Make PSV/DebugBreak unit tests robust to validator version#8625
Merged
alsepkow merged 2 commits intoJul 16, 2026
Conversation
Capping the highest shader model to 6.9 (PR microsoft#8618) lowers the validator version to 1.9, so the compiler emits PSVRuntimeInfo3 (52 bytes) instead of PSVRuntimeInfo4 (56 bytes) and rejects cs_6_10, which broke 14 unit tests. - PixTest DebugBreakInstrumentation_{Basic,Multiple}: skip when Dxil/ Validator < 1.10 (DebugBreak is an experimental SM 6.10 feature). - ValidationTest PSV tests: derive the expected PSVRuntimeInfo size from the validator version (v4 only for >= 1.10, else v3) instead of hard-coding sizeof(PSVRuntimeInfo4), and format the expected size into the WrongPSVVersion mismatch messages. Validated locally against a capped x64 build: the 14 tests pass, and the full ValidationTest + PixTest suites (302) and clang FileCheck lit suite show no failures. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 086148ba-a82b-41c4-8fa0-5eb63e5797e2
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 942a5e5f-0252-472b-9529-7218fdab7664
Contributor
Author
|
I have a draft PR running here which has these changes plus the version change: #8623 This gives us proper validation of the change pre checkin. |
V-FEXrt
approved these changes
Jul 15, 2026
joaosaffran
approved these changes
Jul 15, 2026
alsepkow
enabled auto-merge (squash)
July 15, 2026 23:54
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The PSV and DebugBreak unit tests hard-code the newest PSV runtime-info version, which is a latent fragility unrelated to any single release.
ValidationTestPSV tests hard-codedsizeof(PSVRuntimeInfo4)(and the literal string56). That silently assumes the validator always emits the newest PSV struct. It is wrong for any validator< 1.10(which emitsPSVRuntimeInfo3), including downlevel/external-validator runs and any future shader-model cap. The expected size is now derived from the validator version via aGetExpectedPSVRuntimeInfoSize()helper.PixTest::DebugBreakInstrumentation_{Basic,Multiple}hard-codecs_6_10; they now skip when Dxil/Validator< 1.10.This is behavior-preserving on
main(validator is 1.10 there, so PSV v4 is still expected and the tests still run), while removing the fragility. The latent bug was introduced in #7871, which addedPSVRuntimeInfo4and the< 1.10 -> v3gating together; main has always run validator 1.10 since, so thev3path was never exercised.Intended to be cherry-picked to
release-1.9.2607(where #8618 caps the validator to 1.9), unblocking the SM 6.9 cap.Assisted by co-pilot