Re: nffactor() bug

Bill Allombert on Mon, 17 Nov 2003 16:19:54 +0100

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: nffactor() bug

To: Developers Pari <pari-dev@list.cr.yp.to>
Subject: Re: nffactor() bug
From: Bill Allombert <allomber@math.u-bordeaux.fr>
Date: Mon, 17 Nov 2003 16:16:34 +0100
Delivery-date: Mon, 17 Nov 2003 16:19:54 +0100
In-reply-to: <20031114173825.GQ30106@txc.com>
Mail-followup-to: Developers Pari <pari-dev@list.cr.yp.to>
Mailing-list: contact pari-dev-help@list.cr.yp.to; run by ezmlm
References: <20031114173825.GQ30106@txc.com>
User-agent: Mutt/1.3.28i

> When I wanted to tried building a debugging binary on the same platform (via Configure -g) I got this compilation error: 
> 
> ../src/kernel/none/mp.c: In function `muluu':
> ../src/kernel/none/mp.c:620: `asm' operand requires impossible reload

As far as I can tell this bug is due to a register starvation in gcc.
Compiling this file with -O fix the problem (probably the optimised code
use less registers).

It is not clear whether it a bug or a limitation in gcc or a bug in
PARI.

The problem with the functions returning `int' on 64bit is not solved
yet. Maybe it is time do implement something ?  

Cheers,
Bill.

Follow-Ups:
- Re: nffactor() bug
  - From: Bill Allombert <allomber@math.u-bordeaux.fr>

References:
- nffactor() bug
  - From: Igor Schein <igor@txc.com>

Prev by Date: Re: user install
Next by Date: Re: nffactor() bug
Previous by thread: Re: nffactor() bug
Next by thread: Re: nffactor() bug
Index(es):
- Date
- Thread