Front page | perl.perl5.porters |
Postings from November 2010
On Sun, Nov 28, 2010 at 09:35:30AM -0600, Craig A. Berry wrote:
> In article <E1PM0no-00078L-Re@camel.ams6.corp.booking.com>,
> davem@iabyn.com ("Dave Mitchell") wrote:
>
> > In perl.git, the branch blead has been updated
> >
> > <http://perl5.git.perl.org/perl.git/commitdiff/abf9167d3fff002ddaed53abb44d638
> > 387bca978?hp=507a68aa3c321b422f95b772611c878ce13952df>
> >
> > - Log -----------------------------------------------------------------
> > commit abf9167d3fff002ddaed53abb44d638387bca978
> > Author: David Mitchell <davem@iabyn.com>
> > Date: Mon Nov 22 19:18:49 2010 +0000
> >
> > Make PerlIO marginally reentrant
>
> > A t/io/eintr.t
>
> perlport says:
>
> =item select
>
> Only implemented on sockets. (Win32, VMS)
>
>
> which may explain why we haven't seen any Win32 smokes since Friday.
> On VMS, eintr.t hangs on the very first read(), which never gets
> interrupted by the alarm. I have some ideas about why that might be,
> but there's no easy fix so I'll probably have to put in a skip. If it
> were possible to implement the test in terms of a socket rather than a
> pipe, it might be more portable.
The only select that eintr.t is doing is a C<select FILEHANDLE>, i.e. to
change the default file handle. So that's probably a red herring, and any
hang is likely due to some different reason.
--
Nothing ventured, nothing lost.