[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: 2.4.14 prerelease call for testing #2



I wrote:
>Howard Chu writes:
>> That's interesting. That would require receiving a Hangup event on a 
>> descriptor we've already removed from epoll's event list. How can that be? 
> 
> OS bug?  Race condition?
> Valgrind bug, since it so far will only happen under Valgrind?

Happened without Valgrind too, though it took 55 runs to trigger.
Same backtrace except this time it's not on the sole connection:

  sd=11 which is conn=4.
  *slap_daemon.sd_index@128 =
    {-1, -1, -1, -1,  0,  -1, -1,  1, -1,  2,
      3, -1, -1, -1, -1,  -1, -1, -1,  4, -1 <repeats 109 times>}.

Activity on another connection (conn=5) between the conn=4 UNBIND
and the crash.  Testrun dir in <http://folk.uio.no/hbf/testrun.2.tbz>.

It has not happened so far on 2.4.13.

-- 
Hallvard