[PATCH 0/7] hwmon: zero-initialization instead of memset

[PATCH 0/7] hwmon: zero-initialization instead of memset

[Manish Baing]

Hi all,

This patch series cleans up memory initialization across several
hardware monitoring (hwmon) drivers by replacing explicit memset() calls
with empty brace initialization (= {}).

Following similar cleanups in the IIO subsystem [1], this series updates 
these drivers to stop using memset() for stack memory initialization.
As noted in those discussions [2], using empty brace initialization (= {})
is the preferred approach. 

Beyond simple replacements, a few drivers (nct6683, nct6775-platform, it87)
were using memset() inside a for-loop. To fix this, the variable 
declarationwas moved directly inside the loop and zero-initialized there. 
This safely resets the data on every iteration and makes the code much 
cleaner.

Testing:
- Compiled all modified files using `make W=1` with no warnings or errors. 

[1]: https://lore.kernel.org/all/20250611-iio-zero-init-stack-with-instead-of-memset-v1-0-ebb2d0a24302@baylibre.com/
[2]: https://lore.kernel.org/linux-iio/202505090942.48EBF01B@keescook/

Manish Baing (7):
  hwmon: drivetemp: Use zero-initialization instead of memset()
  hwmon: f71882fg: Use zero-initialization instead of memset()
  hwmon: ltc2947-core: Use zero-initialization instead of memset()
  hwmon: nct6683: Use zero-initialization instead of memset()
  hwmon: nct6775-platform: Use zero-initialization instead of memset()
  hwmon: pmbus: adm1266: Use zero-initialization instead of memset()
  hwmon: it87: Use zero-initialization instead of memset()

 drivers/hwmon/drivetemp.c        | 3 +--
 drivers/hwmon/f71882fg.c         | 4 +---
 drivers/hwmon/it87.c             | 3 +--
 drivers/hwmon/ltc2947-core.c     | 4 +---
 drivers/hwmon/nct6683.c          | 4 ++--
 drivers/hwmon/nct6775-platform.c | 4 ++--
 drivers/hwmon/pmbus/adm1266.c    | 4 +---
 7 files changed, 9 insertions(+), 17 deletions(-)

-- 
2.43.0

[PATCH 1/7] hwmon: drivetemp: Use zero-initialization instead of memset()

[Manish Baing]

Use empty brace initialization (={}) instead of explicit memset()
to zero-initialize stack memory to simplify the code.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/drivetemp.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/hwmon/drivetemp.c b/drivers/hwmon/drivetemp.c
index 002e0660a0b8..b75cd5f5246e 100644
--- a/drivers/hwmon/drivetemp.c
+++ b/drivers/hwmon/drivetemp.c
@@ -161,11 +161,10 @@ static int drivetemp_scsi_command(struct drivetemp_data *st,
 				 u8 ata_command, u8 feature,
 				 u8 lba_low, u8 lba_mid, u8 lba_high)
 {
-	u8 scsi_cmd[MAX_COMMAND_SIZE];
+	u8 scsi_cmd[MAX_COMMAND_SIZE] = { };
 	enum req_op op;
 	int err;
 
-	memset(scsi_cmd, 0, sizeof(scsi_cmd));
 	scsi_cmd[0] = ATA_16;
 	if (ata_command == ATA_CMD_SMART && feature == SMART_WRITE_LOG) {
 		scsi_cmd[1] = (5 << 1);	/* PIO Data-out */
-- 
2.43.0

[PATCH 2/7] hwmon: f71882fg: Use zero-initialization instead of memset()

[Manish Baing]

