Danek Duvall wrote: > On Tue, Jan 06, 2009 at 06:22:39PM -0600, Shawn Walker wrote: > >> I'm not sure what to do here then; I can't see an os.stat() being done for >> every single add_fmri being great for publishing performance. > > Well, you could just chmod the files every time, without doing any stats. > As it is, you're already doing two stats on each file, assuming it exists.
That might be best with a few twists; I just realised we re-create the catalog file every time we add an fmri, which is atrocious anyway. The stat is probably the least of our performance concerns. I've also addressed the race condition comment you and johansen pointed out and modified the test case to account for permissions checking after publishing. Updated webrev: http://cr.opensolaris.org/~swalker/pkg-5603-3/ Sorry, I have no diff from last webrev due to a recommit faux pas on my part. Cheers, -- Shawn Walker _______________________________________________ pkg-discuss mailing list pkg-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
