Subject: Re: egcs
To: None <port-powerpc@netbsd.org>
From: Dan Jacobowitz <drow@false.org>
List: port-powerpc
Date: 08/05/1998 17:11:36
On Wed, Aug 05, 1998 at 01:18:06PM -0400, Todd Vierling wrote:
> On Wed, 5 Aug 1998, Dan Jacobowitz wrote:
>
> : > egcs 1.0.3 does not work on NetBSD/powerpc, period. egcs 1.0.3 is pretty
> : > broken in code generation on anything other than i386 and sparc, basically,
> ^^^^^^^^^^^^^^^
> : Not true, completely - I use egcs on powerpc-unknown-linux-gnu as a primary
> : compiler, and it screams.
>
> Doesn't matter how fast it runs - there are some types of constructs at
> higher optimization levels that will make the resulting code fall over. In
> short, you can use it to generate reliable code on powerpc if you never use
> -O2 or -O3.
>
Actually, the only problems I've had were a fit of unrecognizable insns
in -O2 and they were fixed in 1.1 mainline more or less :)
> : The 1.1 snapshots are a lot more reliable - although, unfortunately,
> : they STILL can't get through Lynx....
>
> "Internal compiler error" or "signal 11" (SIGSEGV)?
VM exhausted, actually. It pfffts on the large number of string
inlines in LYMainLoop.c at -O2. But then, so does gcc 2.7.2.x on
powerpc :)
Dan