Use empty brace initialization (={}) instead of explicit memset()
to zero-initialize stack memory to simplify the code.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/f71882fg.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/hwmon/f71882fg.c b/drivers/hwmon/f71882fg.c
index 204059d2de6c..c1d02d4e7b76 100644
--- a/drivers/hwmon/f71882fg.c
+++ b/drivers/hwmon/f71882fg.c
@@ -2667,9 +2667,7 @@ static int __init f71882fg_init(void)
 {
 	int err;
 	int address;
-	struct f71882fg_sio_data sio_data;
-
-	memset(&sio_data, 0, sizeof(sio_data));
+	struct f71882fg_sio_data sio_data = { };
 
 	address = f71882fg_find(0x2e, &sio_data);
 	if (address < 0)
-- 
2.43.0

[PATCH 3/7] hwmon: ltc2947-core: Use zero-initialization instead of memset()

[Manish Baing]

Use empty brace initialization (={}) instead of explicit memset()
to zero-initialize stack memory to simplify the code.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/ltc2947-core.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/hwmon/ltc2947-core.c b/drivers/hwmon/ltc2947-core.c
index ad7120d1e469..d95265986b18 100644
--- a/drivers/hwmon/ltc2947-core.c
+++ b/drivers/hwmon/ltc2947-core.c
@@ -273,11 +273,9 @@ static int ltc2947_alarm_read(struct ltc2947_data *st, const u8 reg,
 {
 	u8 offset = reg - LTC2947_REG_STATUS;
 	/* +1 to include status reg */
-	char alarms[LTC2947_ALERTS_SIZE + 1];
+	char alarms[LTC2947_ALERTS_SIZE + 1] = { };
 	int ret = 0;
 
-	memset(alarms, 0, sizeof(alarms));
-
 	ret = regmap_write(st->map, LTC2947_REG_PAGE_CTRL, LTC2947_PAGE0);
 	if (ret)
 		return ret;
-- 
2.43.0

[PATCH 4/7] hwmon: nct6683: Use zero-initialization instead of memset()

[Manish Baing]

Reduce the scope of 'struct resource res' by moving it inside the for-loop
and using empty brace initialization (= {}). This ensures the structure is
safely zero-initialized on every iteration, allowing for the safe removal
of explicit memset() call.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/nct6683.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/hwmon/nct6683.c b/drivers/hwmon/nct6683.c
index 0581770380cc..bd30867f9e7c 100644
--- a/drivers/hwmon/nct6683.c
+++ b/drivers/hwmon/nct6683.c
@@ -1437,7 +1437,6 @@ static int __init sensors_nct6683_init(void)
 {
 	struct nct6683_sio_data sio_data;
 	int sioaddr[2] = { 0x2e, 0x4e };
-	struct resource res;
 	bool found = false;
 	int address;
 	int i, err;
@@ -1454,6 +1453,8 @@ static int __init sensors_nct6683_init(void)
 	 * nct6683 hardware monitor, and call probe()
 	 */
 	for (i = 0; i < ARRAY_SIZE(pdev); i++) {
+		struct resource res = { };
+
 		address = nct6683_find(sioaddr[i], &sio_data);
 		if (address <= 0)
 			continue;
@@ -1471,7 +1472,6 @@ static int __init sensors_nct6683_init(void)
 		if (err)
 			goto exit_device_put;
 
-		memset(&res, 0, sizeof(res));
 		res.name = DRVNAME;
 		res.start = address + IOREGION_OFFSET;
 		res.end = address + IOREGION_OFFSET + IOREGION_LENGTH - 1;
-- 
2.43.0

[PATCH 5/7] hwmon: nct6775-platform: Use zero-initialization instead of memset()

[Manish Baing]

Reduce the scope of 'struct resource res' by moving it inside the for-loop
and using empty brace initialization (= {}). This ensures the structure is
safely zero-initialized on every iteration, allowing for the safe removal
of explicit memset() call.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/nct6775-platform.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/hwmon/nct6775-platform.c b/drivers/hwmon/nct6775-platform.c
index 1975399ac440..b73086a56d97 100644
--- a/drivers/hwmon/nct6775-platform.c
+++ b/drivers/hwmon/nct6775-platform.c
@@ -1528,7 +1528,6 @@ static int __init sensors_nct6775_platform_init(void)
 	int i, err;
 	bool found = false;
 	int address;
-	struct resource res;
 	struct nct6775_sio_data sio_data;
 	int sioaddr[2] = { 0x2e, 0x4e };
 	enum sensor_access access = access_direct;
@@ -1562,6 +1561,8 @@ static int __init sensors_nct6775_platform_init(void)
 	 * nct6775 hardware monitor, and call probe()
 	 */
 	for (i = 0; i < ARRAY_SIZE(pdev); i++) {
+		struct resource res = { };
+
 		sio_data.sio_outb = superio_outb;
 		sio_data.sio_inb = superio_inb;
 		sio_data.sio_select = superio_select;
@@ -1596,7 +1597,6 @@ static int __init sensors_nct6775_platform_init(void)
 			goto exit_device_put;
 
