Skip to contentSkip to conversation

staging: iio: adc: ad7816: replace sprintf with sysfs_emit

PATCH In review Maxwell Doose 3 participants 5 messages Apr 19 – Apr 20

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 staging: iio: adc: ad7816: replace sprintf with sysfs_emit

This patch replaces sprintf calls in ad7816.c with sysfs_emit to help modernize the driver, ensure bounds checking, and to increase stability.

---
 drivers/staging/iio/adc/ad7816.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/staging/iio/adc/ad7816.c +8 -8
index 172acf135f3b..0e32a2295990 100644
@@ -124,8 +124,8 @@ static ssize_t ad7816_show_mode(struct device *dev,
 	struct ad7816_chip_info *chip = iio_priv(indio_dev);
 
 	if (chip->mode)
-		return sprintf(buf, "power-save\n");
-	return sprintf(buf, "full\n");
+		return sysfs_emit(buf, "power-save\n");
+	return sysfs_emit(buf, "full\n");
 }
 
 static ssize_t ad7816_store_mode(struct device *dev,
@@ -156,7 +156,7 @@ static ssize_t ad7816_show_available_modes(struct device *dev,
 					   struct device_attribute *attr,
 					   char *buf)
 {
-	return sprintf(buf, "full\npower-save\n");
+	return sysfs_emit(buf, "full\npower-save\n");
 }
 
 static IIO_DEVICE_ATTR(available_modes, 0444, ad7816_show_available_modes,
@@ -169,7 +169,7 @@ static ssize_t ad7816_show_channel(struct device *dev,
 	struct iio_dev *indio_dev = dev_to_iio_dev(dev);
 	struct ad7816_chip_info *chip = iio_priv(indio_dev);
 
-	return sprintf(buf, "%d\n", chip->channel_id);
+	return sysfs_emit(buf, "%d\n", chip->channel_id);
 }
 
 static ssize_t ad7816_store_channel(struct device *dev,
@@ -231,9 +231,9 @@ static ssize_t ad7816_show_value(struct device *dev,
 		data &= AD7816_TEMP_FLOAT_MASK;
 		if (value < 0)
 			data = BIT(AD7816_TEMP_FLOAT_OFFSET) - data;
-		return sprintf(buf, "%d.%.2d\n", value, data * 25);
+		return sysfs_emit(buf, "%d.%.2d\n", value, data * 25);
 	}
-	return sprintf(buf, "%u\n", data);
+	return sysfs_emit(buf, "%u\n", data);
 }
 
 static IIO_DEVICE_ATTR(value, 0444, ad7816_show_value, NULL, 0);
@@ -281,9 +281,9 @@ static ssize_t ad7816_show_oti(struct device *dev,
 		value = AD7816_BOUND_VALUE_MIN +
 			(chip->oti_data[chip->channel_id] -
 			AD7816_BOUND_VALUE_BASE);
-		return sprintf(buf, "%d\n", value);
+		return sysfs_emit(buf, "%d\n", value);
 	}
-	return sprintf(buf, "%u\n", chip->oti_data[chip->channel_id]);
+	return sysfs_emit(buf, "%u\n", chip->oti_data[chip->channel_id]);
 }
 
 static inline ssize_t ad7816_set_oti(struct device *dev,
Signature · 1 line
2.53.0
maintainer
#1 lore raw
Quoted 4 lines · On Sat, 18 Apr 2026 20:05:37 -0500 Maxwell Doose <[email protected]> wrote:

This patch replaces sprintf calls in ad7816.c with sysfs_emit to help modernize the driver, ensure bounds checking, and to increase stability.

Hi Maxwell,

There is a more substantial rework of this code undergoing revisions.
https://lore.kernel.org/all/[email protected]/

Both because that's been underway for a little time and because it makes a wider set of improvements, I'd rather we took that one forwards. However, as I noted in that thread, the ABI is likely to go away anyway as part of dragging this driver into a state suitable for moving out of staging. I don't mind cleaning it up as an intermediate state however as that graduation from staging may take a while!

Thanks,

Jonathan

#2 lore raw
Quoted 3 lines · Jonathan Cameron

Both because that's been underway for a little time and because it makes a wider set of improvements, I'd rather we took that one forwards.

Sorry, I didn't see the other patch. I'll focus on the v2 for the srf08 patch for the time being then.

best regards,
maxwell

#3 lore raw
In reply to Maxwell Doose
Quoted 2 lines · Maxwell Doose

This patch replaces sprintf calls in ad7816.c with sysfs_emit to help modernize the driver, ensure bounds checking, and to increase stability.

This says that changing to sysfs_emit() will increase the stability of the driver, but actually that's not true. I am fine with these sorts of hardening patches but I want the commit message to say that "The original code is fine as is" instead of implying it is unstable.

If the commit did fix a real issue, then we would need a Fixes tag.

regards,
dan carpenter

#4 lore raw
Quoted 3 lines · Dan Carpenter

This says that changing to sysfs_emit() will increase the stability of the driver, but actually that's not true.

Sorry. Like I've said in previous emails to you, I'm still figuring it all out. When I made this patch, I was under the impression that since sysfs_emit() improved security by preventing overflow errors, that it would inherently improve stability as well. Thanks for pointing out my error.