On 2/22/06, Philip Martin <philip@codematters.co.uk> wrote:
> "Garrett Rooney" <rooneg@electricjellyfish.net> writes:
>
> > There's also a question in 3 of the cases, the check might not be
> > needed at all. If we're using svn_wc_adm_access_path to get the path
> > argument, does that mean the entry will already be cached and thus
> > can't be NULL?
>
> I believe every access baton must have a corresponding entry, so it
> can't be NULL.
Makes sense to me.
> > If so, the parts marked with /* XXX ??? */ comments in
> > libsvn_wc/log.c are safe, otherwise they need attention.
> >
> > As far as this patch goes, there are 16 places where it inserts checks
> > that return error if the entry is NULL, and 1 place where it fixes the
> > check that was there to refer to the correct variable. It makes it
> > through make check, but I haven't done a whole lot else with it at
> > this point.
> >
[snip nice long review]
Thanks for the review, I think the version I committed in r18587
should take all your comments into account, although I didn't tackle
the "maybe we should be passing an entry here" part.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Feb 23 23:05:37 2006