On Monday 02 July 2007, Oliver Neukum wrote: > Am Montag, 2. Juli 2007 schrieb David Brownell:
> > As noted above, for full speed devices we could get similar throughput > > with slightly more clever implementation of scatterlist handling. If > > the HCD has good hardware support for queueing, and the system has fair > > IRQ latency, recycling as few as five URBs might give similar throughput > > to the current "URB per element" approach. > > I don't like the idea of the sg code differentiating between the HCDs. It would differentiate between HCD characteristics, not caring specifically about the HCD. > BTW, it seems to me that the current sg code is underutilizing the sg > capability inherent in EHCI's qtds. If you mean that if HCDs were given scatterlists directly they might in some cases be able to pack TDs more efficiently ... sure. But the downside of that would be multiplying the complexity of those parts of HCDs. It's a lot simpler if they can know that they only need to handle one buffer at a time. - Dave ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
