flip wrapping of struct related functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
421a04f5ba compiler/types: Flip wrapping of size related functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00
Caio Oliveira
418e3be14c compiler/types: Flip wrapping of CL related functions
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25129>
2023-10-25 01:51:12 +00:00