 		if (sio_data.access == access_direct) {
-			memset(&res, 0, sizeof(res));
 			res.name = DRVNAME;
 			res.start = address + IOREGION_OFFSET;
 			res.end = address + IOREGION_OFFSET + IOREGION_LENGTH - 1;
-- 
2.43.0

[PATCH 6/7] hwmon: pmbus: adm1266: Use zero-initialization instead of memset()

[Manish Baing]

Use empty brace initialization (={}) instead of explicit memset()
to zero-initialize stack memory to simplify the code.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/pmbus/adm1266.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/hwmon/pmbus/adm1266.c b/drivers/hwmon/pmbus/adm1266.c
index 6f6ad7b20e9a..806b92473a4f 100644
--- a/drivers/hwmon/pmbus/adm1266.c
+++ b/drivers/hwmon/pmbus/adm1266.c
@@ -449,13 +449,11 @@ static int adm1266_config_nvmem(struct adm1266_data *data)
 static int adm1266_set_rtc(struct adm1266_data *data)
 {
 	time64_t kt;
-	char write_buf[6];
+	char write_buf[6] = { };
 	int i;
 
 	kt = ktime_get_real_seconds();
 
-	memset(write_buf, 0, sizeof(write_buf));
-
 	for (i = 0; i < 4; i++)
 		write_buf[2 + i] = (kt >> (i * 8)) & 0xFF;
 
-- 
2.43.0

[PATCH 7/7] hwmon: it87: Use zero-initialization instead of memset()

[Manish Baing]

Reduce the scope of 'struct it87_sio_data sio_data' by moving it inside
the for-loop and using empty brace initialization (= {}). This ensures
the structure is safely zero-initialized on every iteration, allowing for
the safe removal of the explicit memset() call.

No functional change.

Signed-off-by: Manish Baing <manishbaing2789@gmail.com>
---
 drivers/hwmon/it87.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/hwmon/it87.c b/drivers/hwmon/it87.c
index 5fd310662ee4..f3f8d835ff92 100644
--- a/drivers/hwmon/it87.c
+++ b/drivers/hwmon/it87.c
@@ -3777,7 +3777,6 @@ MODULE_DEVICE_TABLE(dmi, it87_dmi_table);
 static int __init sm_it87_init(void)
 {
 	int sioaddr[2] = { REG_2E, REG_4E };
-	struct it87_sio_data sio_data;
 	unsigned short isa_address[2];
 	bool found = false;
 	int i, err;
@@ -3789,7 +3788,7 @@ static int __init sm_it87_init(void)
 	dmi_check_system(it87_dmi_table);
 
 	for (i = 0; i < ARRAY_SIZE(sioaddr); i++) {
-		memset(&sio_data, 0, sizeof(struct it87_sio_data));
+		struct it87_sio_data sio_data = { };
 		isa_address[i] = 0;
 		err = it87_find(sioaddr[i], &isa_address[i], &sio_data, i);
 		if (err || isa_address[i] == 0)
-- 
2.43.0

Re: [PATCH 1/7] hwmon: drivetemp: Use zero-initialization instead of memset()

[sashiko-bot]

> Use empty brace initialization (={}) instead of explicit memset()
> to zero-initialize stack memory to simplify the code.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=1

Re: [PATCH 2/7] hwmon: f71882fg: Use zero-initialization instead of memset()

