Subject: Re: sun-lamp CVS commits, -O3 on libkern
To: Rob Healey <>
From: None <>
List: current-users
Date: 09/16/1994 21:47:38
>Forwarded message:
>> deraadt
>> Thu Sep 15 18:27:59 PDT 1994
>> Update of /b/source/CVS/src/sys/lib/libkern
>> In directory
>> Modified Files:
>> 	Makefile 
>> Log Message:
>> compile -O3
>	I'm I safe in assuming this particular change was fully tested on
>	ALL architectures before going in? I ask because I know the gcc 2.5
>	has a nasty talent to wack up m68k code generation at higher levels
>	of optimization. Since m68k got royally burned in the ptrace()
>	stuff it would be nice to verify changes on this order on all
>	major architectures, sparc, m68k, mips and x86.

Well, unless gcc gets upgraded to 2.6.0, anything above -O2 is meaningless
anyways. -O2 and higher before 2.6.0 meant the same thing. gcc 2.6.0
defined -O3 to include -finline-functions now however. This was 
probably a good idea for them to wait till then anyways since I've had
code break hard with inline stuff before that version....