]> Cypherpunks repositories - gostls13.git/commit
iter: deflake TestPull by letting exiting goroutines finish
authorMichael Anthony Knyszek <mknyszek@google.com>
Thu, 23 May 2024 20:26:02 +0000 (20:26 +0000)
committerMichael Knyszek <mknyszek@google.com>
Fri, 24 May 2024 19:15:10 +0000 (19:15 +0000)
commit54efef99b2b9432c2eb6cd63a287a13d43b9bb7b
tree28d7e3a9e9f10984ad7201afbc0e379420a9baaf
parent1c924572d07df9c267028a20ee8934a94bfd7f8c
iter: deflake TestPull by letting exiting goroutines finish

Currently TestPull is flaky because goroutines spawned to run subtests
exit asynchronously when they finish and TestPull has explicit checks
for the number of existing goroutines.

This is pretty much only a problem between subtests executing, because
within each subtest the coroutine goroutine spawned for iter.Pull always
exits fully synchronously before the final `next` or `stop` returns.

So, we can resolve the problem by ensuring the first goroutine count the
test takes likely doesn't contain any exiting goroutines. The trick is
to set GOMAXPROCS=1 and spin in runtime.Gosched until the number of
goroutines stabilizes to some reasonable degree (we pick 100 consecutive
iterations; there are only a handful of possible goroutines that can
run, so this is giving that handful around 20 chances to actually run to
completion).

When running TestPull under stress2, this issue is easily reproducible
before this CL. After this CL, it no longer reproduces under these
conditions.

Fixes #66017.

Change-Id: I4bf0a9771f7364df7dd58f8aeb3ae26742d5746f
Reviewed-on: https://go-review.googlesource.com/c/go/+/587917
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
src/iter/pull_test.go