On Fri, Apr 23, 2010 at 03:39:14PM -0700, Greg KH wrote: > On Wed, Mar 17, 2010 at 11:06:10PM +0100, Johan Hovold wrote: > > Switch to the generic, multi-urb, write implementation. > > > > Note that this will also make it fairly easy to use the generic > > fifo-based write implementation: simply unset the multi_urb_write flag > > and modify prepare_write_urb (or unset if not using a legacy SIO > > device). This may be desirable for instance on an embedded system where > > optimal throughput at high baudrates may not be as important as other > > factors (e.g. no allocations during runtime and less pressure on host > > stack). > > > > Signed-off-by: Johan Hovold <jhovold@xxxxxxxxx> > > --- > > > > Perhaps the BUG_ON in prepare_write_buffer is a bit unnecessary (also in > > aircable patch later). If someone changes ftdi_sio to use fifo-based writes > > without changing prepare_write_buffer they are sure to find out the hard way > > pretty soon. > > > > Maybe a comment next to the flag in usb_serial_driver would suffice? Or add > > BUG_ON and comment to ftdi_init? > > A comment would be good to have, adding BUG() calls to the kernel is > frowned apon these days. > > I've applied all of the patches in your series up to here. Care to > respin these last 3 without the BUG_ON() call? I'll try to get around to it as soon as possible. > Oh, and it all looks great, thanks so much for doing this work. Thanks, Johan -- 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
