The default message transfer implementation - spi_transfer_one_message -
invokes the specific device driver's transfer_one(), then waits for
completion. However, there is no mechanism for the device driver to
report failure in the middle of the transfer.

Introduce SPI_TRANS_FAIL_IO for drivers to report transfer failure.

Signed-off-by: Nam Cao <namcao@linutronix.de>
---
 drivers/spi/spi.c       | 3 +++
 include/linux/spi/spi.h | 1 +
 2 files changed, 4 insertions(+)

diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 8ead7acb99f3..a4b8c07c5951 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -1361,6 +1361,9 @@ static int spi_transfer_wait(struct spi_controller *ctlr,
 				"SPI transfer timed out\n");
 			return -ETIMEDOUT;
 		}
+
+		if (xfer->error & SPI_TRANS_FAIL_IO)
+			return -EIO;
 	}
 
 	return 0;
diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h
index 7b4baff63c5c..ddadae2e1ead 100644
--- a/include/linux/spi/spi.h
+++ b/include/linux/spi/spi.h
@@ -1040,6 +1040,7 @@ struct spi_transfer {
 	unsigned	len;
 
 #define SPI_TRANS_FAIL_NO_START	BIT(0)
+#define SPI_TRANS_FAIL_IO	BIT(1)
 	u16		error;
 
 	dma_addr_t	tx_dma;
-- 
2.39.2

On Wed, Nov 29, 2023 at 2:32 PM Nam Cao <namcao@linutronix.de> wrote:

> The default message transfer implementation - spi_transfer_one_message -
> invokes the specific device driver's transfer_one(), then waits for
> completion. However, there is no mechanism for the device driver to
> report failure in the middle of the transfer.
>
> Introduce SPI_TRANS_FAIL_IO for drivers to report transfer failure.
>
> Signed-off-by: Nam Cao <namcao@linutronix.de>

This looks useful to me
Acked-by: Linus Walleij <linus.walleij@linaro.org>

>  #define SPI_TRANS_FAIL_NO_START        BIT(0)
> +#define SPI_TRANS_FAIL_IO      BIT(1)

Is it obvious from context when these flags get set?
from transfer_one() and in which flag field?

Otherwise maybe we should add a comment.

Yours,
Linus Walleij

On Wed, Nov 29, 2023 at 02:55:18PM +0100, Linus Walleij wrote:
> On Wed, Nov 29, 2023 at 2:32 PM Nam Cao <namcao@linutronix.de> wrote:
> >  #define SPI_TRANS_FAIL_NO_START        BIT(0)
> > +#define SPI_TRANS_FAIL_IO      BIT(1)
> 
> Is it obvious from context when these flags get set?
> from transfer_one() and in which flag field?
> 
> Otherwise maybe we should add a comment.

Agree that the purpose of this flag is not clear. I will add some
description in v2.

Best regards,
Nam