develooper Front page | perl.perl5.porters | Postings from December 2017

[perl #120758] [5.19.6] Failing M::BI tests, ppc, -Duselongdouble

From:
Sisyphus via RT
Date:
December 11, 2017 11:38
Subject:
[perl #120758] [5.19.6] Failing M::BI tests, ppc, -Duselongdouble
Message ID:
rt-4.0.24-6291-1512992310-760.120758-15-0@perl.org
On Wed, 11 Dec 2013 02:19:03 -0800, sisyphus wrote:

> Unpacking
> the long double reveals that perl assigns:
> 
> 5ab7151b377c247e5707b80b00474380
> instead of the correct:
> 5ab7151b377c247e5707b80b00474440

Actually, I don't think 5ab7151b377c247e5707b80b00474440 is correct either.
It's the value that C's strtold() assigns - but the mpfr library tells me that the correct value (for rounding to nearest, ties to even) is 5ab7151b377c247e5707b80b0047445d.

Having spent some time tonight playing around with various ad-hoc values, it seems that strtold() is little more reliable than perl when it comes to assigning these values.

Cheers,
Rob


---
via perlbug:  queue: perl5 status: open
https://rt.perl.org/Ticket/Display.html?id=120758



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