[PATCH] adcxx: Fix for single-channel ADCs

[José Miguel Gonçalves]

Hi,

While testing a ADC121S021 in an embedded board with a S3C2142 SoC (ARM 
core), I have found that the 'adcxx' driver does not handle correctly 
single channel ADCs from this chip family.
For single channel chips you must only issue one read transfer for 
correct measurement.

Regards,
José Gonçalves
Signed-off-by: Jose Miguel Goncalves <jose.goncalves@xxxxxxx>

diff --git a/drivers/hwmon/adcxx.c b/drivers/hwmon/adcxx.c
index 5e9e095..e645ce5 100644
--- a/drivers/hwmon/adcxx.c
+++ b/drivers/hwmon/adcxx.c
@@ -62,18 +62,23 @@ static ssize_t adcxx_read(struct device *dev,
 	struct spi_device *spi = to_spi_device(dev);
 	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
 	struct adcxx *adc = dev_get_drvdata(&spi->dev);
-	u8 tx_buf[2] = { attr->index << 3 }; /* other bits are don't care */
+	u8 tx_buf[2];
 	u8 rx_buf[2];
 	int status;
-	int value;
+	u32 value;
 
 	if (mutex_lock_interruptible(&adc->lock))
 		return -ERESTARTSYS;
 
-	status = spi_write_then_read(spi, tx_buf, sizeof(tx_buf),
-					rx_buf, sizeof(rx_buf));
+	if (adc->channels == 1) {
+		status = spi_read(spi, rx_buf, sizeof(rx_buf));
+	} else {
+		tx_buf[0] = attr->index << 3; /* other bits are don't care */
+		status = spi_write_then_read(spi, tx_buf, sizeof(tx_buf),
+						rx_buf, sizeof(rx_buf));
+	}
 	if (status < 0) {
-		dev_warn(dev, "spi_write_then_read failed with status %d\n",
+		dev_warn(dev, "SPI synch. transfer failed with status %d\n",
 				status);
 		goto out;
 	}