new database code

[Andrew Tridgell]

I've just committed the first of the big changes in the head (Samba
3.0) branch to do with the new database code.

For those who don't remember, the database code will be replacing a
bunch of shared data structures and lock files that have been used in
Samba previously. This should greatly simplify the internal structures
in Samba and also has some external benefits.

what has been done in CVS head so far is:

  - new tdb/tdb.c database module. I've written this to be like gdbm
    but to use shared memory and allow multiple simultaneous writers
    and locking. It also includes a test suite and a database tool

  - changed the existing connection databases (what was in STATUS..LCK
    and sharename.LCK) to use a single tdb database instead

  - changd all the existing share mode and oplock code that used
    sysv,mmap or lock files to instead use a tdb database. A nice side
    effect is that the storage for this stuff now takes up much less
    space and auto-expands as needed.

  - modified smbstatus and swat to know about the new databases.

there are still several more places in Samba that will be converted to
use tdb databases, especially the WINS code. Expect some more commits
soon.

Cheers, Tridge