Date: Wed, 15 Sep 1999 04:13:11 +0200 From: Marc Lehmann To: pgcc AT delorie DOT com Subject: Re: -fsoftware-pipe Message-ID: <19990915041311.A3983@cerebro.laendle> Mail-Followup-To: pgcc AT delorie DOT com References: <19990915002056 DOT A3481 AT win DOT tue DOT nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <19990915002056.A3481@win.tue.nl>; from Ronald de Man on Wed, Sep 15, 1999 at 12:20:57AM +0200 X-Operating-System: Linux version 2.2.12 (root AT cerebro) (gcc driver version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release) executing gcc version 2.7.2.3) Sender: Marc Lehmann Reply-To: pgcc AT delorie DOT com X-Mailing-List: pgcc AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Wed, Sep 15, 1999 at 12:20:57AM +0200, Ronald de Man wrote: > The option that seems to make pgcc-2.95.1 crash on almost every source > file appears to be -fsoftware-pipe. Compiling with -O6 -fno-software-pipe Maybe it makes more snese when I tell you that this is one of the most often bugfixed optimization in pgcc ever. I remember many nights of debug sessions.. *sigh* > I don't get internal compiler errors anymore on my program. > (However, -O2 -Os -fomit-frame-pointer generates much faster code > on a P233 MMX, at least in the case of this program.) There is rumour that -Os is indeed faster than -O2 in gcc (and thus in pgcc). The new snapshots might fix that, though. > Would it be helpful if I come up with a, say, 10 line program > that exhibits the error? I could probably extract something like > that from my source. That would be great! A small test case is the first step I'd have to do anyway. -- -----==- | ----==-- _ | ---==---(_)__ __ ____ __ Marc Lehmann +-- --==---/ / _ \/ // /\ \/ / pcg AT goof DOT com |e| -=====/_/_//_/\_,_/ /_/\_\ XX11-RIPE --+ The choice of a GNU generation | |