David Brownell <david-b@xxxxxxxxxxx> writes: > Right; such fine grain locking tends to be more hassle than it's > worth. ISTR studies from a while back showing locks need to be > quite heavily contended before splitting them is worth much. Any > per-controller lock for a UDC is unlikely to see much contention. I would gladly look at a reference for the studies if you had it somewhere. The pxa27x_udc driver is not in a position where parallelism will be a problem (as there are no pxa multi-core device AFAIK). But if there was an evolution with multiple cores, slow enough to be unable to saturate the USB bandwith alone, but which could saturate it with both cores attacking the UDC, the trouble of per-ep spinlock would be worth it ... Well, maybe a future 48-cores pxa4xx, who knows ? :) As a side note, from a very quick glance at pxa25x_udc and omap_udc, I'm a bit surprised they didn't suffer from the same behaviour. Cheers. -- Robert -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
