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

Re: [perl #118175] threaded perl-5.18.0 fails re/pat_rt_report_thr.ton i586

Thread Previous
From:
Nicholas Clark
Date:
June 7, 2013 15:38
Subject:
Re: [perl #118175] threaded perl-5.18.0 fails re/pat_rt_report_thr.ton i586
Message ID:
20130607153803.GL3729@plum.flirble.org
On Wed, May 29, 2013 at 05:22:05PM -0700, Tony Cook via RT wrote:
> On Tue May 28 03:06:14 2013, tonyc wrote:
> > Our test regexp is UTF-8, if scan were under 0x80000000 this code
> > would be sane, unfortunately, the string in my test cases we being
> > allocated at very high address, from debugging code I added:
> 
> Last night I spent some time trying to figure out if I can run the
> entire test suite allocating above the 2G line, but I couldn't see a
> simple way to do it.
> 
> The glibc supplied with Mageia appears to switch to mmap() for allocation
> (instead of brk()) inside threads, and mmap() is allocating above the
> 2G line.
> 
> The only other way I can see to do that would be a custom malloc(), but
> I'm not sure I want to work that hard.

I suspect that Perl's malloc can be (ab)used so that it always allocates
"more core" by using mmap. But in turn I don't know if mmap() can be told
to allocate address above 2G (distinct from "allocate at this address).
And I agree, I'm not sure if it's worth it.

Nicholas Clark

Thread Previous


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