Commit Graph

27 Commits

Author SHA1 Message Date
Bill Hollings
40141ffddf MSL: Selectively enable fast-math in MSL code to match Vulkan CTS results.
Based on CTS testing, math optimizations between MSL and Vulkan are inconsistent.
In some cases, enabling MSL's fast-math compilation option matches Vulkan's math
results. In other cases, disabling it does. Broadly enabling or disabling fast-math
across all shaders results in some CTS test failures either way.

To fix this, selectively enable/disable fast-math optimizations in the MSL code,
using metal::fast and metal::precise function namespaces, where supported, and
the [[clang::optnone]] function attribute otherwise.

Adjust SPIRV-Cross unit test reference shaders to accommodate these changes.
2021-09-22 18:58:31 -04:00
Hans-Kristian Arntzen
7277c7ac46 Use to_unpacked_row_major_expression to unify row-major in MSL/GLSL. 2019-07-23 11:36:54 +02:00
Hans-Kristian Arntzen
47a18b9f1b Simplify row-major matrix/vector multiplies. 2019-07-23 10:56:57 +02:00
Hans-Kristian Arntzen
19f5cd3e90 Declare correct matrix type when unpacking. 2019-07-22 13:25:45 +02:00
Hans-Kristian Arntzen
6c1f97b4a9 Fix unpacking of packed but not remapped types on load. 2019-07-19 14:50:35 +02:00
Hans-Kristian Arntzen
e90d816cdd Deal with scalar layout of entire structs.
Mark all candidate struct types.
2019-07-19 14:18:14 +02:00
Hans-Kristian Arntzen
e2c95bdcbc MSL: Rewrite how resource indices are fallback-assigned.
We used to use the Binding decoration for this, but this method is
hopelessly broken. If no explicit MSL resource remapping exists, we
remap automatically in a manner which should always "just work".
2019-06-21 12:54:08 +02:00
Chip Davis
27af716c3a MSL: Emit F{Min,Max,Clamp} as fast:: and N{Min,Max,Clamp} as precise::.
This roughly matches their semantics in SPIR-V and MSL. For `FMin`,
`FMax`, and `FClamp`, and the Metal functions `fast::min()`,
`fast::max()`, and `fast::clamp()`, the result is undefined if any
operand is NaN. For the 'N' operations and their corresponding MSL
`precise::` functions, the result is consistent with IEEE 754 (first
non-NaN wins; result is NaN if all operands are NaN).

We can only do this with 32-bit floats, though, because Metal only
provides these variants for `float`. `half` only has one variant of
these functions that is presumably consistent with IEEE 754. I guess
that's OK; the SPIR-V spec only says that `F{Min,Max,Clamp}` are
undefined for NaNs. Performance might suffer, though.
2018-09-01 23:01:46 -05:00
Bill Hollings
ab2ea93e35 Merge branch 'master' of https://github.com/KhronosGroup/SPIRV-Cross 2018-06-12 11:42:56 -04:00
Bill Hollings
9b4defe202 CompilerMSL support matrices & arrays in stage-in & stage-out.
Support flattening StorageOutput & StorageInput matrices and arrays.
No longer move matrix & array inputs to separate buffer.
Add separate SPIRFunction::fixup_statements_in & SPIRFunction::fixup_statements_out
instead of just  SPIRFunction::fixup_statements.
Emit SPIRFunction::fixup_statements at beginning of functions.
CompilerMSL track vars_needing_early_declaration.
Pass global output variables as variables to functions that access them.
Sort input structs by location, same as output structs.
Emit struct declarations in order output, input, uniforms.
Regenerate reference shaders to new formats defined by above.
2018-06-12 11:41:35 -04:00
Hans-Kristian Arntzen
192a882df3 Also unpack regular unary/binary operations on MSL.
Apparently MSL gets confused when you have packed_float3 op float3 ...
2018-06-11 16:23:09 +02:00
Hans-Kristian Arntzen
f65120c147 Deal with packed expressions in more scenarios.
Make a new "to_extract_component_expression" helper.
2018-05-25 10:57:02 +02:00
Hans-Kristian Arntzen
ac81a0ce68 Use declared binding in SPIR-V as a fallback for explicit MSL binds. 2018-04-04 12:25:11 +02:00
Hans-Kristian Arntzen
b46910e9f5 Access chain into packed vectors as arrays.
Cleaner and should be more compatible with storing to packed vector
elements.

