Moriah Waterland writes:
> > usr/src/cmd/svr4pkg/hdrs/install.h
> >
> > Not sure why this is a new file.
> > usr/src/cmd/svr4pkg/hdrs/libinst.h
> > Not sure why new.
> This file is actually not new, but when I was creating the webrev of
> changes I forgot to grab these files from the Legacy Nevada Install
> gate. The only difference between the files that I updated and the
> original is the sccs versioning :
> o install.h, I removed:
> > #pragma ident "@(#)install.h 1.24 06/02/27 SMI"
>
> o libinst.h, I removed:
> > #pragma ident "@(#)libinst.h 1.56 08/01/18 SMI"
OK. (If that's really the only change, I would have reverted it. But
that change is right otherwise, so it's your call.)
> > usr/src/cmd/svr4pkg/pkgadm/Makefile
> >
> > 48-51,53-55: lots of duplicated stuff across makefiles; consider
> > creating "usr/src/cmd/svr4pkg/Makefile.com" to hold it.
> Done. I created the following Makefiles:
> o usr/src/cmd/svr4pkg/Makefile.svr4pkg
> o usr/src/cmd/svr4pkg/Makefile.svr4pkg.targ
OK; thanks.
> > usr/src/cmd/svr4pkg/pkgadm/main.c
[...]
> > 59: why was this made a global symbol?
> Honesty, I really don't know why it was originally declared that way.
> It doesn't seem to be necessary. With struct cmd defined as static, gcc
> pukes a warning and cc chokes on compilation.
Oh -- of course. My mistake. It's not a symbol at all; it's a type
declaration, and having a storage qualifier on a type declaration is
purely nonsense. Your change is exactly right.
> Note 0.
> =======
> I don't know what the hell to do about this situation. gcc won't
> compile without the returns.
> default:
> usage();
> /*NOTREACHED*/
> /*
> * Although usage() calls a noreturn function,
> * needed to add return (1); so that main() would
> * pass compilation checks. The statement below
> * should never be executed.
> */
> return (1);
>
> I added the following line at the end of the definition of main() :
> /* LINTED: no return */
As long as it's handled consistently compiles/lints cleanly, I'm
happy.
> Note 4.
> ======
>
> Change was putback to this source after code should have been frozen.
>
> D 1.60 08/12/17 04:46:38 jk217608 68 67 00010/00004/01250
> MRs:
> COMMENTS:
> 6776920 removef doesn't cope with duplicate path on command line
>
>
>
> I went back and reviewed all the other putbacks since I last pulled
> the code and verified that I am not missing anything.
Great; thanks!
> Note 5.
> ======
> Added as part of
> D 1.16 94/10/21 Added functions to suppoort faster install.
>
>
> 719
> 720 int
> 721 is_an_inst_root(void)
> 722 {
> 723 return (install_root_exists);
> 724 }
> 822 client_offset = (is_an_inst_root) ? install_root_len : 0;
Ouch.
> fixpath.c: In function `eval_path':
> fixpath.c:822: warning: the address of `is_an_inst_root', will always
> evaluate as `true'
Indeed it does.
> Note 6.
> --------
> I fixed these problems. I had introduced the "" regression when I was
> originally working on just getting this code to build in the ON gate and
> missed this in my cleanup.
OK.
--
James Carlson, Solaris Networking <james.d.carlson at sun.com>
Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
