On Thu, Nov 22, 2018 at 11:17:22AM -0500, Jeff King wrote: > The script I use is at: > > https://github.com/peff/git/blob/meta/stress > > which you invoke like "/path/to/stress t5562" from the top-level of a > git.git checkout. It basically just runs a loop of twice as many > simultaneous invocations of the test script as you have CPUs, and waits > for one to fail. The load created by all of the runs tends to flush out > timing effects after a while. > > It fails for me on t5562 within 30 seconds or so (but note that in this > particular case it sometimes takes a while to produce the final output > because invoke-with-content-length misses the expected SIGCLD and sleeps > the full 60 seconds). I have observed it caught failure at the very first run. However I could not fail t again. I tried running up to 20 instances, with 1 or 2 active cores (that's all I have here), also edited the test to include only push_plain case, and repeat it several times, to avoid running irrelevant cases, the failure never happened again. The first failure was a bit unusual, in the ouput actually all tests were marked as passed, but it still failed somehow. Unfortunately, I did not save the output. I submitted the perl patch -- Max
