Re: bug in lang/perl5 5.18.0?

[Amitai Schlair <schmonz%schmonz.com@localhost>]

On Sun, Jun 9, 2013 at 10:07 PM, Tom Spindler (moof)
<dogcow%babymeat.com@localhost> wrote:

> I suspect the next thing to do is to recompile with -O instead of -O2,
> especially for the Encode/ and cpan/Encode bits.

I recompiled with -O1 on i386 6.1. No failing tests started passing,
and one new Benchmark test failed. (For reference, the output:
<URL:http://www.netbsd.org/~schmonz/netbsd6.1-perl5.18.0-O1-tests.txt>)

Assuming NetBSD development is continuous and differentiable, and that
this isn't an MD problem, there must exist an earliest date-based
build on HEAD, somewhere between the 6.1 branch point and -current, on
which Perl 5.18.0's tests start passing.