Any more samldb or similar patches pending?
Andrew Bartlett
abartlet at samba.org
Wed Jun 12 00:31:29 MDT 2013
On Tue, 2013-06-11 at 20:51 +0200, Matthias Dieter Wallnöfer wrote:
> I remember that this was the issue:
> > @@ -775,7 +777,13 @@ int ltdb_modify_internal(struct ldb_module *module,
> > valued attributes or aren't
> > provided twice */
> > /* TODO: This is O(n^2) - replace with
> > more efficient check */
> > for (j = 0; j < el->num_values; j++) {
> > - if (ldb_msg_find_val(el2,
> > &el->values[j]) != NULL) {
> > + struct ldb_val *matched_val;
> > + ret =
> > ldb_msg_find_val_schema(ldb, a, el2,
> > + &el->values[j], &matched_val);
> > + if (ret != LDB_SUCCESS) {
> > + return ret;
> > + }
> > + if (matched_val != NULL) {
> ^^^ we compared "matched_val != &el->values[j]" rather than "!= NULL".
>
> Matthias
>
> Andrew Bartlett schrieb:
> > On Tue, 2013-06-11 at 10:00 +0200, Matthias Dieter Wallnöfer wrote:
> >> Hi Andrew,
> >>
> >> what do you think?
> >>
> >> Matthias
> >>
> >> Matthias Dieter Wallnöfer schrieb:
> >>> Hi Andrew,
> >>>
> >>> you could try reviewing my "ldb_schema" branch. I think it should be
> >>> working now.
> >>>
> > It looks good. What changed since we last had so much trouble with
> > this?
My remaining concern is how are linked attributes handled. I'm starting
to think that it isn't a concern, because of the way repl_meta_data
handles them, but I want to keep thinking about it for a little bit.
The issue is the extended components, which could differ even if the DN
is the same.
I also worry about the O(n^2) now we have made the comparison inside
much more expensive.
Regardless of my concerns, thank you so much for your continued efforts
here.
Andrew Bartlett
--
Andrew Bartlett http://samba.org/~abartlet/
Authentication Developer, Samba Team http://samba.org
More information about the samba-technical
mailing list