Repository navigation
ci: pinned-memory diagnostics with isolated wheel imports - #3458
Draft
jpascucci-nv wants to merge 4 commits into
Draft
jpascucci-nv wants to merge 4 commits into
jpascucci-nv wants to merge 4 commits into
Conversation
Run dedicated Windows L4 diagnostics for pinned-memory failures using the wheels selected by CI. Launch pytest in isolated Python processes so the source checkout cannot shadow the installed cuda.core wheel and cause the missing cuda.core._version import error. Verify wheel imports, reject all-skipped diagnostic runs, and preserve per-case logs, JUnit results, and first-OOM reports. Validation: 46 targeted tests passed; Ruff and actionlint passed. GPU validation pending CDCI.
jpascucci-nv
added
bug
Contributor
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
Contributor
Author
|
/ok to test 9ee75ba |
Contributor
Compare legacy, per-thread, and nonblocking streams with and without access revocation. Alternate watermark resets with minimal logging, preserve the last API boundary across crashes, and time out hung probes. Capture fixture OOM diagnostics before popping the CUDA context, preserving the original failure and the first diagnostic snapshot. Validation: 29 CPU checks, crash/hang checks, and Ruff passed. GPU validation pending CDCI.
Contributor
Author
|
/ok to test 2a7d7b0 |
Build a standalone C++ driver-API probe to isolate failures from Python bindings. Compare stream and access-revocation combinations, plus controls omitting trim, omitting watermark reset, or reusing a pool. Preserve first-failure diagnostics and progress through crashes and timeouts, and upload the executable, source, build log, and results. Validate with 22 focused tests and compilation against CUDA headers.
Contributor
Author
|
/ok to test 10b05ff |
Contributor
Author
|
/ok to test e77c968 |
Contributor
Author
|
opened https://nvbugspro.nvidia.com/bug/7092346 with reproducer cpp. |
This branch has not been deployed
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.
Description
Run dedicated Windows L4 diagnostics for pinned-memory failures using the wheels selected by CI. Launch pytest in isolated Python processes so the source checkout cannot shadow the installed cuda.core wheel and cause the missing cuda.core._version import error.
Verify wheel imports, reject all-skipped diagnostic runs, and preserve per-case logs, JUnit results, and first-OOM reports.
Validation: 46 targeted tests passed; Ruff and actionlint passed. GPU validation pending CDCI.
Checklist