This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: suggest gcc option -Wallall to add absolutely every -W* option
- To: amylaar at cygnus dot co dot uk, oliva at dcc dot unicamp dot br
- Subject: Re: suggest gcc option -Wallall to add absolutely every -W* option
- From: mrs at wrs dot com (Mike Stump)
- Date: Wed, 18 Nov 1998 12:09:18 -0800
- Cc: egcs at cygnus dot com, vonbrand at inf dot utfsm dot cl, wheeler at ipl dot rpi dot edu
> To: Joern Rennecke <amylaar@cygnus.co.uk>
> Cc: vonbrand@inf.utfsm.cl, wheeler@ipl.rpi.edu, egcs@cygnus.com
> From: Alexandre Oliva <oliva@dcc.unicamp.br>
> Date: 17 Nov 1998 15:55:45 -0200
> How about accepting a number after -W, so that one could ask the
> compiler to be more or less picky? -W0 would be no warnings, -W1
> the default, -W2 == -Wall, -W3 == -W -Wall, and -W99 would enable
> all possible warnings.
No, bad idea. With optimizations it makes sense because they all mean
about the same thing (make program go faster, or make program smaller
at the expense of speed), with the only tradoff being how much time
you want to spend doing it. With warnings, you talk about coding
conventions and other more artful things subject to more debate and
less clearcutness, so we wind up with some thing some people like and
other things other people like, and a number to select which implies
there is only one dimension possible (untrue). With optimizations,
there are only two dimensions (speed and size).