Date: Fri, 13 Jan 1995 21:44:32 -0700 (MST)
From: Calimath / Sliced Bread <sl5h9 AT cc DOT usu DOT edu>
Subject: Re: gcc = gcc -O2 ?
To: djgpp AT sun DOT soe DOT clarkson DOT edu



On Sat, 14 Jan 1995, Stephen Turnbull wrote:

>       That sounds like a good idea.  Are there very many times when you 
>    don't want to optimize, after all?
> 
>    ====================== -Jon (SL5H9 AT cc DOT usu DOT edu) =========================
> 
> (1) when you're debugging and you want the debugger to know where in
> the code you are
> (2) when you're porting to the same hardware but different OS and the
> objects can be the same and you'd like them to be the same so that you
> know it's a OS problem (or a problem that only shows up when the OS
> changes, not at all the same thing)
> (3) when you're trying to write a portable makefile and the hardware
> is buggy (Pentium, did someone say Pentium?) and supports assorted
> OSes.
>     (These are not necessarily reasons to avoid optimization; they are
> reasons for DJGPP to have the same default behavior as other GCCs.)
>     I don't have time to think of more, is 3 reasons enough?

   Hmmm...you could, of course, use -O0 on those.  Then again, these 
cases probably occur more often then the times that you want 
optimization.  And I do like the point about having djgpp behave like 
other gcc's.  I guess my vote is for gcc == gcc -O0 as well.

====================== -Jon (SL5H9 AT cc DOT usu DOT edu) =============================
The optimist sees a glass that's half full.
The pessimist sees a glass that's half empty.
An engineer sees a glass that's twice as big as it needs to be!
================== http://www.declab.usu.edu:8080/~sl5h9/ ==================