Re: CMT_EventLoop timing & crash in psm server

[Colin Blake]

OK, the hang I reported in the previous reply I think was a result of some
other debug code I still had in, because now with a fixed nsSSLIOLayerRead I
have a PSM server that stays up a lot longer than ever before. And the
browser seems a LOT faster too (we're saving a whole bunch of client/server
exchanges).

Still not perfect though. Maybe there's another "spurious" client/server
exchange I need to track down.