develooper Front page | perl.perl5.porters | Postings from March 2013

Re: [perl #116989] S_croak_memory_wrap breaks gcc warning flagsdetection

From:
Andy Dougherty
Date:
March 27, 2013 20:11
Subject:
Re: [perl #116989] S_croak_memory_wrap breaks gcc warning flagsdetection
Message ID:
alpine.DEB.2.02.1303271606000.15645@fractal.phys.lafayette.edu
On Tue, 26 Mar 2013, Tony Cook via RT wrote:

> On Mon Mar 25 02:12:34 2013, nicholas wrote:
> > On Mon, Mar 25, 2013 at 09:36:05AM +0100, Dagfinn Ilmari Manns???ker
> > wrote:
> > > Darin McBride (via RT) <perlbug-followup@perl.org> writes:
> > >
> > > > This seems to be a regression:
> > > >
> > > > $ perl5.16.3 -MTime::HiRes -le 'print Time::HiRes::clock_gettime
> > (Time::HiRes::CLOCK_MONOTONIC ())'
> > > > 1597643.86352558
> > > > $ perl5.17.10 -MTime::HiRes -le 'print Time::HiRes::clock_gettime
> > (Time::HiRes::CLOCK_MONOTONIC ())'
> > > > Time::HiRes::clock_gettime(): unimplemented in this platform at -e
> > line 1.
> > >
> > > This is caused by the same header problem as:
> > >
> > > https://rt.perl.org/rt3/Public/Bug/Display.html?id=117319
> > > https://rt.perl.org/rt3/Public/Bug/Display.html?id=116989
> > 
> > I have merged all these tickets.
> > 
> > I haven't yet been hit with inspiration on how to solve the problem
> > whilst
> > preserving the intent of the changes which inadvertently caused it.
> 
> Making croak_memory_map() into an exported API function would retain
> most of the savings while reducing the pain, I think.

This is the tact I tried in commit f8fd8bfff8eb57156a0ab858d4c185701258b0e3
in branch remotes/origin/smoke-me/andyd/croak_memory_wrap.

I'd welcome feedback.  I'm no longer sure it ends up saving anything
compared to the original Perl_croak_nocontext("%s", PL_memory_wrap),
but this seemed like the best way forward for now.

-- 
    Andy Dougherty		doughera@lafayette.edu



nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About