Add: run deferred_notify_demo onboard on a2a3 - #1584
Open
ChaoWao wants to merge 1 commit into
Open
Conversation
hw-native-sys#1568 left this demo sim-only because it failed in a full onboard sweep. That sweep carried no `--ignore` sets, so it included the SDMA demos that `ci.yml:600` quarantines out of the general sweep precisely because they are only correct in isolation — an invalid baseline, by the standard hw-native-sys#1580 has since written into the testing skill. Re-tested against a CI-shaped sweep (`$SDMA_IGNORE $HEAVY_IGNORE`, mirroring `ci.yml:607`): the demo passes onboard a2a3 in three independent runs on three different device pairs — [7, 9], [3, 5], [9, 1] — at ~14 s each, alongside the rest of `examples/a2a3`. It also still passes under a2a3sim, and passed standalone at `max_diff=0.000e+00`. Nothing in the demo is sim-specific: it uses the same CommDomain / CommBufferSpec path as `async_notify_demo`, which has always run onboard, and differs only in also writing a peer mailbox in the comm window. The a5 copy stays sim-only — this box is a2a3 silicon, so widening it there would be an unverified claim. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughThe deferred notify demo now documents coverage for onboard and sim environments and runs on both the ChangesDeferred notify coverage
Estimated code review effort: 1 (Trivial) | ~2 minutes Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
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.
Why this was sim-only, and why that was wrong
#1568 left
deferred_notify_demomarkedplatforms(["a2a3sim"])because it failed when I temporarily widened it and ran a full onboard sweep.That sweep carried no
--ignoresets, so it included the SDMA demos thatci.yml:600quarantines out of the general sweep — precisely because they are only correct in isolation. By the standard #1580 has since written into the testing skill, that was an invalid baseline. The demo was condemned on evidence that does not stand.Re-test against a CI-shaped sweep
Mirroring
ci.yml:607($SDMA_IGNORE $HEAVY_IGNORE, no--runtime, so the real dispatcher runs), onboard a2a3:examples tests/stdevices=[9, 1]examples tests/stdevices=[3, 5]examples/a2a3(ran to completion, exit 0)devices=[7, 9]Three independent runs, three different device pairs. Runs 1 and 2 were cut short at 60–70% by the
task-submitdaemon restarting (exit=137) — infrastructure, not a test: both showed 55 passed / 0 failed at the point of interruption. Run 3 used a narrower scope and completed cleanly.Also still green under
a2a3sim(PASS 14.7 s), and passes standalone atmax_diff=0.000e+00.Why it was always plausible
Nothing in the demo is sim-specific. It uses the same
CommDomain/CommBufferSpecpath asasync_notify_demo, which has always run onboard; a line-by-line diff shows the only real delta is thatdeferredadditionally writes a peer mailbox in the comm window.Scope
a2a3 only. The a5 copy stays
["a5sim"]— this box is a2a3 silicon, so widening it there would be an unverified claim. Worth a look by someone with a5 hardware, since the same reasoning likely applies.🤖 Generated with Claude Code