fs: cancel in-flight stat on abort

[mertcanaltin]

The AbortError behavior was landed in main with #63143, leaving only the uv_cancel call that actually dropped the request from the threadpool queue in this PR.

[bakkot]

Drive-by: this looks like it's still not actually canceling the underlying call, which would involve invoking uv_cancel somewhere.

[codecov]

Codecov Report

❌ Patch coverage is 85.71429% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.07%. Comparing base (46a7dbd) to head (6e2fc83).
⚠️ Report is 25 commits behind head on main.

Files with missing lines	Patch %	Lines
src/node_file.cc	83.33%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #63142      +/-   ##
==========================================
- Coverage   90.21%   90.07%   -0.14%     
==========================================
  Files         751      751              
  Lines      253550   254427     +877     
  Branches    47813    47976     +163     
==========================================
+ Hits       228733   229168     +435     
- Misses      16076    16436     +360     
- Partials     8741     8823      +82     

Files with missing lines	Coverage Δ
lib/fs.js	98.43% <100.00%> (-0.01%)	⬇️
src/node_file.cc	74.15% <83.33%> (+0.01%)	⬆️

... and 66 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mertcanaltin]

Thanks for review @bakkot I added a new binding function for cancel

[trivikr]

cc @nodejs/fs

[mertcanaltin]

@Renegade334, could you review this pr?

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/76286/

[mertcanaltin]

I measuring uv_cancel.

[trivikr]

@mertcanaltin Merge commits don't work with Node.js tooling. Can you please rebase?

[mertcanaltin]

Thanks @trivikr, I've rebased now and measured.

UV_THREADPOOL_SIZE=1, pool blocked by a pbkdf2 job, all 10,000 stadiums that were aborted while the pool was clogged are actually cancelled.

	executed	canceled	settled
without cancel	10000	0	528 ms
with cancel	0	10000	14 ms

[mertcanaltin]

Update: The AbortError behavior was landed in main with #63143, leaving only the uv_cancel call that actually dropped the request from the threadpool queue in this PR.

[atlowChemi]

Should there be some new test ensuring this works?

[mertcanaltin]

Should there be some new test ensuring this works?

Thanks, added.

[gurgunday]

lgtm

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/76513/