Skip Menu |

This queue is for tickets about the EV CPAN distribution.

Report information
The Basics
Id: 35132
Status: rejected
Priority: 0/
Queue: EV
People
Owner: Nobody in particular
Requestors: vpit [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: Unimportant
Broken in: (no value)
Fixed in: (no value)

History Show all quoted text
  Wed Apr 16 15:53:46 2008 vpit [...] cpan.org - Ticket created
Subject: Undefined symbol ev_stat64_start
Cf http://www.nntp.perl.org/group/perl.cpan.testers/2008/04/msg1272851.html I don't know if it's caused by ICC or the system headers, but stat seems to be #define'd to stat64, causing RESET() to expand to an unknown symbol : EV.xs(1049): warning #266: function "ev_stat64_stop" declared implicitly RESET (stat, w, (w, SvPVbyte_nolen (w->fh), interval)); ^ Which logically leads to tests failures of this kind : Can't load '/home/smoke/.cpan/build/EV-3.2-PergV1/blib/arch/auto/EV/EV.so' for module EV: /home/smoke/.cpan/build/EV-3.2-PergV1/blib/arch/auto/EV/EV.so: undefined symbol: ev_stat64_start at /home/smoke/perl/builds/5.10.0/lib/5.10.0/x86_64-linux-thread-multi/DynaLoader.pm line 203 This can be solved just by #undef-ining stat around the top of EV.xs. Regards, Vincent Pit (VPIT)
  Wed Apr 16 17:08:38 2008 schmorp [...] schmorp.de - Correspondence added
CC: undisclosed-recipients: ;
Subject: Re: [rt.cpan.org #35132] Undefined symbol ev_stat64_start
Date: Wed, 16 Apr 2008 23:08:18 +0200
To: Vincent Pit via RT <bug-EV [...] rt.cpan.org>
From: Marc Lehmann <schmorp [...] schmorp.de>
On Wed, Apr 16, 2008 at 03:53:48PM -0400, Vincent Pit via RT <bug-EV@rt.cpan.org> wrote: Show quoted text
> > Wed Apr 16 15:53:46 2008: Request 35132 was acted upon. > Transaction: Ticket created by VPIT > Queue: EV
Please close your ticket and resubmit it via the documented contact address - rt.cpan.org is not supported but unfortunately cannot be switched off and continues to spam people :/ It is by pure chance that I saw your report. Show quoted text
> EV.xs(1049): warning #266: function "ev_stat64_stop" declared implicitly > RESET (stat, w, (w, SvPVbyte_nolen (w->fh), interval));
EV requires a working C or C++ compiler, which ICC is apparently not. stat is a macro argument and therefore not expanded (c99 6.10.3), and the replacement text contains no stat identifier (because ## suppresses expansion further). I suggest using a working C compiler such as gcc, which has a working preprocessor, or maybe ICC has a switch that converts it into a C compiler that you have to specify. Show quoted text
> This can be solved just by #undef-ining stat around the top of EV.xs.
The problem is not in EV.xs, and therefore cannot be solved there. The problem is that you are not using a C compiler to compile EV.c, which is a requirement. If you really want to use ICC you should report this to your vendor so the compiler can be fixed. Thanks for bringing this to my attention, though! -- The choice of a Deliantra, the free code+content MORPG -----==- _GNU_ http://www.deliantra.net ----==-- _ generation ---==---(_)__ __ ____ __ Marc Lehmann --==---/ / _ \/ // /\ \/ / pcg@goof.com -=====/_/_//_/\_,_/ /_/\_\
  Wed Apr 16 17:08:41 2008 The RT System itself - Status changed from 'new' to 'open'
  Wed Apr 16 17:37:55 2008 vpit [...] cpan.org - Correspondence added
Subject: Re: [rt.cpan.org #35132] Undefined symbol ev_stat64_start
Date: Wed, 16 Apr 2008 23:37:26 +0200
To: bug-EV [...] rt.cpan.org
From: Vincent Pit <perl [...] profvince.com>
Show quoted text
> <URL: http://rt.cpan.org/Ticket/Display.html?id=35132 > > > On Wed, Apr 16, 2008 at 03:53:48PM -0400, Vincent Pit via RT <bug-EV@rt.cpan.org> wrote: >
