Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 01 Oct 2007 11:00:36 +0900 (JST)
From:      Maho NAKATA <chat95@mac.com>
To:        tomdean@speakeasy.org
Cc:        freebsd-current@freebsd.org, sgk@troutmask.apl.washington.edu
Subject:   Re: Compiler Problems?
Message-ID:  <20071001.110036.102571438.chat95@mac.com>
In-Reply-To: <200709160456.l8G4udLl023614@dv6000.tddhome>
References:  <200709160102.l8G12bAl001481@dv6000.tddhome> <20070916020726.GB6447@troutmask.apl.washington.edu> <200709160456.l8G4udLl023614@dv6000.tddhome>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi Thomas and Steve,

Thanks for your clarification. I rewrote this port completely and
update to 3.1.1. Now at least dlamch/slamch are build with -O0.
Hope now it works well. I haven't tried yet with g77 and may not work with it.

% cd /usr/ports/math/lapack
% rm -rf work ; make > & log &
% grep slamch.f log
gfortran42 -O0 -c slamch.f -o slamch.o
gfortran42 -O0 -fpic -c slamch.f -o slamch.o
gfortran42 -O0 -pg -c slamch.f -o slamch.o

From: "Thomas D. Dean" <tomdean@speakeasy.org>
Subject: Re: Compiler Problems?
Date: Sat, 15 Sep 2007 21:56:39 -0700 (PDT)

> Aha!  The problem may be due to optimization.
> 
> http://www.netlib.org/lapack/faq.html#1.25
> 
>  1.25) Problems compiling dlamch.f?
> 
> The routine dlamch.f (and its dependent subroutines dlamc1, dlamc2,
> dlamc3, dlamc4, dlamc5) MUST be compiled without optimization. 

-- Nakata Maho (maho@FreeBSD.org)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20071001.110036.102571438.chat95>