RE: Powerpc signal handler with floating point registers

["Fraser, Doug" <D.Fraser%flarion.com@localhost>]

To answer my own question, apparently not.
Though for some odd reason it rings a bell. Hmmmmm. Prior release?
Do they get used for anything besides 'long long' operations?

-----Original Message-----
From: port-powerpc-owner%NetBSD.org@localhost
[mailto:port-powerpc-owner%NetBSD.org@localhost]On Behalf Of Fraser, Doug
Sent: Monday, February 28, 2005 7:24 PM
To: Matt Thomas; David Edelsohn
Cc: port-powerpc%NetBSD.org@localhost
Subject: RE: Powerpc signal handler with floating point registers 


Isn't there some kind of -fno-fp-moves or something like that????

-----Original Message-----
From: port-powerpc-owner%NetBSD.org@localhost
[mailto:port-powerpc-owner%NetBSD.org@localhost]On Behalf Of Matt Thomas
Sent: Monday, February 28, 2005 5:58 PM
To: David Edelsohn
Cc: port-powerpc%NetBSD.org@localhost
Subject: Re: Powerpc signal handler with floating point registers 


At 02:40 PM 2/28/2005, David Edelsohn wrote:
> >>>>> chris zhang writes:
>
>chris> Apparently, gcc generates code by using FP when
>chris> handling "long long" arithmatics. We couldn't find
>chris> compiler flags to turn off this yet.
>
>         -msoft-float.

That seems to be a rather big hammer to smash that fly.  Any other
way to make moves not use FP?  Allow floating point, but only if
the program is using floating point types.


-- 
Matt Thomas                     email: matt%3am-software.com@localhost
3am Software Foundry              www: http://3am-software.com/bio/matt/
Cupertino, CA              disclaimer: I avow all knowledge of this message.



===========================================================
This email may contain confidential and privileged material for the sole use
 of the intended recipient.  Any review or distribution by others is strictly
 prohibited.  If you are not the intended recipient please contact the sender
 and delete all copies.
===========================================================