Re: [Lxr-dev] Re: /bin/true as a zombie

[Jan-Benedict G. <jb...@lu...>]

On Tue, Jul 31, 2001 at 10:57:30AM +0200, Peder O. Klingenberg wrote:
> Jan-Benedict Glaw <jb...@lu...> writes:
> and less obvious, but also less broken way (for CVS trees).  Could you
> try --allversions with that version of genxref?  It's available from
> <http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/~checkout~/lxr/lxr/genxref?rev=1.18&content-type=text/plain>,
> and I think thats the same version as in the old ping CVS tree.

- I've reverted the change to ./lib/LXR/CVS.pm
- Checked out the above version og genxref
- Result:
	- This version also *only* tries to index head
	- It seems to handle some things wrong:

[...]
### /lib/wait.c  
--- /lib/wait.c head 1.2
/lib/wait.c was already referenced
### /lib/write.c  
--- /lib/write.c head 1.2
/lib/write.c was already referenced
### /mm/  
### /mm/RCS/  
### /mm/Makefile  
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
co: /home/data/CVSROOT/linux//mm/Makefile,v: branch number 0 too low
### /mm/memory.c  
--- /mm/memory.c head 1.4
/mm/memory.c was already referenced
### /mm/page.s  
[...]

> > disconnect(DBI::db=HASH(0x83a7a30)) invalidates 1 active statement. Either destroy statement handles or call finish on them before disconnecting. at lib//LXR/Index/Postgres.pm line 272.
> 
> This is an unrelated, but nevertheless irritating bug.

Which is also triggered with the older CVS version of gnexref you mentioned
above.

I've also tried the old copy of ping's genxref version. It triggers the
"branch number 0 too low" bug also

> > Yes. Another copy of LXR (that one from the CVS tree on cvs.ping.uio.no)
> > running on the same source tree(but from another directory with its
> > own config etc.) behaves just fine with it:-)
> 
> weird.

If it could help you, I'd package my complete LXR installation (
functional and non-functional as well) and send it to you...

> > It doesn't exit ungracefully. Just like above (with cancelling one
> > active statement by closing the DB handle too early). My CVS
tree is also not that large (~1.5MB) uncompressed...

> And I don't think genxref explains why you're having trouble
> browsing, so that must be a separate issue.

I'm unsure. Those oloder versions show up this "branch 0 too low"
thing I could not associate to any usage of CVS I ever had when
accessing it by hand...

MFG, JBG