Em Sun, 16 Dec 2012 19:23:28 +0100
Frank Schäfer <fschaefer.oss@xxxxxxxxxxxxxx> escreveu:
> The em2800 can transfer up to 4 bytes per i2c message.
> All other em25xx/em27xx/28xx chips can transfer at least 64 bytes per message.
>
> I2C adapters should never split messages transferred via the I2C subsystem
> into multiple message transfers, because the result will almost always NOT be
> the same as when the whole data is transferred to the I2C client in a single
> message.
> If the message size exceeds the capabilities of the I2C adapter, -EOPNOTSUPP
> should be returned.
>
> Signed-off-by: Frank Schäfer <fschaefer.oss@xxxxxxxxxxxxxx>
> ---
> drivers/media/usb/em28xx/em28xx-i2c.c | 44 ++++++++++++++-------------------
> 1 Datei geändert, 18 Zeilen hinzugefügt(+), 26 Zeilen entfernt(-)
>
> diff --git a/drivers/media/usb/em28xx/em28xx-i2c.c b/drivers/media/usb/em28xx/em28xx-i2c.c
> index 44533e4..c508c12 100644
> --- a/drivers/media/usb/em28xx/em28xx-i2c.c
> +++ b/drivers/media/usb/em28xx/em28xx-i2c.c
> @@ -50,14 +50,18 @@ do { \
> } while (0)
>
> /*
> - * em2800_i2c_send_max4()
> - * send up to 4 bytes to the i2c device
> + * em2800_i2c_send_bytes()
> + * send up to 4 bytes to the em2800 i2c device
> */
> -static int em2800_i2c_send_max4(struct em28xx *dev, u8 addr, u8 *buf, u16 len)
> +static int em2800_i2c_send_bytes(struct em28xx *dev, u8 addr, u8 *buf, u16 len)
> {
> int ret;
> int write_timeout;
> u8 b2[6];
> +
> + if (len < 1 || len > 4)
> + return -EOPNOTSUPP;
> +
Except if you actually tested it with all em2800 devices, I think that
this change just broke it for em2800.
Maybe Sascha could review this patch series on his em2800 devices.
Those devices are limited, and just like other devices (cx231xx for example),
the I2C bus need to split long messages, otherwise the I2C devices will
fail.
Btw, there was already a long discussion with regards to splitting long
I2C messages at the I2C bus or at the I2C adapters. The decision was
to do it at the I2C bus logic, as it is simpler than making a code
at each I2C client for them to properly handle -EOPNOTSUPP and implement
a fallback logic to reduce the transfer window until reach what's
supported by the device.
So, for now, I won't apply this patch series (except for patch 1, with
is obviously correct).
Cheers,
Mauro
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
