On Sat, Feb 26, 2005 at 01:48:09PM +0000, Richard Earnshaw wrote:

> On Sun, 27 Feb 2005 00:39:18 +1100, Simon Burge wrote:
> > What do folks think of making assembler warnings fatal?  On mips at
> > least we've hit a few new warnings that appeared with binutils 2.15 and
> > have been fixed, but these would have been caught much easier if we
> > enabled the --fatal-warnings flag to the assembler.
> > 
> > Just adding -Wa,--fatal-warnings to AFLAGS in bsd.sys.mk didn't help.
> > Adding that to AFLAGS in sys/conf/conf/Makefile.kern.inc works, but this
> > probably doesn't catch LKMs etc?  Where would be the best place to add
> > this?  Also, it probably should be enabled for userland assembly code
> > too.
> > 
> > Any arguments against not enabling --fatal-warnings ?
> > 
> 
> Does that apply to assembly warnings from .c files, or only from .S/.s?
> 
> If the former, then it will break ARM ports where an inline ASM generates 
> such a warning (though it happens to be safe).

If we only use this option on AFLAGS it wont affect .c files.

Do you have an example of the ARM warning?  I'm curious...

Simon.
--
Simon Burge                                   <simonb@wasabisystems.com>
NetBSD Development, Support and Service:   http://www.wasabisystems.com/