Re: cvs problems on 1998 Sep 7 current

["William Allen Simpson" <wsimpson_(_at_)_greendragon_(_dot_)_com>]

> From: Tobias Weingartner <weingart_(_at_)_wolfram_(_dot_)_com>
> > Some folks have moved around a fair number of files and directories, and
> > make obj failed.  I had to rm -rf /usr/obj and run make obj again.
>
> A simple 'make obj' should have "fixed" things for you.  Granted, you would
> have stale objects in your obj tree...
>
Yes, it should.  But it didn't.  That's why I'm reporting it.  There is
a bug in error handling in the "make obj", that occurs whenever any part
of the tree is moved (as happened with ppp and e2fsprogs).  It was
repeatable and verifiable (I tried several times before I gave up and
just removed the whole thing).

Any chance that someone who understands make better than I will take a
look and fix it?

Maybe the only fix is to add rm -rf /usr/obj in the /usr/src makefile?
After all, the first thing that make build does is clean out all the obj.


> > The update is failing in one section of the tree, although the make
> > build went OK (as far as I can tell).
> >
> > [...]
> > cvs server: Updating gnu/usr.sbin/e2fsprogs/lib/et
> > [...lots and lots more of the same...]
>
> Unless you have local changes in that part of the tree, the best way (that I
> know of) to update that portion, is to remove it, and run the update again.
>
I'll try that.  Of course, I don't have any local changes.  We'll see
whether the diff3 stops segment faulting.  But it would have been nicer
for someone to figure out _why_ diff3 is segment faulting....

WSimpson_(_at_)_UMich_(_dot_)_edu
    Key fingerprint =  17 40 5E 67 15 6F 31 26  DD 0D B9 9B 6A 15 2C 32