Skip Menu |

This queue is for tickets about the IO-Socket-Socks CPAN distribution.

Report information
The Basics
Id: 97619
Status: resolved
Priority: 0/
Queue: IO-Socket-Socks
People
Owner: oleg [...] cpan.org
Requestors: sri [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: 0.65_1

History Show all quoted text
  Wed Jul 30 07:36:06 2014 sri [...] cpan.org - Ticket created
Subject: Test failures on OS X
About 9 out of 10 times i see tests fail just like this on OS X 10.9.4 with Perl 5.20.0. http://www.cpantesters.org/cpan/report/78d47f7e-90da-11e3-830c-efd9924679b0 Here's what i have so far, the failures are pretty consistent, and i've tracked it down to this line. https://github.com/olegwtf/p5-IO-Socket-Socks/blob/6e75bda0355f7d4c9fbb353e241136113113e03e/lib/IO/Socket/Socks.pm#L2010 The sysread() call on line 2000 returns a 0, which then makes _socks_read() return an empty string, which then triggers a _fail() call with that empty string, resulting in the "Timeout" diagnostics message. P.S.: Afraid i can't offer you shell access to an OS X box, but if you need someone to run more tests for you, feel free to poke me on IRC. (sri on irc.perl.org) -- sebastian
  Wed Jul 30 07:39:24 2014 sri [...] cpan.org - Correspondence added
One more data point, i've been running a lot of non-blocking SOCKS tests with Mojolicious, and there have been no random failures. So this may very well be a flaw in the accept tests. -- sebastian
  Wed Jul 30 07:59:54 2014 oleg [...] cpan.org - Correspondence added
Срд Июл 30 07:39:24 2014, SRI писал: Show quoted text
> One more data point, i've been running a lot of non-blocking SOCKS > tests with Mojolicious, and there have been no random failures. So > this may very well be a flaw in the accept tests. > > -- > sebastian
Can you tell me on which sockets this fails always occures? On sockets created by accept (%server_clients) or client sockets (%local_clients) or on both?
  Wed Jul 30 07:59:54 2014 The RT System itself - Status changed from 'new' to 'open'
  Wed Jul 30 08:53:08 2014 sri [...] cpan.org - Correspondence added
Show quoted text
> Can you tell me on which sockets this fails always occures? On sockets > created by accept (%server_clients) or client sockets (%local_clients) > or on both?
Only %local_clients, and always 9 out of 10. -- sebastian
  Thu Jul 31 14:07:32 2014 oleg [...] cpan.org - Correspondence added
On Wed Jul 30 08:53:08 2014, SRI wrote: Show quoted text
> > Can you tell me on which sockets this fails always occures? On sockets > > created by accept (%server_clients) or client sockets (%local_clients) > > or on both?
> > Only %local_clients, and always 9 out of 10. > > -- > sebastian
Looks fixed for me now. Please fetch osx_bug branch from https://github.com/olegwtf/p5-IO-Socket-Socks.git , run tests and report results here
  Fri Aug 01 13:35:43 2014 sri [...] cpan.org - Correspondence added
Show quoted text
> Looks fixed for me now.
Just tried IO::Socket::Socks 0.65_1 and everything works perfect now! :) -- sebastian
  Fri Aug 01 13:41:19 2014 oleg [...] cpan.org - Correspondence added
On Fri Aug 01 13:35:43 2014, SRI wrote: Show quoted text
> > Looks fixed for me now.
> > Just tried IO::Socket::Socks 0.65_1 and everything works perfect now! :) > > -- > sebastian
Ok, let's see CPAN Testers result before closing it.
  Tue Aug 05 04:47:44 2014 oleg [...] cpan.org - Correspondence added
Птн Авг 01 13:41:19 2014, OLEG писал: Show quoted text
> Ok, let's see CPAN Testers result before closing it.
No fail reports received.
  Tue Aug 05 04:47:45 2014 oleg [...] cpan.org - Status changed from 'open' to 'resolved'
  Tue Aug 05 04:47:45 2014 oleg [...] cpan.org - Taken
  Tue Aug 05 04:47:46 2014 oleg [...] cpan.org - Fixed in 0.65_1 added