[sashiko-bot]

> Use empty brace initialization (={}) instead of explicit memset()
> to zero-initialize stack memory to simplify the code.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=2

Re: [PATCH 3/7] hwmon: ltc2947-core: Use zero-initialization instead of memset()

[sashiko-bot]

> Use empty brace initialization (={}) instead of explicit memset()
> to zero-initialize stack memory to simplify the code.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=3

Re: [PATCH 4/7] hwmon: nct6683: Use zero-initialization instead of memset()

[sashiko-bot]

> Reduce the scope of 'struct resource res' by moving it inside the for-loop
> and using empty brace initialization (= {}). This ensures the structure is
> safely zero-initialized on every iteration, allowing for the safe removal
> of explicit memset() call.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=4

Re: [PATCH 5/7] hwmon: nct6775-platform: Use zero-initialization instead of memset()

[sashiko-bot]

> Reduce the scope of 'struct resource res' by moving it inside the for-loop
> and using empty brace initialization (= {}). This ensures the structure is
> safely zero-initialized on every iteration, allowing for the safe removal
> of explicit memset() call.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=5

Re: [PATCH 6/7] hwmon: pmbus: adm1266: Use zero-initialization instead of memset()

[sashiko-bot]

> Use empty brace initialization (={}) instead of explicit memset()
> to zero-initialize stack memory to simplify the code.
> 
> No functional change.
> 
> Signed-off-by: Manish Baing <manishbaing2789@gmail.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=6

Re: [PATCH 7/7] hwmon: it87: Use zero-initialization instead of memset()

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Replacing memset() with empty brace initialization leaves struct padding bytes uninitialized, causing a KMSAN uninit-value warning when copied via platform_device_add_data().
--

commit 429f8b305f54cb034b99374e3bdaa61a1ba8f424
Author: Manish Baing <manishbaing2789@gmail.com>

hwmon: it87: Use zero-initialization instead of memset()

This commit reduces the scope of 'sio_data' by moving it inside the for-loop
and replacing the explicit memset() with empty brace initialization.

> diff --git a/drivers/hwmon/it87.c b/drivers/hwmon/it87.c
> index 5fd310662ee43..f3f8d835ff92a 100644
> --- a/drivers/hwmon/it87.c
> +++ b/drivers/hwmon/it87.c
> @@ -3777,7 +3777,6 @@ MODULE_DEVICE_TABLE(dmi, it87_dmi_table);
>  static int __init sm_it87_init(void)
>  {
>  	int sioaddr[2] = { REG_2E, REG_4E };
> -	struct it87_sio_data sio_data;
>  	unsigned short isa_address[2];
>  	bool found = false;
>  	int i, err;
> @@ -3789,7 +3788,7 @@ static int __init sm_it87_init(void)
>  	dmi_check_system(it87_dmi_table);
>  
>  	for (i = 0; i < ARRAY_SIZE(sioaddr); i++) {
> -		memset(&sio_data, 0, sizeof(struct it87_sio_data));
> +		struct it87_sio_data sio_data = { };

[Severity: Medium]
Does replacing memset() with empty brace initialization leave struct padding
bytes uninitialized?

Empty brace initialization is often lowered by compilers to member-wise
zeroing, which skips padding bytes.

When this structure is passed to it87_device_add(), it is copied
byte-by-byte via platform_device_add_data():

drivers/hwmon/it87.c:it87_device_add() {
	...
	err = platform_device_add_data(pdev, sio_data,
				       sizeof(struct it87_sio_data));
	...
}

Because platform_device_add_data() uses kmemdup() to copy the entire size of
the structure, won't reading the uninitialized padding bytes trigger a KMSAN
uninitialized-value warning during driver probe?

>  		isa_address[i] = 0;
>  		err = it87_find(sioaddr[i], &isa_address[i], &sio_data, i);
>  		if (err || isa_address[i] == 0)

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260530221353.159461-1-manishbaing2789@gmail.com?part=7