Hi,
On Thu, Apr 29, 2010 at 05:05:56PM +0200, matthias wrote:
> 2010/4/29 Greg KH <greg@xxxxxxxxx>:
> > On Thu, Apr 29, 2010 at 09:54:08AM +0200, matthias wrote:
> >> 2010/4/28 Greg KH <gregkh@xxxxxxx>:
> >> > On Wed, Apr 28, 2010 at 08:45:35PM +0200, matthias wrote:
> >> >> This drivers allows a communication with the Analog Devices ADIS16255
> >> >> Low Power Gyroscope over SPI.
> >> >
> >> > Why is this going to be a staging driver? Is there a problem with it
> >> > going into the main portion of the kernel now?
> >>
> >> - No one else has tested it, so I think if it stays in staging, others
> >> have the possibility to test it.
> >> - It has checkpatch issues
> >> - Odd stuff like the AD_CHK(_ss) macro
> >> - I'm not sure to which subsystem it should be added (maybe industrial
> >> io, but this is still staging. Any suggestions?)
> >> - Improvements of code on protocol hierarchy needed (e.g. shutdown
> >> device commands in remove function; it is not obvious which sample
> >> rate the device uses)
> >
> > Ok, can you resend it after fixing the email problem, and add a TODO
> > file that lists these things and puts your email as the person to
> > contact about the driver?
>
> Yes, here we go. I added most of the things Jiri mentioned.
>
> Signed-off-by: Matthias Brugger <mensch0815@xxxxxxxxx>
There's a little typo in this function:
> +static int spi_adis16255_write_data(struct spi_adis16255_data *spiadis,
> + u8 adr1,
> + u8 adr2,
> + u8 *wbuf)
> +{
> + struct spi_device *spi = spiadis->spi;
> + struct spi_message msg;
> + struct spi_transfer xfer1, xfer2;
> + u8 *buf, *rx;
> + int ret;
> +
> + buf = kmalloc(4, GFP_KERNEL);
> + if (buf == NULL)
> + return -ENOMEM;
> +
> + rx = kzalloc(4, GFP_KERNEL);
> + if (rx == NULL) {
> + ret = -ENOMEM;
> + goto err_buf;
> + }
> +
> + spi_message_init(&msg);
> + memset(&xfer1, 0, sizeof(xfer1));
> + memset(&xfer2, 0, sizeof(xfer2));
> +
> + buf[0] = adr1 | 0x80;
> + buf[1] = *wbuf;
> +
> + buf[2] = adr2 | 0x80;
> + buf[3] = *(wbuf + 1);
> +
> + xfer1.tx_buf = buf;
> + xfer1.rx_buf = rx;
> + xfer1.len = 2;
> + xfer1.delay_usecs = 9;
> +
> + xfer2.tx_buf = buf+2;
> + xfer2.rx_buf = rx+2;
> + xfer2.len = 2;
> +
> + spi_message_add_tail(&xfer1, &msg);
> + spi_message_add_tail(&xfer2, &msg);
> +
> + ret = spi_sync(spi, &msg);
> + if (ret != 0)
> + dev_warn(&spi->dev, "wirte data to %#x %#x failed\n",
> + buf[0], buf[2]);
wirte -> write
--
Charles Clément
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
