> Hmm, as we're currently thinking about replacing the dreaded SL811 > for our next PXA263 design with an ISP 116x, that's bad news for me ;) > > I thought, skimming the specs from Philips, that the ISP 116x is OHCI, > but does'nt have memory-mapped registers (it's PIO).
Well, the registers are memory-mapped in a way. It's just that the data and memory lines are multiplexed. I guess that's closer to PIO. But the ISP1161's can do DMA. One of the drivers I mentioned previously does use DMA. > Now, If one would add some layer to the OHCI code which does the register > access, i.e. for memory-mapped do readx()/writex(), and for the PIO types > do PIO, then one could add a "simple" bus glue for ISP 116x chips and the > like. Isn't this what is happening in the 2.6.x kernel??? > If that's not a Really Good Idea (TM), then I wonder which options one > has, e.g. what USB Host chips are out there which do OHCI _and_ can talk > to > a processor w/o PCI? There are a number of other OHCI-like embedded host processors, but they all have their little quirks. About the only other one worth mentioning is the Transdimension TD242 / TD243. Apparently there are some other GPL Linux drivers around for those chips, but their URL escapes me atm... Philips are also bringing out a USB 2.0 Embedded Host Controller, later in the year (ISP1761), and there is also the ISP1362, but these don't have GPL drivers (yet) AFAIK... At the moment, the best "open source" controller with drivers, seems to be the Philips ISP1161A1, but it does require a 1kHz interrupt... ------------------------------------------------------- This SF.Net email is sponsored by Sleepycat Software Learn developer strategies Cisco, Motorola, Ericsson & Lucent use to deliver higher performing products faster, at low TCO. http://www.sleepycat.com/telcomwpreg.php?From=osdnemail3 _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
