[Commented On] D9952: revlog: add a mechanism to verify expected file position before appending

Alphare (Raphaël Gomès) phabricator at mercurial-scm.org
Tue Apr 20 08:49:19 UTC 2021


Alphare added a comment.


  In D9952#159118 <https://phab.mercurial-scm.org/D9952#159118>, @spectral wrote:
  
  > I'm unable to reproduce. I've run the test over 10,000 times (I added a `#testcases a b c d e f g h i j k l m n o p q r s t u v w x y z` so it ran 26 times each run, and I've run over 300 instances of that like `run-tests.py -j26 -l --chg test-racy-mutations.t`, over 100 with -j108, and other combinations (with and without the added testcases, with and without `--chg`, etc.)
  > I'm on a rather beefy linux machine, based on Debian testing, using Python3.9, on a commit descended from 856820b4 <https://phab.mercurial-scm.org/rHG856820b497fcf31b7904e0b160c318437581d0c0>.
  > It hasn't failed a single time, let alone like that. I also don't know what would cause that kind of failure. :(
  
  Mhhh. It may be possible that your machine is too fast to cause the race. Your runners are not the fastest machines in the world and they are under quite a bit of load; it wouldn't be the first time I've seen that happen. Could you try to artificially load up your CPU while you run the tests with chg? Otherwise we'll have to dig deeper.

REPOSITORY
  rHG Mercurial

CHANGES SINCE LAST ACTION
  https://phab.mercurial-scm.org/D9952/new/

REVISION DETAIL
  https://phab.mercurial-scm.org/D9952

To: spectral, indygreg, #hg-reviewers
Cc: Alphare, martinvonz, joerg.sonnenberger, mercurial-patches
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial-patches/attachments/20210420/35f29aa1/attachment-0002.html>


More information about the Mercurial-patches mailing list