ACL module patches

Matthias Dieter Wallnöfer mdw at
Tue Jul 6 13:53:46 MDT 2010


I did only add a check that means: if the control was specified then 
please handle this the same way as if we would have one delete and one 
add password mod (password modify operation). If the control isn't 
specified then do perform the other checks which you've introduced.


Nadezhda Ivanova wrote:
> Just one quick question - obviously I do not understand this control, 
> but are you sure we should perform the access check only if the 
> control has been specified?
> Regards,
> Nadya
> On Tue, Jul 6, 2010 at 8:50 PM, Nadezhda Ivanova <nivanova at 
> <mailto:nivanova at>> wrote:
>     Hi Matthias,
>     I basically made sure to do what you did in password_hash, and I
>     did not see this control there, but it looks all right. As for the
>     memory context patches - I am not so sure. They will not have any
>     effect but as far as I know contexts should be as granular as
>     possible, otherwise you risk a method freeing a context that was
>     passed as parameter instead of the local one. And the habit of
>     moving line endings around even if they do not exceed the
>     character limit... Well, in short, I am not convinced that the
>     second patch is necessary, but I am not against applying it.
>     However, since they are your patches, I think you should push them
>     :). Just make sure to run both and tests before
>     that.
>     Regards,
>     Nadya
>     On Tue, Jul 6, 2010 at 7:55 PM, Matthias Dieter Wallnöfer
>     <mdw at <mailto:mdw at>> wrote:
>         Hi Nadya,
>         the ACL password work does work well beside one exception of
>         which you probably wasn't aware. I put here the link to the
>         patch from which the commit message should explain the reason:
>         The other patch should fix the outstanding memory contexts:
>         Would be nice if you could integrates this one too.
>         Greets,
>         Matthias

More information about the samba-technical mailing list