On Thu, Sep 08, 2005 at 12:57:00PM -0400, Daniel Kristjansson wrote: > On Thu, 2005-09-08 at 17:38 +0100, Stuart Auchterlonie wrote: > > > Still not correct in 7211. > > You should stick with passing in 5000ms and do > I want 5 ms not 5000 ms. 5 ms is a LONG time for a computer running at > greater than about 4.7 Mhz... i.e. anything post 8086. :)
The 5000ms was rather arbitrary anyway. :-) I was thinking in terms of how long it might take to move a satellite dish. Technically you will still need to update the timeout setup. It breaks for anything over 1000. > > All we really want to do is wait for a context switch to the kernel > and back. As soon as we yield the backend driver should pick up the > frontend command and write to a few i/o ports on the card and then > return. The reason we don't just do a 50 ns sleep is because we want > a yield and not a busy wait loop. A 5 ms sleep also lets the kernel > do any other stuff it must do, such as set up DMA for disk writes, > reschedule the thread priorities, etc. > This makes sense. Stuart
_______________________________________________ mythtv-dev mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
