On Sat, Mar 11, 2017 at 01:54:57PM +0100, Matthias Fechner wrote: > Am 11.03.2017 um 13:45 schrieb Willy Tarreau: > > I don't understand, the bisection didn't end ? > > > > Otherwise I'm inclined to think that the regression comes from > > "BUG/MEDIUM: tcp: don't poll for write when connect() succeeds", which > > it the one I proposed you to revert and which didn't change anything. > > > > Did "git bisect" end up telling you "This is the first bad commit" ? > > I had a typo, here the result (typed git bisect goog instead of good): > > cd4c5a3ecf5e77fb4734c423c914f7280199c763 is the first bad commit > commit cd4c5a3ecf5e77fb4734c423c914f7280199c763 > Author: Willy Tarreau <[email protected]> > Date: Wed Jan 25 14:12:22 2017 +0100 > > BUG/MEDIUM: tcp: don't poll for write when connect() succeeds
(...) OK so this is the one that I initially suspected and that after you reverted, didn't fix the issue for you. Are you sure you didn't have a problem when you reverted it ? (eg: failed to restart the process or something like this). At least now we're certain it's this one so we can try to imagine any possible corner case or any remaining bug this one could have uncovered. Thanks! Willy
