Messages in this thread | | | Date | Mon, 28 Jan 2008 02:20:00 -0800 | From | Andrew Morton <> | Subject | Re: [PATCH -mm v4 7/9] atmel_serial: Add DMA support |
| |
On Mon, 28 Jan 2008 10:59:09 +0100 Haavard Skinnemoen <hskinnemoen@atmel.com> wrote:
> > ho-hum. The generic uart buffer-handling code does ringbuffers the wrong > > way. Maybe it has to handle non-power-of-two buffer sizes. > > Hmm...I don't understand. What does it do wrong?
An faq ;) If the buffer size is a power-of-two it's better to allow the head and tail indices wrap through 0xffffffff and only mask them when subscripting. It ends up faster (usually) and you can use all of the elements of the buffer (rather than all-1) and you get nice things like:
is_empty = (head == tail) is_full = (tail - head == size) nr_items_in_ring = (tail - head)
> > > + /* more to transmit - setup next transfer */ > > > + > > > + /* disable PDC transmit */ > > > + UART_PUT_PTCR(port, ATMEL_PDC_TXTDIS); > > > + dma_sync_single_for_device(port->dev, > > > + pdc->dma_addr, > > > + pdc->dma_size, > > > + DMA_TO_DEVICE); > > > + > > > + if (xmit->tail < xmit->head) > > > + count = xmit->head - xmit->tail; > > > + else > > > + count = SERIAL_XMIT_SIZE - xmit->tail; > > > > Doesn't uart_circ_chars_pending() do this? > > Hmm...no. I think we really want something CIRC_CNT_TO_END()-ish. > > > All those uart_circ_*() macros reference their arg more than once and ... > > you know the deal. > > Yeah. Would you like a patch that inline-ifies <linux/circ.h>?
uh, if you're feeling especially keen. We have bigger problems than this.
| |