Commit Graph

878 Commits

Author SHA1 Message Date
dwang102
070863af69 Add SPV_AMD_shader_early_and_late_fragment_tests 2022-07-18 14:20:18 +08:00
Greg Fischer
c1ae2f33b5 Do not generate samplerBuffer for spirv1.6 and beyond
This type was removed from spirv1.6. If samplerBuffer is specified in
GLSL, generate textureBuffer. If samplerBuffer type is constructed,
just return the buffer.

Fixes #2956
2022-06-21 17:42:53 -06:00
James0124
ea024d2bfc CInterface: Add OpSource support.
Add interface for `TIntermediate::addSourceText` and `TIntermediate::setSourceFile`.
2022-06-07 01:13:21 +09:00
Greg Fischer
9e2b914722 Restore legacy interface for remap()
Fixes ABI breakage caused by #2933
2022-06-01 16:40:29 -06:00
Andrea Faulds
6cdae46314 Avoid duplicate BuiltIn variables for ray tracing matrices (fix #2921)
Fixes an issue where invalid SPIR-V was generated when
gl_ObjectToWorldEXT and gl_ObjectToWorld3x4EXT, or
gl_WorldToObjectEXT and gl_WorldToObject3x4EXT, were used in the same
shader. The SPIR-V specification requires that there be at most one
OpVariable decorated with a given BuiltIn value.
2022-06-01 10:43:13 +02:00
Greg Fischer
316f12ac1d
Merge pull request #2950 from qingyuanzNV/insert_opline_before_opfunction2
Generate OpLine Before OpFunction
2022-05-31 13:25:22 -06:00
Qingyuan Zheng
61d244145d avoid using make_unique for c++11 compatibility 2022-05-31 10:40:25 -07:00
Qingyuan Zheng
b6df89b470 use unique_ptr to avoid calling deleted move ctor 2022-05-30 23:08:50 -07:00
stusmith
ebf45697be Add support for VK_EXT_fragment_shader_barycentric 2022-05-25 13:02:02 +01:00
Qingyuan Zheng
279c28e70a generate OpLine before OpFunction 2022-05-23 23:05:43 -07:00
David Neto
ea7e64cf50 Remove unused variable 2022-05-19 13:27:12 -04:00
Greg Fischer
2f5bc0b741
Merge pull request #2933 from AaronHaganAMD/whitelist
Add whitelist filtering for debug comments in SPIRV-Remap.
2022-05-11 13:45:33 -06:00
ahagan
b5aae62731 Add whitelist filtering for debug comments in SPIRV-Remap. 2022-05-11 12:08:25 -04:00
Marius Bjorge
3015d00ee0 Adding support for GL_EXT_ray_cull_mask 2022-05-05 12:56:04 +02:00
Greg Fischer
b7968b7dbc Generate Int8, Int16 and Float16 capabilities for constants
Fixes #2905
2022-03-25 14:36:01 -06:00
Greg Fischer
f25ca1ec5e
Merge pull request #2912 from greg-lunarg/edic1
Add eliminate-dead-input-components to -Os
2022-03-24 13:43:40 -06:00
Greg Fischer
abbdf63cbe Add eliminate-dead-input-components to -Os 2022-03-24 13:22:44 -06:00
Biswapriyo Nath
e80f2f8760
cmake: Install SPVRemapper runtime library
This installs SPVRemapper.dll file which was missing previously.
2022-03-15 19:26:57 +05:30
Greg Fischer
69565b1884 Fix size_t to int warning 2022-01-27 11:51:04 -07:00
Greg Fischer
9ebd8ff6c1 Add Vulkan 1.3 support
Also update known goods to Vulkan 1.3 support
Also re-enable SPIR-V 1.6 tests with vulkan1.3 target
Also re-cache SPIRV 1.6 header which somehow regressed back to 1.5
2022-01-26 17:57:31 -07:00
Jeremy Hayes
294c8f1d51 Fix GCC sign-compare warnings 2022-01-04 19:55:41 -07:00
Greg Fischer
df609a01b3
Merge pull request #2800 from mhillenbrand/string_literal_endianness
Fix encoding/decoding of string literals for big-endian systems
2022-01-03 17:57:38 -07:00
Greg Fischer
7a49192d23 Add support for spirv1.6
Add command line support which enables the following behavior:

- Remap discard

  Map discard to DemoteToHelperInvocation for HLSL shaders. Map to
  OpTerminateInvocation for GLSL shaders.

