spoolss talloc problem

[tridge at samba.org]

Hi Simo,

 > Cool, and I have to say: I was so close ... I knew the sibling list was
 > faulty, too bad I couldn't create a reproducer.
 > Were you able to create a talloc test that reproduces the issue ?

not really ...

What I realised yesterday was that the spoolss probem was really just
a new varient of the bug that Rusty had previously found, and that was
in our test suite as test_rusty(). 

So what I did was to remove the 'fix' I had put in for that bug, and
then looked at it carefully again. The fix I'd previously used for
that bug was really just a workaround for the sibling list being
corrupt, rather than actually preventing the corruption. So when I
then put in the right fix for the test_rusty() case I also found that
the spoolss case was fixed.

It would still be nice to create a test case for the exact situation
you came across, but I don't think we have to wait for that to push
the fix. The fix is also much more obviously correct than the
workaround we previously had for test_rusty().

Cheers, Tridge