Fix CompositeExtract bug with packed vectors on MSL.
2018-03-13 12:13:33 +01:00
Bill Hollings
6371d9e43a CompilerMSL emit no-warning pragma when emitting spvConvertFromRowMajorCxR functions. 2018-01-06 00:51:25 -05:00
Bill Hollings
5ee6b46087 Fixes from review of PR #373.
Code fixes from review.
Refactor MSL tests back to using the SPIRV-Tools
and glslang loaded by checkout_glslang_spirv_tools.sh.
2018-01-05 23:22:36 -05:00
Bill Hollings
d8d2da9d8d CompilerMSL allow swizzle of packed_float3 vectors by unpacking to float3.
Pass packed indicator back through OpAccessChain and OpLoad.
Unpack packed vector before applying swizzle.
Add packed swizzle test.
2018-01-05 17:46:56 -05:00
Bill Hollings
1c18078811 Enhancements to MSL compute and entry point naming.
Support Workgroup (threadgroup) variables.
Mark if SPIRConstant is used as an array length, since it cannot be specialized.
Resolve specialized array length constants.
Support passing an array to MSL function.
Support emitting GLSL array assignments in MSL via an array copy function.
Support for memory and control barriers.
Struct packing enhancements, including packing nested structs.
Enhancements to replacing illegal MSL variable and function names.
Add Compiler::get_entry_point_name_map() function to retrieve entry point renamings.
Remove CompilerGLSL::clean_func_name() as obsolete.
Fixes to types in bitcast MSL functions.
Add Variant::get_id() member function.
Add CompilerMSL::Options::msl_version option.
Add numerous MSL compute tests.
2017-11-05 21:34:42 -05:00
Bill Hollings
f591bc0d4a CompilerMSL enhancements.
Support BuiltInFragDepth.
Emit interface block for StorageClassUniformConstant.
Throw exception when output or fragment input structs contain matrix or array.
Dynamically created interface structs sorted by location number instead of alphabetically.
Add Compiler::is_array() function.
2017-06-30 19:10:46 -04:00
Hans-Kristian Arntzen
de33d89074 Add explicit in/out locations everywhere.
Needed for newer glslang. With Vulkan semantics for SPIR-V, all
locations must be explicitly defined.
2017-06-21 09:39:08 +02:00
Bill Hollings
192bdc9516 CompilerMSL elide unused builtins from entry function input and output structs.
Add Compiler::has_active_builtin() function.
Update test reference shaders that included unused builtins.
2017-05-24 09:31:38 -04:00
Hans-Kristian Arntzen
91379fb0d0 Implement more sophisticated check for point_size. 2017-05-23 10:15:22 +02:00
Bill Hollings
f9f87ca391 CompilerMSL output [[point_size]] attribute for BuiltInPointSize member by default.
CompilerMSL::Options::is_rendering_points defaults to true.
2017-05-05 16:13:55 -04:00
Bill Hollings
5550c87b1f CompilerMSL options access and UBO alignment test case.
CompilerMSL accesses options using same design pattern as CompilerGLSL and CompilerHLSL.
CompilerMSL support setting VA & rez binding specs via either constructor or compile() method overload.
CompilerMSL support single UBO packing and padding in single pass.
spriv_cross app (main.cpp) supports turning off UBO packing and padding via command line option.
Add MSL UBO alignment test shader.
2017-03-12 17:42:51 -04:00
Bill Hollings
de72b3328e Removed Travis-killing MSL test: flatten/copy.flatten.vert. 2017-01-31 12:01:13 -05:00
Bill Hollings
d1737c65cf Removed Travis-killing MSL test: flatten/dynamic.flatten.vert. 2017-01-31 11:53:20 -05:00
Bill Hollings
be4cb17a14 Enhance MSL testing and add numerous MSL test cases.
Add to suite of MSL tests and references any existing GLSL tests
that successfully convert GLSL->SPIRV->MSL and compile as MSL.
test_shaders_helper() ignores hidden files that start with '.',
to avoid accidentally finding hidden OSX files such as .DS_Store.
Use xcrun to compile MSL shaders instead of hard-coded path to Metal compiler.
Wrap calls to xcrun in exception handling to ignore if Xcode not installed.
For MSL tests, move call to validate_shader_msl() to after call to
regression_check() to allow a converted MSL shader to be saved for
manual review even if it doesn't successfully compile as MSL.
2017-01-30 22:55:21 -05:00