- Decorate HelperInvocation with Volatile

- Use localSizeId for execution mode

  WorkGroupSize is deprecated in spirv1.6

Also update known goods to SPIRV 1.6
2021-12-16 15:01:42 -07:00
Greg Fischer
a0f98ad401 Pickup header for SPIR-V 1.6 2021-12-16 13:44:54 -07:00
Ashwin Lele
e306f02927 Do not output location decoration for certain variables in ray tracing storage
classes.
2021-12-06 13:09:03 -08:00
Rex Xu
fe54126c05 Fix an issue of spirv_by_reference
When using this qualifier for a parameter, we make it as a pointer.
However, the function TranslateStorageClass() is therefore called and
the storage class should only be set to Function when it is invoked
to translate parameter types rather than actual argument types.
2021-11-23 16:40:45 +08:00
Rex Xu
3e9e60af86 Remove the test spv.intrinsicsSpecConst
According to current SPIR-V spec, we couldn't write a valid case that
references OpExecutionModeId and OpDecorateId.
2021-11-22 21:33:36 +08:00
Rex Xu
bee91eb479 Fix an issue of spirv_type used in local variable definitions
We previously use createOp() in SPV builder to create type declaration.
However, all type declarations should be placed in const-type-variable
declaration section. And duplicated type defintions ought to be avoided.
We now make a method in SPV builder to perform this operation with a
more general solution: makeGenericType().
2021-11-18 20:17:04 +08:00
Greg Fischer
77415296b2 Update to latest spirv-headers 2021-11-11 08:43:30 -07:00
Marius Hillenbrand
78ce7e567f Fix encoding/decoding of string literals for big-endian systems
Per SPIR-V spec, a string literal's UTF-8 octets are encoded packed into
words with little-endian convention. Explicitly perform that encoding
instead of assuming that the host system is little-endian.

Note that this change requires corresponding fixes in SPIRV-Tools.

Fixes #202
2021-11-08 14:12:01 +01:00
Greg Fischer
38fd9221db
Merge pull request #2777 from amdrexu/bugfix
Make modifications of GL_EXT_spirv_intrinsics
2021-10-21 12:59:22 -06:00
Greg Fischer
82b2668d58 Allow 8/16-bit integer as array index
Also enable 8/16 bit int capability in SPIR-V in such cases.

Also enable 64 bit capabilities when used in operations.

Fixes #2779
2021-10-20 15:15:30 -06:00
Rex Xu
07aec25f82 Make modifications of GL_EXT_spirv_intrinsics
1. spirv_execution_mode_id and spirv_decorate_id could support
   specialization constants. The original implementation always
   assume only normal frontend constants are valid. It is not true.

2. spirv_type donesn't support type_specifier as an option of
   spirv_type_parameter. At present, only constant_expression is the
   valid option.
2021-10-15 20:22:36 +08:00
Chris Forbes
dd097e5adb Untangle use of core glslang version enums in SpvPostProcess
This is the only thing requiring GlslangToSpv.h to be included here.

Change-Id: I72e9e278aa1e6d717e83f9dbf11e089aafe2eb0e
2021-10-14 08:43:34 -07:00
Nathaniel Cesario
c8ef4f8a9f cmake: Remove "conditions" from endif
See https://cmake.org/cmake/help/latest/command/endif.html and
https://cmake.org/cmake/help/latest/command/if.html.

