I'm pretty well convinced that this is a separate issue. (Although
it might be related.) I tried all the fixes and patches I could find
that relate to the select/poll 64k bug you're talking about. I tried
them against Apache 2.0.54 and 2.0.55 (the latest). The crashes I
experienced only went away when I removed "--with-mpm=worker" from
the Apache configure command line.
Maybe multi-threading exposes the select/poll bug in a way that isn't
covered by the fix that got in to 2.0.55. Or maybe it's a completely
different, unknown bug!
On 25 Nov 2005, at 12:54 PM, Will Senn wrote:
>
> On Nov 25, 2005, at 12:29 PM, Michael Sinz wrote:
>
>> Will Senn wrote:
>>> Hi all,
>>> I originally posted this to user@, but was directed to post to
>>> this list as a more appropriate venue. Could y'all take a look
>>> and tell me what might be the matter?
>>
>> This is a "known" issue with OS-X 10.4 and Apache 2.0.x...
>>
>> The issue is actually both complex and simple at the same time.
>> There
>> are fixes to the Apache code that address the issue. It mainly
>> has to
>> do with the Apache 2.0.x server early termination of longer I/O
>> writes
>> on Mac OS-X 10.4 - but the code path clearly shows that it could
>> happen
>> in other conditions but clearly has not been seen to happen.
>>
>> See http://issues.apache.org/bugzilla/show_bug.cgi?id=34332
>>
>> (note that it is marked as "fixed" - it depends on the build of
>> Apache
>> you have (2.0.55 should be good but I am not 100% sure on that)
>>
>> --
>> Michael Sinz Technology and Engineering
>> Director/Consultant
>> "Starting Startups"
>> mailto:michael.sinz@sinz.org
>> My place on the web http://www.sinz.org/
>> Michael.Sinz
>
> Michael,
>
> Thanks for the link, that explains what's happening. However, I'm
> not sure that the 'fix' is in... Certainly, it doesn't appear to be
> working in my build. We'll have to wait and see if it's corrected
> in the next version. In the meantime, I'm going to try Paul's
> suggestion of removing the --with-mpm=worker flag, as that seems
> related to the fact that it will execute ok in the debugger single
> process mode, run -X and with the maintainer flag on - not sure
> what all that does, but it 'works'.
>
> Thanks,
>
> Will
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
> For additional commands, e-mail: dev-help@subversion.tigris.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Nov 25 21:28:03 2005