[PATCH] atmel_spi: support zero length transfer (resend)

[PATCH] atmel_spi: support zero length transfer (resend)

[Atsushi Nemoto]

A spi transfer with zero length is not invalid.  For example, such
transfer (len == 0 && delay_usecs != 0) can be used to achieve delay
before first CLK edge after chipselect assertion.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
---
Though some discussion about zero length transfer were raised by this
patch last time, I think there were no explicit objection to this
patch itself.

diff --git a/drivers/spi/atmel_spi.c b/drivers/spi/atmel_spi.c
index 293b7ca..5dff5e0 100644
--- a/drivers/spi/atmel_spi.c
+++ b/drivers/spi/atmel_spi.c
@@ -606,7 +606,7 @@ static int atmel_spi_transfer(struct spi_device *spi, struct spi_message *msg)
 		return -ESHUTDOWN;
 
 	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
-		if (!(xfer->tx_buf || xfer->rx_buf)) {
+		if (!(xfer->tx_buf || xfer->rx_buf) && xfer->len) {
 			dev_dbg(&spi->dev, "missing rx or tx buf\n");
 			return -EINVAL;
 		}

Re: [PATCH] atmel_spi: support zero length transfer (resend)

[Haavard Skinnemoen]

Atsushi Nemoto <anemo@mba.ocn.ne.jp> wrote:
> A spi transfer with zero length is not invalid.  For example, such
> transfer (len == 0 && delay_usecs != 0) can be used to achieve delay
> before first CLK edge after chipselect assertion.
> 
> Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
> ---
> Though some discussion about zero length transfer were raised by this
> patch last time, I think there were no explicit objection to this
> patch itself.

Well, the patch doesn't seem to do any harm, but I can't see much of a
point to it either if zero-length transfers aren't going to be
allowed...

Also, if the length is zero, the driver will end up doing a memset with
length zero at some point. Is that allowed?

Haavard

Re: [PATCH] atmel_spi: support zero length transfer (resend)

[Atsushi Nemoto]

On Sat, 29 Mar 2008 17:56:51 +0100, Haavard Skinnemoen <haavard.skinnemoen@atmel.com> wrote:
> > Though some discussion about zero length transfer were raised by this
> > patch last time, I think there were no explicit objection to this
> > patch itself.
> 
> Well, the patch doesn't seem to do any harm, but I can't see much of a
> point to it either if zero-length transfers aren't going to be
> allowed...

Well, some wired device might want long delay before first CLK edge.
I think most device do not have such constraint, but ...

Anyway, SPI framework does not reject it.  So I think the controller
driver should accept it unless it is hard to support on that hardware.

> Also, if the length is zero, the driver will end up doing a memset with
> length zero at some point. Is that allowed?

I believe memset() with zero length should be allowed.  It seems
ambiguous that dma_sync_single_for_device() with zero length is legal
or not.  Actually, it seems OK on ARM and AVR32.

---
Atsushi Nemoto