On Wed, Feb 27, 2002 at 03:15:09PM -0800, Peter Wemm wrote: > If there are files that are too hard to fix, or vendor files, or the fix > isn't clear, we should use the nowerror conf/files* flags. > > It is important that we stop new warnings turning up when the compile > output is so damn large that it hides things.
I definatly agree. The warnings in gif(4) were mostly lame and should have been fixed, but no one bothered. > I will do a pass over things now and see what I can do. Please make sure to test the !SMP case. The following slipped through the initial sweep: Index: kern_sig.c =================================================================== RCS file: /usr/cvs/src/sys/kern/kern_sig.c,v retrieving revision 1.149 diff -u -r1.149 kern_sig.c --- kern_sig.c 23 Feb 2002 11:12:54 -0000 1.149 +++ kern_sig.c 26 Feb 2002 19:37:00 -0000 @@ -1233,7 +1233,9 @@ register int prop; register sig_t action; struct thread *td; +#ifdef SMP struct ksegrp *kg; +#endif KASSERT(_SIG_VALID(sig), ("psignal(): invalid signal %d\n", sig)); -- Brooks -- Any statement of the form "X is the one, true Y" is FALSE. PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4
Description: PGP signature