If the else/endif condition does not match the if condition verbatim,
an error is produced on some versions of cmake. This change removes
these "legacy conditions."
2021-09-30 11:00:09 -06:00
Greg Fischer
fb0e4983e4
Merge pull request #2767 from greg-lunarg/i2763
Fix unreachable code in getSampledType()
2021-09-23 10:41:34 -06:00
Greg Fischer
3d03b7822e
Scalarize vector readFirstInvocationARB (#2766)
Fixes #2761
2021-09-23 10:40:43 -06:00
Greg Fischer
4ea41b650d Fix unreachable code in getSampledType()
Fixed #2763
2021-09-22 15:16:30 -06:00
Jeremy Hayes
012436d680 Fix GCC warnings
Fix -Wsign-compare warnings.
Fix -Wunused-parameter warnings.
2021-08-31 16:40:35 -06:00
alelenv
da8027a9df
Merge branch 'KhronosGroup:master' into rt_motion 2021-08-10 12:02:09 -07:00
alelenv
fc60f77aa2 Add support for GL_NV_ray_tracing_motion_blur. 2021-08-10 10:40:28 -07:00
Greg Fischer
9f18258e41 Fix seperate stores to swizzled lvalue.
Seen with += and ++ ops on swizzled lvalues. Was using stale access chain.

Fixes 2725.
2021-08-09 17:58:57 -06:00
Greg Fischer
d433cccb8a
Merge pull request #2702 from jeremy-lunarg/hayes-2518
Generate separate stores for partially swizzled memory stores
2021-07-28 16:49:27 -06:00
Jeremy Hayes
6d5b40f051 Generate separate stores for partially swizzled memory stores
Full vector and fully specified vector swizzle stores are not affected by this change, only partial swizzles ie swizzles with fewer components than the vector being stored to.

Previously the vector being stored to loaded and any components not specified in the swizzle were used to create a full store to the vector.

While this change generates more SPIR-V instructions, it is necessary for correctness.

Fixes #2518.
2021-07-28 14:55:45 -06:00
Nathan Reed
fc9897d1ba
Fix OpImageRead result type when compiling HLSL (#2706)
Fix OpImageRead result type when compiling HLSL

- Per the Vulkan spec, OpImageRead must return a 4-component vector always. When compiling HLSL, loads from a RWTexture of a template type with < 4 components would incorrectly generate an OpImageRead with a < 4 component result, resulting in validation errors.
- This was previously fixed for OpImageFetch in commit 4425f24; this commit does the same thing for OpImageRead.
- Added associated tests and expanded existing image fetch tests to check all the different types of textures, in both float and int incarnations, for completeness.
- Update other HLSL tests involving OpImageRead
2021-07-28 09:29:17 -06:00
Peter Kasting
e8cffa5b7f Fix some instances of -Wunused-but-set-variable.
Bug: chromium:1203071
2021-07-26 07:34:37 -07:00
Shahbaz Youssefi
097215f618 Fix loading bool arrays from interface blocks
SPIR-V disallows bool in interface blocks, which is emulated with uint.
When loading a bool variable (through accessChainLoad()), it's converted
from uint to bool if it came from an interface block.

This was handled for bool and bvecN, but not for bool arrays.  This
change implements the conversion for bool arrays.

Closes #2694
2021-07-15 22:29:48 -04:00
Greg Fischer
b9d06bdb01
Merge pull request #2668 from cmarcelo/GLSL_EXT_shader_atomic_float2
Implement GLSL_EXT_shader_atomic_float2
2021-06-17 11:27:47 -06:00
Greg Fischer
71612a7e5d
Merge pull request #2625 from amdrexu/feature
Implement the extension GL_EXT_spirv_intrinsics
2021-06-16 14:52:35 -06:00
Jason Ekstrand
d352577a99 Implement GLSL_EXT_shader_atomic_float2 2021-06-16 09:11:14 -07:00