Skip to content

Commit

Permalink
[SPIR-V] Remove dead code and correct Float16 docs (#6127)
Browse files Browse the repository at this point in the history 
The docs were out of date stating that the
`SPV_AMD_gpu_shader_half_float` extension was added when half types were
used. This is now corrected to show that the Float16 capability is added
and no extension.

The remaining code that could have added that extension in certain cases
has been dead code since #2503 since none of the instructions checked in
that list are ones that are emitted by DXC, so it is removed to clean
up.

Evidence that this is dead code:
1. `git grep "InterpolateAt"` finds no uses other than those removed
here.
2. All tests still pass when it's removed.
3. The comments in #2503 also say they were not able to find a mapping
from any HLSL intrinsics to these instructions, which is why no tests
were added.

Follow-up from: KhronosGroup/SPIRV-Tools#5519
  • Loading branch information
@sudonatalie
sudonatalie authored Jan 4, 2024
1 parent 30114a5 commit 1574879
Show file tree
Hide file tree
Showing 4 changed files with 11 additions and 37 deletions.
22 changes: 11 additions & 11 deletions docs/SPIR-V.rst
View file
Original file line number Diff line number Diff line change
Expand Up @@ -671,21 +671,21 @@ Normal scalar types
in HLSL are relatively easy to handle and can be mapped directly to SPIR-V
type instructions:

============================== ======================= ================== =========== =================================
HLSL Command Line Option SPIR-V Capability Extension
============================== ======================= ================== =========== =================================
============================== ======================= ================== ===========
HLSL Command Line Option SPIR-V Capability
============================== ======================= ================== ===========
``bool`` ``OpTypeBool``
``int``/``int32_t`` ``OpTypeInt 32 1``
``int16_t`` ``-enable-16bit-types`` ``OpTypeInt 16 1`` ``Int16``
``uint``/``dword``/``uin32_t`` ``OpTypeInt 32 0``
``uint16_t`` ``-enable-16bit-types`` ``OpTypeInt 16 0`` ``Int16``
``half`` ``OpTypeFloat 32``
``half``/``float16_t`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``SPV_AMD_gpu_shader_half_float``
``half``/``float16_t`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``Float16``
``float``/``float32_t`` ``OpTypeFloat 32``
``snorm float`` ``OpTypeFloat 32``
``unorm float`` ``OpTypeFloat 32``
``double``/``float64_t`` ``OpTypeFloat 64`` ``Float64``
============================== ======================= ================== =========== =================================
============================== ======================= ================== ===========

Please note that ``half`` is translated into 32-bit floating point numbers
if without ``-enable-16bit-types`` because MSDN says that "this data type
Expand All @@ -705,20 +705,20 @@ We use the 16-bit variants if '-enable-16bit-types' command line option is prese
For more information on these types, please refer to:
https://github.com/Microsoft/DirectXShaderCompiler/wiki/16-Bit-Scalar-Types

============== ======================= ================== ==================== ============ =================================
HLSL Command Line Option SPIR-V Decoration Capability Extension
============== ======================= ================== ==================== ============ =================================
============== ======================= ================== ==================== ============
HLSL Command Line Option SPIR-V Decoration Capability
============== ======================= ================== ==================== ============
``min16float`` ``OpTypeFloat 32`` ``RelaxedPrecision``
``min10float`` ``OpTypeFloat 32`` ``RelaxedPrecision``
``min16int`` ``OpTypeInt 32 1`` ``RelaxedPrecision``
``min12int`` ``OpTypeInt 32 1`` ``RelaxedPrecision``
``min16uint`` ``OpTypeInt 32 0`` ``RelaxedPrecision``
``min16float`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``SPV_AMD_gpu_shader_half_float``
``min10float`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``SPV_AMD_gpu_shader_half_float``
``min16float`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``Float16``
``min10float`` ``-enable-16bit-types`` ``OpTypeFloat 16`` ``Float16``
``min16int`` ``-enable-16bit-types`` ``OpTypeInt 16 1`` ``Int16``
``min12int`` ``-enable-16bit-types`` ``OpTypeInt 16 1`` ``Int16``
``min16uint`` ``-enable-16bit-types`` ``OpTypeInt 16 0`` ``Int16``
============== ======================= ================== ==================== ============ =================================
============== ======================= ================== ==================== ============

Vectors and matrices
--------------------
Expand Down
21 changes: 0 additions & 21 deletions tools/clang/lib/SPIRV/CapabilityVisitor.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -736,27 +736,6 @@ bool CapabilityVisitor::visit(SpirvExtInstImport *instr) {
return true;
}

bool CapabilityVisitor::visit(SpirvExtInst *instr) {
// OpExtInst using the GLSL extended instruction allows only 32-bit types by
// default for interpolation instructions. The AMD_gpu_shader_half_float
// extension adds support for 16-bit floating-point component types for these
// instructions:
// InterpolateAtCentroid, InterpolateAtSample, InterpolateAtOffset
if (SpirvType::isOrContainsType<FloatType, 16>(instr->getResultType()))
switch (instr->getInstruction()) {
case GLSLstd450::GLSLstd450InterpolateAtCentroid:
case GLSLstd450::GLSLstd450InterpolateAtSample:
case GLSLstd450::GLSLstd450InterpolateAtOffset:
addExtension(Extension::AMD_gpu_shader_half_float, "16-bit float",
instr->getSourceLocation());
break;
default:
break;
}

return visitInstruction(instr);
}

bool CapabilityVisitor::visit(SpirvAtomic *instr) {
if (instr->hasValue() && SpirvType::isOrContainsType<IntegerType, 64>(
instr->getValue()->getResultType())) {
Expand Down
1 change: 0 additions & 1 deletion tools/clang/lib/SPIRV/CapabilityVisitor.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -36,7 +36,6 @@ class CapabilityVisitor : public Visitor {
bool visit(SpirvImageOp *) override;
bool visit(SpirvImageSparseTexelsResident *) override;
bool visit(SpirvExtInstImport *) override;
bool visit(SpirvExtInst *) override;
bool visit(SpirvAtomic *) override;
bool visit(SpirvDemoteToHelperInvocation *) override;
bool visit(SpirvIsHelperInvocationEXT *) override;
Expand Down
4 changes: 0 additions & 4 deletions tools/clang/lib/SPIRV/FeatureManager.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -179,8 +179,6 @@ Extension FeatureManager::getExtensionSymbol(llvm::StringRef name) {
Extension::EXT_shader_stencil_export)
.Case("SPV_EXT_shader_viewport_index_layer",
Extension::EXT_shader_viewport_index_layer)
.Case("SPV_AMD_gpu_shader_half_float",
Extension::AMD_gpu_shader_half_float)
.Case("SPV_AMD_shader_early_and_late_fragment_tests",
Extension::AMD_shader_early_and_late_fragment_tests)
.Case("SPV_GOOGLE_hlsl_functionality1",
Expand Down Expand Up @@ -240,8 +238,6 @@ const char *FeatureManager::getExtensionName(Extension symbol) {
return "SPV_EXT_shader_stencil_export";
case Extension::EXT_shader_viewport_index_layer:
return "SPV_EXT_shader_viewport_index_layer";
case Extension::AMD_gpu_shader_half_float:
return "SPV_AMD_gpu_shader_half_float";
case Extension::AMD_shader_early_and_late_fragment_tests:
return "SPV_AMD_shader_early_and_late_fragment_tests";
case Extension::GOOGLE_hlsl_functionality1:
Expand Down

0 comments on commit 1574879

Please sign in to comment.