On Tue, Apr 5, 2011 at 11:58 AM, Felipe Balbi <balbi@xxxxxx> wrote: > Hi, > > On Tue, Apr 05, 2011 at 11:50:23AM +0200, Mian Yousaf Kaukab wrote: >> > @@ -1333,10 +1311,12 @@ static int do_write(struct fsg_dev *fsg) >> > if (amount_left_to_req == 0) >> > get_some_more = 0; >> > >> > - /* amount is always divisible by 512, hence by >> > - * the bulk-out maxpacket size */ >> > + /* Except at the end of the transfer, amount will be >> > + * equal to the buffer size, which is divisible by >> > + * the page size and hence by the bulk-out maxpacket >> > + * size. >> > + */ >> > bh->outreq->length = amount; >> > - bh->outreq->short_not_ok = 1; >> >> Instead of removing short_not_ok can we use something like following? >> bh->outreq->short_not_ok = (amount % maxp) ? 0 : 1; >> short_not_ok is useful for example when it is not possible to do DMA >> tranfers from usb controller's buffer to the memory for unaligned >> lengths. >> controller driver can decide to use dma or pio mode based on if short >> packets are expected or not. > > IMHO, we should avoid PIO as much as possible. Specially with USB3 > comming. It's fairly simple for driver to always set req->length as a > multiple of wMaxPacketSize and that'll make all controllers happy, not > only just a few. I didn't say anything about the req->length ;). What I am trying to say is that please do not remove short_not_ok. Even for the CBW request, it is ok to bring back set_bulk_out_req_length(). Just make sure that short_not_ok is set to 0 for that. > > -- > balbi > Thanks, -- Mian Yousaf Kaukab -- 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
