Re: Proposed libf2c patch

[Toon Moene <toon at moene dot indiv dot nluug dot nl>]

Jeffrey A Law wrote:

[ Bug report on complex division by John W. Eaton, author of MathLab
  work-alike Octave ]

>> This problem is due to a bug in the complex division routine used by
>> g77.  The following patch will make it behave the same as the complex
>> division code used by g++ (which was adapted from the libf2c code).

> I have no idea if this patch is reasonable or not.  Can one of the 
> Fortran folks take appropriate action for this patch.

It has been obsolescened (or is that obsoleted ?) by the inlining of
complex division in optabs.c which happens :-) to DTRT.  g77 will now
(i.e. from 2.95 onwards) only generate a call to this routine if -Os is
specified.  Of course, that means that we still have to think about
whether to apply this patch, just to make complex division behave the
same for all -Ox flags.
But, AFAIK, Octave is not built using the -Os flag.

> [ If it's already been dealt with, chide me for falling way behind on 
>   patches during the gcc-2.95 release cycle :-) ]

I'm more worried you're falling behind on sleep ;-)  The date on the
mail I'm replying to reads:  Sun, 29 Aug 1999 05:05:29 -0600

Hey - it's weekend; go lie on your back and do nothing - like I do !

Cheers,

-- 
Toon Moene (toon@moene.indiv.nluug.nl)
Saturnushof 14, 3738 XG  Maartensdijk, The Netherlands
Phone: +31 346 214290; Fax: +31 346 214286
GNU Fortran: http://world.std.com/~burley/g77.html