HLSL allows image and texture types to be templatized on sub-vec4 types,
or even structures. This was mostly handled already during creation of
sampling operations. However, for operator[] which can generate image
loads, this wasn't happening.
It also isn't very easy to do at that point in time, because operator[]
does not know where the results it produces will end up. They may be
an lvalue or an rvalue, and there's a post-process to convert loads to
stores. They may end up in atomic ops.
To bypass that difficulty, GlslangToSpv now looks for this case and
adds the appropriate conversion. LIMITATION: this only works for
cases for which a simple conversion opcode suffices. That is to say,
it will not work if the type is templatized on a struct.
Update SPIRV-Tools. Relevant functional changes:
- Optimizer enhancements:
- ADCE now removes OpSwitch
- Block merging occurs in more cases
- Optimizer fixes:
- Constant propagation (CCP): support matrix constants
- #1199: Optimizer: Fix CCP: don't propagate spec constants.
- #1203: Optimizer: Fix common uniform elim bug introduced by refactoring.
- #1210: Optimizer: Aggressive dead code elimination: Fix 'break' identification.
- #1212: Optimizer: Aggressive dead code elimination: Was skipping too many instructions.
- #1214: Optimizer: Aggressive dead code elimination: Fix infinite loop.
- #1228: Optimizer: Fix CCP: Handling of varying Phi nodes; was resulting in infinite loop.
- #1245: Optimizer: Dead branch elimination: Avoid a null pointer dereference.
- #1250: Optimizer: Dead branch elimination: Avoid spuriously reporting a change.
- #1262: Support building on VisualStudio 2013 again
Update SPIRV-Headers, with "unified1" directory.
Updated one Glslang legalization test base result due to better block merging.
- make it sharable with GLSL
- correct the case insensitivity
- remove the map; queries are not needed, all entries need processing
- make it easier to build bottom up (will help GLSL parsing)
- support semantic checking and reporting
- allow front-end dependent semantics and attribute name mapping
- correct inheritence (or not) of the right XFB buffer
- compute implicit stride (fixes#1212)
- semantic check block-member redeclarations
- inherit stride from a member
- fixes#1209, addresses most of #1187
- only query feature availability on seeing the feature
(was doing it for every single token)
- correct case-sensitive checks for multi-character suffixes
- partially addresses #1209 and #1187
- only query 64-bit extension on seeing use of a 64-bit literal
(was doing it for every single token)
- correct HLSL acceptance of 64-bit literal syntax (still an int though)
- error on overflow of 32-bit literal type
The grammar for no semicolon and no object name for cbuffer/tbuffer
was correct, but the production still skipped the anonymous declarations
if an identifier followed.
Set type to r-value resulting from indexing vector, to prevent
float->uint conversion when source is already uint. Resulting
OpConvertFToU would otherwise fail validation because source is
already uint.
For LoadN, incorrect uint->float->uint can be avoided; fixing
potential truncation of big integer values.
Also, only emit this XFB information where the SPIR-V spec says
it should be emitted: essentially, on objects.
This and the previous commit together fix#1185.
Some stage (e.g, hull shaders) have arrayed builtin outputs (e.g, position).
When copying from the internal structure to the split form, it is necessary
to propagate that indirection to the actual arrayed outputs. This was not
happening.
Addresses #1181
This makes struct returns from functions work, but breaks
structs containing arrays, due to limitations in subsequent
transforms in spirv-opt. This is expected to be fixed soon.
Adds command line options:
--invert-y
--iy
(synonyms) which invert position.Y on vertex shader output. Handles these cases:
* Direct single variable return
* Member of direct returned struct
* Single variable output parameter
* Member of struct output parameter
API:
// Enables position.Y output negation in vertex shader
void TShader::setInvertY(bool invert);
Fixes#1173
This continues to prevent writing output buffers (out from a function),
but fixes the problem where the copy-in/out was not getting done.
Making everything work will require knowing both in/out-ness and bufferness,
but these are currently mutually exclusive, because both are storage
qualifiers.
Issue #791 was partially fixed by PR #1161 (the mat mul implicit
truncations were its main point), but it still wouldn't compile due to
the use of ConstantBuffer as an identifier. Apparently those fall into
the same class as "float float", where float is both a type and an
identifier.
This allows struct definitions with such keyword-identifiers,
and adds ConstantBuffer to the set. 'cbuffer int' is legal in HLSL,
and 'struct int' appears to only be rejected due to the redefinition
of the 'int' type.
Fixes#791
HLSL truncates the vector, or one of the two matrix dimensions if there is a
dimensional mismatch in m*v, v*m, or m*m.
This PR adds that ability. Conversion constructors are added as required.
If a shader includes a mixture of several stages, such as HS and GS,
the non-stage output geometry should be ignored, lest it conflict
with the stage output.
Per feedback on PR #1111, this reverses the order of the parameters for the setShiftBinding API.
It is now:
void TShader::setShiftBindingForSet(TResourceType res, unsigned int base, unsigned int set);
This script will crank a supplied set of glslang test shaders through the
spirv-val tool, reporting on the results.
There are some important things to note:
* Like 'runtests', this must be invoked from the 'Test' subdirectory.
* This is mostly useful on the hlsl.* tests, although it is not strictly
limited to those. The reason is that most of the glsl tests either contain
validation error cases, and so fail to compile, or are not using a #version
compatible with producing SPIR-V modules.
* Some tests, such as negative tests, or most of the glsl tests, have
intentional compilation errors. This script treats that as OK. Failures
are successfully compiling shaders which proceed to fail spirv-val.
* spirv-val is looked for in either the External directory, or if not
found there, in a sibling directory of glslang, and if not found there
either, in /usr/local/bin.
* There are a bunch of command line options. ./validate-shaders.sh --help
will describe them.
Some examples to try:
./validate-shaders.sh hlsl.* # exercise all hlsl.* tests.
./validate-shaders.sh --terse hlsl.* # same, but tersely.
# dump validator results for problems in something.frag:
./validate-shaders.sh --quiet --dump-val something.frag
Both debug and release clang builds have segfaulted on recent
changes, non deterministically, while doing the single/multi-thread
test all test files. Removing recent test files, to see if it gives
a clue.
This PR adds the ability to provide per-descriptor-set IO mapping shift
values. If a particular binding does not land into a per-set value,
then it falls back to the prior behavior (global shifts per resource class).
Because there were already 6 copies of many different methods and internal
variables and functions, and this PR would have added 6 more, a new API is
introduced to cut down on replication and present a cleaner interface.
For the global (non-set-specific) API, the old entry points still exist
for backward compatibility, but are phrased internally in terms of the
following.
// Resource type for IO resolver
enum TResourceType {
EResSampler,
EResTexture,
EResImage,
EResUbo,
EResSsbo,
EResUav,
EResCount
};
Methods on TShader:
void setShiftBinding(TResourceType res, unsigned int base);
void setShiftBindingForSet(TResourceType res, unsigned int set, unsigned int base);
The first method replaces the 6 prior entry points of various spellings, which
exist now in depreciated form. The second provides per-resource-set functionality.
Both accept an enum from the list above.
From the command line, the existing options can accept either a single shift value as
before, or a series of 1 or more [set offset] pairs. Both can be provided, as in:
... --stb 20 --stb 2 25 3 30 ...
which will use the offset 20 for anything except descriptor set 2 (which uses 25) and
3 (which uses 30).
