On Wed, Aug 7, 2013 at 11:08 PM, Ben Reser <b...@reser.org> wrote: > Yes that's true. I know there are people out there with very large > authz files though. Your cache isn't going to use much extra memory > for most connections. >
We could shave off some more if we integrated the cache into svn_config_t. But that would break encapsulation. Actually the current svn_config_t contains stuff totally irrelevant for the authz case, e.g. cfg_option_t->x_value - authz files don't contain variables that need to be expanded. It's more than that. In the worst case scenario where the users > change for every request then you're doing the work to build a cache > that's torn down without ever being used. > Correct, but the cache gets filled lazily, so we only call svn_config_enumerate2() when using the cache if we would have called it without the cache, too. And these calls are were the cycles are spent. Actually I think that your worst case scenario should be illegal and that Apache should close a connection when the user changes between requests. Anyway, I'm about to take a few days off, so I won't follow this up until Monday. Cheers, Roderich
