On Fri, 2005-04-15 at 15:45, Peter N. Lundblad wrote: > This idea is so simple and ellegant! My only concern is about contention > if that function is used much (say every function in a library has to use > it to ensure it was initialized). That might be over-paranoid, however.
Well, when possible, we'd retrieve a mutex once and hold onto it for a while, e.g. when we create an FS object. For the svn_utf case, we don't have an object to hold onto, but we could stuff the mutex into the current pool as a contention-free cache.