On Sat, 4 Sep 2010, Martin Fuzzey wrote: > > Why would a driver want to sync the buffers manually when usbcore > > already takes care of all that? Besides, manual syncing won't mix with > > usbmon. > > > However Documentation/usb/dma.txt says: > - Some drivers may prefer to work with the model that they're mapping large > buffers, synchronizing their safe re-use. (If there's no re-use, then let > usbcore do the map/unmap.) Large periodic transfers make good examples > here, since it's cheaper to just synchronize the buffer than to unmap it > each time an urb completes and then re-map it on during resubmission. > > Is this out of date then? Yes, or inappropriate -- take your pick. In that situation it's simplest for drivers to use coherent mappings. On some platforms this might be less efficient because coherent buffers get placed in uncached memory. We don't really have a good solution when that happens. Drivers doing their own synchronization works okay _provided_ usbmon isn't being used and the HCD doesn't need to touch the data. Both of these exceptions arose long after Documentation/usb/dma.txt was written. Alan Stern -- 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
