Re: Error building if_fatm.c

On Tue, Dec 23, 2003 at 07:11:53PM -0600, Ryan Sommers wrote:
> On Tue, 2003-12-23 at 18:50, Craig Rodrigues wrote:
> > /usr/src/sys/dev/fatm/if_fatm.c: In function `fatm_open_vcc':
> > /usr/src/sys/dev/fatm/if_fatm.c:2285: warning: comparison between pointer and integer
> > *** Error code 1
> 
> This is just a warning, it isn't what is causing the error. It looks
> like you are compiling with the -j option and this will sometimes hide
> the real cause of the failure. Try compiling without -j and see what the
> error condition is it is hitting.
> 
> PS As for the redefinition of NULL you could change it to that, or just
> cast NULL to an int. 


Wrong.  I am not compiling with -j.

I am simply doing: make buildkernel KERNCONF=MYKERNEL1

If you look at my previous message, you will see that -Werror
is being passed to the compiler...this is causing the warning to
be flagged as an error.  (NOTE: My /etc/make.conf file is empty, so
this behavior is obtained from the default FreeBSD kernel makefile rules).

I think my patch to if_fatm.c fixes things (since it unbreaks things for me)
but I just wanted to clarify things on this mailing list.

I think this problem emerged due to the following change
to /usr/src/sys/sys/_null.h :

==========================================================================
 Revision 1.3 / (download) - annotate - [select for diffs], Tue Dec 23
   02:34:25 2003 UTC (22 hours, 54 minutes ago) by peter
   Branch: MAIN 
   Changes since 1.2: +4 -0 lines
   Diff to previous 1.2 (colored)
Don peril sensitive sunglasses and set NULL to an actual pointer type,
but *only* for the kernel.  We can do this because the kernel is not a
standard C application environment.  This would have stopped the recent
mtx_* arg NULL/MTX_DEF mixups from going unnoticed for so long.
==========================================================================


-- 
Craig Rodrigues        
http://crodrigues.org
rodrigc_at_crodrigues.org