(and GL_NV_shader_subgroup_partitioned) based on GL/ES version
instead of predicating it on vulkan SPV generation
Also add AST testing.
The glsl.450.subgroup* files are largely the same as the spv.subgroup*
The glsl.es320.subgroup* files are the same as the 450 versions, but modified to be ES compatible.
Prior to this change, OES_primitive_bounding_box and EXT_primitive_bounding_box were both recognised as extensions, but only the name gl_BoundingBoxOES could be used. However the EXT version uses the name gl_BoundingBoxEXT instead. In addition, since GLES 3.2, the extension has been included in the core standard and the name gl_BoundingBox may be used instead. This change aims to make both extensions and the 3.2 core version all work.
This is an alternate fix for the issue described in commit be63facd, whose
solution didn't work if there were non-trivial operations involved in computing
a constant initializer which caused the 'constant unfolding' code to kick in
(addConstantReferenceConversion). Instead, this change does the 'unfolding'
later in createSpvConstantFromConstUnionArray. If a reference-type constant has
survived that long, then folding is already done, this must be a 'real' (inside
a function) use of the constant, and it should be safe to unfold and apply the
bitcast.
Last year we changed 'volatile' to also act as 'coherent', but when I
resolved the memory model changes against that change I missed handling
volatile in a couple places that we check for coherent. There was also
a place in post-processing that acted as if the volatile memory access
flag has a literal number associated with it, when it doesn't.
Allow constructors to and from references to be constant folded. Section 4.3.3
says constructors whose arguments are all constant expressions must fold.
Disallow 'const' on buffer reference types. It is not a 'non-void transparent
basic data type' (it is not considered 'basic').
Handle buffer reference constants (which can be assigned to a non-const reference,
or can be further folded to another type of constant) by converting to
'constructor(uint64_t constant)' in addConversion.
Disallow == and != operators on reference types.
- This change also allows redeclaration of gl_PrimitiveIndicesNV and
adds error checks against incorrect explicit array size.
- Also modifies gtests to check array bound limits and redeclare gl_PrimitiveIndicesNV[].
* This comes from the resolution of issues 4, 5 & 6 in
ARB_program_interface_query, stating that uniform buffers should have their
members expanded out as normal and arrays should have elements added.
* If a buffer block has a large array e.g. [10000] we don't want to iterate over
every array element. Instead we should only expand out the first [0] element,
then expand as normal from there.
* The array name should still be appended with [0] to indicate that it's an
array.
This commit undoes the code on the commit "ParseHelper: assign
xfb_offset for struct members too" (commit
af8c1bdb161e4b2ddf019b61bdfca61f84cb1ca8), except the method renaming,
and updates the tests.
During the review of the implementation of Vulkan extension
VK_EXT_transform_feedback for the Linux Intel driver, it was pointed
that the Vulkan environment spec for SPIR-V only ever refers to block
or block members being decorated. It would be strange to not do the
same for OpenGL (ARB_gl_spirv). That would also fit better to what
GLSL does, where setting explicit xfb offsets for struct members is
not allowed.
FWIW, the original patch was proposed based on the fact that
ARB_gl_spirv is relying on OpenGL for how xfb offsets should be
assigned to members, and it was not clear (at least to me) which is
the responsible of such.
* Previously the offset was only updated with EOpIndexDirectStruct, but we also
need to update it for the other index types into arrays, and when expanding an
aggregate reached after dereferencing.
