[Pkg-octave-devel] Bug#871214: [pkg-octave/master] d/rules: Pass -ffloat-store option to g++ on i386 architecture

[Rafael Laboissiere]

tag 871214 pending
thanks

Date: Sat Aug 12 06:30:37 2017 -0300
Author: Rafael Laboissiere <rafael at debian.org>
Commit ID: d383f06a10084cf511f385f46d569071fb2feef1
Commit URL: https://anonscm.debian.org/cgit/pkg-octave/octave-image.git;a=commitdiff;h=d383f06a10084cf511f385f46d569071fb2feef1
Patch URL: https://anonscm.debian.org/cgit/pkg-octave/octave-image.git;a=commitdiff_plain;h=d383f06a10084cf511f385f46d569071fb2feef1

    d/rules: Pass -ffloat-store option to g++ on i386 architecture

    This will ensure the precise definition of IEEE floating point, what
    will, hopefully, avoid the endless loop (and the consequent FTBFS) in
    the unit test for bwdist.

    Gbp-Dch: Full
    Closes: #871214
    Thanks: Adrian Bunk for the diagnosis of the problem and the suggestion of the fix