This is the mail archive of the cygwin mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Perl bug?


Corinna Vinschen wrote:
> On Jan 27 23:17, Reini Urban wrote:
>> Václav Haisman schrieb:
>> >If I were you I would report it as a bug to their bug tracker.
>>
>> It's no bug, it's a perl feature,
>
> Uh, right, a *feature* ;)
>
>>  and often defended.
>> Even dll's are not unloaded.
>>
>> If you want to free it, free it explicitly with "undef"
>> or with lexicals ("my") go out of scope.

Doesn't help in this case, Reini.

>> Same with PHP and python btw. Only GC languages like lisp, ml and its
>> derivates have a proper GC.
>> The perl GC they are talking about only "garbage collects" cyclic
>> referenced objects on final destruction, to enable proper free() of
>> externals.
>
> Thanks for the info.  It's interesting to know.  What I still don't get,
> however, is the fact that the same statement does not waste memory on
> the x86 Linux Perl 5.8.5, but does on the x86 Cygwin Perl 5.8.7 and the
> x86_64 Linux 5.8.8.  So it has been introduced only in later versions?
> And why is it defended?  It doesn't seem to make sense, rather on the
> contrary.

This behaviour hasn't changed that I know of.  I verified that a 5.8.5
cygwin perl behaves the same way.  Could you confirm your x86 Linux
Perl 5.8.5 results again, and send me (privately if you wish) the
output of: perl -MConfig=config_sh -we'print config_sh' from that
perl?


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]