Re: Re: Broken: php/php-src#6708 (PHP-5.6 - 6538dab)

On Tue, Jun 3, 2014 at 3:16 PM, Andrea Faulds <ajf@ajf.me> wrote: > > On 3 Jun 2014, at 12:06, Ferenc Kovacs <tyra3l@gmail.com> wrote: > > > Hi, > > > > anybody have any idea why did > ext/mysqli/tests/mysqli_reap_async_query.phpt > > started to fail after my commit which only changed the NEWS file? > > is this an intermittent test, or maybe the mysql lib on the travis > basebox > > got updated or something? > > At the risk of asking the obvious, did you check the log travis provides? > I’ve seen travis fall apart on me before and have tests fail, not because > my code was bad, but because some software update broke compilation > altogether. > yes, you should be able to check yourself from the mail I've forwarded in my original mail: https://travis-ci.org/php/php-src/builds/26644402 only a single test failed (ext/mysqli/tests/mysqli_reap_async_query.phpt), only in the ENABLE_MAINTAINER_ZTS=0 ENABLE_DEBUG=0 axis, with the following diff: 004+ [013 + 1] There should be 1 links ready to read from, 0 ready 004- [013 + 2] array ( 005- '_two' => '2', 006- ) then the issue resolved itself without anybody fixing anything relevant: https://travis-ci.org/php/php-src/builds/26649071 so it was either a travis hiccup which we have no control over, or our test can fail from time to time (intermittent test failure), in which case we should try to eliminate the non-determinism. -- Ferenc Kovács @Tyr43l - http://tyrael.hu