On Thu, 2008-10-02 at 12:32 -0500, Will Fiveash wrote: > On Thu, Oct 02, 2008 at 04:33:09PM +0200, Mark Phalan wrote: > > On Mon, 2008-09-29 at 12:38 -0400, Peter Shoults wrote: > > ... > > > > > > > > The only difference I see between MIT's copy_addrs.c and our > > > > copy_addrs.c is on line 32 - and its just a lint comment. > > > > Can you elaborate? > > > > > > > > > > > It looked like in other files, the memcpy was the code that was solaris > > > specific. > > > See usr/src/lib/gss_mechs/mech_krb5/krb5/krb/auth_con.c lines 26 & 40. > > > So, my question is if those are Solaris specific in auth_con.c, why not > > > here. > > > > I can add a "/* Solaris Kerberos */" comment to the lint comment but we > > haven't traditionally commented lint comments with "/* Solaris Kerberos > > */" comments. Perhaps we should? MIT code has no lint comments so its > > pretty clear that any lint comments are Solaris specific. > > Either we need give back the lint directive mods to MIT or write up a > resync guide with a comment about the lint directives so the resync'er > doesn't toss those out.
Agreed. Of course it would be best to have them upstream. But we may need that guide until that happens. I've been using nightly(1) to determine what lint fixes we need and which ones we don't. That reasoning may be flawed. -M
