[PATCH] D80293: [clangd] Run PreambleThread in async mode behind a flag

[Pieb, Wolfgang via cfe-commits]

The test failed 7-8 times in a row on the same build machine and then started passing for about 10 runs, and then started failing again for 4 more runs, after which we disabled it.
The configuration is nothing special, simply a Release+Asserts build on Windows. We tried to reproduce it by hand on the same machine, but weren’t able to. The test ran in less than half a second when we tried it (the timeout is 10s).

Sorry I can’t be of more help. Perhaps we can turn on some tracing output that would help diagnosing the problem?

  *   wolfgang

From: Kadir Çetinkaya <kadircet at google.com>
Sent: Tuesday, June 2, 2020 1:20 PM
To: reviews+D80293+public+6cba1e506b8b42d8 at reviews.llvm.org
Cc: Sam McCall <sammccall at google.com>; Pieb, Wolfgang <wolfgang.pieb at sony.com>; Ilya Biryukov <iu.biryukov at gmail.com>; javeda at graphcore.ai; Fangrui Song <maskray at google.com>; Jan Korous <jkorous at apple.com>; Aleksei Lorenz <arphaman at gmail.com>; jfbastien at apple.com; Utkarsh Saxena <usx at google.com>; cfe-commits <cfe-commits at lists.llvm.org>; Theko Lekena <mlekena at skidmore.edu>; Nicolas Lesser <blitzrakete at gmail.com>; Han Shen <shenhan at google.com>; Haojian Wu <hokein at google.com>
Subject: Re: [PATCH] D80293: [clangd] Run PreambleThread in async mode behind a flag

thanks for letting me know. it is annoying that the test is flaky, but I don't see that flakiness on any other buildbot I've access to.

Is there any chance that you are running in a custom build configuration that I can try to repro the failure? Also what's the falkiness ratio?

On Tue, Jun 2, 2020 at 1:43 AM Wolfgang Pieb via Phabricator <reviews at reviews.llvm.org<mailto:reviews at reviews.llvm.org>> wrote:
wolfgangp added a comment.

Hi!

Shortly after this patch went in, we started to see intermittent failures with the unittest TUSchedulerTests.ManyUpdates. The build server (Windows) we're running it on is quite powerful and our own hand-run tests only take about 400 ms, so we don't really understand what could be happening. Any help would be appreciated.

Here is an excerpt from the log:

Note: Google Test filter = TUSchedulerTests.ManyUpdates

[==========] Running 1 test from 1 test case.

[----------] Global test environment set-up.

[----------] 1 test from TUSchedulerTests

[ RUN      ] TUSchedulerTests.ManyUpdates

C:\j\w\...\clang-tools-extra\clangd\unittests\TUSchedulerTests.cpp(508): error: Value of: S.blockUntilIdle(timeoutSeconds(10))

  Actual: false

Expected: true

[  FAILED  ] TUSchedulerTests.ManyUpdates (10576 ms)

[----------] 1 test from TUSchedulerTests (10576 ms total)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D80293/new/

https://reviews.llvm.org/D80293


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20200602/5ff9e51d/attachment.html>