>> Wed Apr 16 15:53:46 2008: Request 35132 was acted upon. >> Transaction: Ticket created by VPIT >> Queue: EV >>
> > Please close your ticket and resubmit it via the documented contact > address - rt.cpan.org is not supported but unfortunately cannot be > switched off and continues to spam people :/ > > It is by pure chance that I saw your report. > >
>> EV.xs(1049): warning #266: function "ev_stat64_stop" declared implicitly >> RESET (stat, w, (w, SvPVbyte_nolen (w->fh), interval)); >>
> > EV requires a working C or C++ compiler, which ICC is apparently not.
I think you may want to tell Intel that their compiler is moot. Or also tell a vast majority of scientists that solely use it that they're wrong. Not sure how they'll take it though. Show quoted text
> stat is a macro argument and therefore not expanded (c99 6.10.3), and the replacement > text contains no stat identifier (because ## suppresses expansion further). >
Ok. Show quoted text
> I suggest using a working C compiler such as gcc, which has a working > preprocessor, or maybe ICC has a switch that converts it into a C compiler > that you have to specify. >
So gcc is a conformant C99 compiler now ? I'm glad to hear it. Show quoted text
>> This can be solved just by #undef-ining stat around the top of EV.xs. >>
> > The problem is not in EV.xs, and therefore cannot be solved there. The > problem is that you are not using a C compiler to compile EV.c, which is a > requirement. >
Just ignore the report then. I was just using Intel CC for the purpose of smoking. We won't get it fixed with that kind of discussion. Sorry for the nuisance. Show quoted text
> If you really want to use ICC you should report this to your vendor so the > compiler can be fixed. > > Thanks for bringing this to my attention, though! > >
  Wed Apr 16 17:57:00 2008 vpit [...] cpan.org - Status changed from 'open' to 'resolved'
  Wed Apr 16 17:57:01 2008 vpit [...] cpan.org - Severity Unimportant added
  Thu Apr 17 09:23:56 2008 schmorp [...] schmorp.de - Correspondence added
CC: undisclosed-recipients: ;
Subject: Re: [rt.cpan.org #35132] Undefined symbol ev_stat64_start
Date: Thu, 17 Apr 2008 15:23:41 +0200
To: Vincent Pit via RT <bug-EV [...] rt.cpan.org>
From: Marc Lehmann <schmorp [...] schmorp.de>
On Wed, Apr 16, 2008 at 05:37:58PM -0400, Vincent Pit via RT <bug-EV@rt.cpan.org> wrote: Show quoted text
> I think you may want to tell Intel that their compiler is moot. Or also
Why should I? I don't use it. Show quoted text
> > I suggest using a working C compiler such as gcc, which has a working > > preprocessor, or maybe ICC has a switch that converts it into a C compiler > > that you have to specify. > >
> So gcc is a conformant C99 compiler now ? I'm glad to hear it.
Nobody claimed that, but it is definitely a working c compiler, unlike icc, which apparently can't even get the relatively simple preprocessor right, Show quoted text
> Just ignore the report then. I was just using Intel CC for the purpose > of smoking. We won't get it fixed with that kind of discussion.
Well, maybe you shouldn't use ICC then, you will create a lot of bogus bug reports with such a marginal compiler. -- The choice of a Deliantra, the free code+content MORPG -----==- _GNU_ http://www.deliantra.net ----==-- _ generation ---==---(_)__ __ ____ __ Marc Lehmann --==---/ / _ \/ // /\ \/ / pcg@goof.com -=====/_/_//_/\_,_/ /_/\_\
  Thu Apr 17 09:23:58 2008 The RT System itself - Status changed from 'resolved' to 'open'
  Thu Apr 17 09:25:33 2008 MLEHMANN [...] cpan.org - Status changed from 'open' to 'rejected'