On Fri, Mar 09, 2007 at 09:04:33AM -0500, Greg Troxel wrote: > +This is preliminary design document on the organization of the host > +component of the USRP inband signaling implementation over USB. > + > +Assumptions: we'll have a single usrp_usb_daemon, implemented as an > +mblock, that will provide the high-level message based interface to > +the USRP. The daemon will handle all resource allocation, muxing and > +demuxing for the control and status messages from multiple clients. > > It would be nice if the design for in-band signaling treated USB as > the (current) special case, and was done with GigE etc. in mind. I > didn't see anything that jumped out from this viewpoint, but earlier > is better to avoid trouble.
Yes, that's what's being done. I don't think that it's written down anywhere, but that's the goal. > Do you expect that allocation of capacity is anything more than > internal control within the usrp daemon? That conjures up isochronous > bus transactions and OS-level reservations, but I don't think you mean > that. Just an internal control, checking that the total allocation is <= 32MB/s. Eric _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
