1. Each OpEntryPoint with the MeshEXT Execution Model can have at most one global OpVariable of storage class TaskPayloadWorkgroupEXT.
2. PerPrimitiveEXT only be used on a memory object declaration or a member of a structure type
3. PerPrimitiveEXT only Input in Fragment and Output in MeshEXT
4. Added Mesh vulkan validation support for following rules:
VUID-Layer-Layer-07039 VUID-PrimitiveId-PrimitiveId-07040,VUID-PrimitivePointIndicesEXT-PrimitivePointIndicesEXT-07042,
VUID-PrimitiveLineIndicesEXT-PrimitiveLineIndicesEXT-07048, VUID-PrimitiveTriangleIndicesEXT-PrimitiveTriangleIndicesEXT-07054,
VUID-ViewportIndex-ViewportIndex-07060 VUID-StandaloneSpirv-ExecutionModel-07330 VUID-StandaloneSpirv-ExecutionModel-07331
This fixes the problem reported in #5623 using the observation that if
we are re-building a type that already exists in the type pool, we
should just return that type.
This makes type rebuilding more efficient, and it also prevents the
type builder from getting itself into infinite recursion (as reported in
this issue).
In fixing this, I found a couple of other bugs in the type builder:
- When rebuilding an Array type, we were not re-building the element
type. This caused stale type references in the rebuilt type.
- This bug had not been caught by the test, because the test itself had
a bug in it: the test was rebuilding types on top of the same ID (the
ID counter was never incremented).
Initially, the bug in the test caused a failure with the new logic in
the builder because we now return types from the pool directly, which
causes a failure when two incompatible types are registered under the
same ID.
Fixing that issue in the test exposed another bug in the rebuilder: we
were not re-building the element type for Array types. This was causing
a stale type reference inside Array types which was later caught by the
type removal logic in the test.
GitHub releases were just the exact same as the LunarG releases,
since the same commit were used for both.
For this reason, we don't think it brings any value, and will stop
publishing new releases on the GitHub page.
Signed-off-by: Nathan Gauër <brioche@google.com>
6598a8ecd5...917047f360
$ git log 6598a8ecd..917047f36 --date=short --no-merges --format='%ad %ae %s'
2024-03-28 junyer Tell Bazel the specific target version of macOS too. Sigh.
2024-03-28 junyer Force a specific target version of macOS.
2024-03-28 junyer Prepare to tag release `2024-04-01`.
Created with:
roll-dep external/re2
Co-authored-by: GitHub Actions[bot] <>
* Roll external/googletest/ b479e7a3c..c231e6f5b (1 commit)
b479e7a3c1...c231e6f5b1
$ git log b479e7a3c..c231e6f5b --date=short --no-merges --format='%ad %ae %s'
2024-03-14 absl-team Add test for move-only type in `Action` signature
Created with:
roll-dep external/googletest
* Roll external/re2/ 108914d28..db46d1e11 (3 commits)
108914d28a...db46d1e11e
$ git log 108914d28..db46d1e11 --date=short --no-merges --format='%ad %ae %s'
2024-03-17 junyer Bump versions of actions to latest releases.
2024-03-17 junyer Fix the link to the Python wrapper.
2024-03-15 oleksandr.red+github doc/mksyntaxgo: use standard generated Go header
Created with:
roll-dep external/re2
---------
Co-authored-by: GitHub Actions[bot] <>
* Roll external/googletest/ 5df0241ea..b479e7a3c (11 commits)
5df0241ea4...b479e7a3c1
$ git log 5df0241ea..b479e7a3c --date=short --no-merges --format='%ad %ae %s'
2024-03-11 dcheng Guard Abseil flags usage in googlemock with GTEST_NO_ABSL_FLAGS.
2024-03-07 absl-team Reland: Optimize Google Test process startup
2024-03-05 absl-team Revert Optimize Google Test process startup
2024-03-05 absl-team Optimize Google Test process startup
2024-02-29 absl-team Accept one-shot callables in InvokeArgument.
2024-02-27 absl-team Reland 9756ee7cba
2024-02-27 absl-team Make sure that current_test_suite and current_test_info are mutex-protected while writing for thread-safety.
2024-02-23 tmiguelf Fixed gcc linker error 58
2024-02-22 absl-team Revert incorrect update.
2024-02-21 dmauro Rollback 9756ee7cba
2024-02-11 155268264+danfabo Get include dirs from target rather than global variables.
Created with:
roll-dep external/googletest
* Roll external/re2/ dbf15a205..108914d28 (3 commits)
dbf15a205c...108914d28a
$ git log dbf15a205..108914d28 --date=short --no-merges --format='%ad %ae %s'
2024-03-11 junyer `p0deje/setup-bazel` has been transferred to `bazel-contrib`.
2024-03-11 junyer Add Clang 18 to the build matrix.
2024-03-07 junyer Delete an old comment.
Created with:
roll-dep external/re2
---------
Co-authored-by: GitHub Actions[bot] <>
* Use bazel 7 and bzlmod
Many of our dependencies are now using bazel 7 and the new bzlmod module
system. This has been breaking our autoroll; this PR should fix the
issues we've been having with that.
It may or may not be worthwhile to update effcee to use bzlmod as well;
this would let us get rid of WORKSPACE entirely.
* Try to force cla check to rerun
All tests treat kAnalysisEnd like STL end iterators, which
means its value must be greater than that of the last valid
Analysis.
Change-Id: Ibfaaf60bb450c508af0528dbe9c0729e6aa07b3b
Signed-off-by: Kevin Petit <kevin.petit@arm.com>
Flagged by -Wredundant-decls
I'm assuming the declarations in libspirv.h are part of the external
interface and need to be kept.
Change-Id: I6b138d3322a7a4ee49ee33b0fbcf0ca35dd92261
Signed-off-by: Kevin Petit <kevin.petit@arm.com>
DXC does not do a good job of recognizing which variables need to be
on the entry point for which functions. This is because it does not
want to have to walk the call tree to determine which instructions
are reachable from which entry points.
This is also useful if the same input variable gets used from two
different shader, but the uses in one get optimized away.
Will parially fix
https://github.com/microsoft/DirectXShaderCompiler/issues/4621. Will not
fix code compiled with -fcgl.
* [OPT] Identify arrays with unknown length in copy prop arrays
The code in copy propagate arrays assumes that the length of an
OpTypeArray is known at compile time, but that is not true when the size
is an OpSpecConstant. We try to fix that assumption.
Fixes https://crbug.com/oss-fuzz/66634
We are removing Kernel from the image channel order and image
channel data type enums because Kernel is already required
transitively, so we need to update the tests to match.
* [OPT] Use new instruction folder for for all opcodes in spec consti folding
When folding and OpSpecConstantOp, we use the new instruction folder for
a small number of opcodes. This enable the new instruction folder for
all opcodes and uses the old one as a fall back. This allows us to
remove some code from the older folder that is now covered by the new
one.
Fixes#5499
Adds folding rules that will fold basic artimetic for signed and
unsigned integers of all sizes, including 64-bit.
Also folds OpSConvert and OpUConvert.
