On Tue, Nov 27, 2012 at 11:42:02AM -0500, Stephen Gallagher wrote:
Actually, I agree with Michal here. Performing the NULL check first
reads better (and is more in line with our coding style elsewhere in
SSSD). The patch's real changes are viewable with 'git diff -w
HEAD^' after applying it, where it's clear that he's only fixed up
three debug messages and added the NULL check before the
previously-indented section.
I'll ack this patch as-is. There is no other appropriate split for it.
I agree. I like the code flow.
I also tested sysdb upgrade from 0.12 to 0.14 (aka from 6.3 to 6.4) and
it worked fine.
Ack