iio: pressure: bmp280: fix stack leak in bmp580 trigger handler
Subsystems (3)
PATCH 1/3 iio: pressure: bmp280: fix stack leak in bmp580 trigger handler
bmp580_trigger_handler() declares its scan buffer on the stack without an initializer and then memcpy()s 3 bytes of 24-bit sensor data into each 4-byte __le32 field. The high byte of comp_temp and comp_press is left uninitialized, and the channel storagebits is 32, so two bytes of stack are pushed to userspace per scan.
This is a regression from when the buffer lived in the private data, the move to a stack-local struct dropped the implicit zeroing. bme280_trigger_handler() was fixed up to handle this bug, but this driver was not fixed because there was no padding hole, but rather a short-fill issue.
Fix this all by just zero-initializing the structure on the stack.
--- drivers/iio/pressure/bmp280-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/pressure/bmp280-core.c +1 -1
@@ -2616,7 +2616,7 @@ static irqreturn_t bmp580_trigger_handler(int irq, void *p)__le32 comp_temp;__le32 comp_press;aligned_s64 timestamp; - } buffer; + } buffer = { };int ret;guard (mutex)(&data->lock);
Signature · 1 line
2.53.0
PATCH 2/3 iio: imu: st_lsm6dsx: fix stack leak in tagged FIFO buffer
The tagged FIFO path declares iio_buff on the stack with __aligned(8) but no initializer, but there is a hole in the structure, which will then leak to userspace as ST_LSM6DSX_SAMPLE_SIZE bytes (6) will be copied, but the space between that and the timestamp are not initialized.
Commit c14edb4d0bdc ("iio:imu:st_lsm6dsx Fix alignment and data leak issues") moved the untagged FIFO path to a kzalloc'd buffer in hw->scan, but for the tagged path it only added the alignment qualifier and not the initializer :(
Fix this by just zero-initializing the structure on the stack.
--- drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c +1 -1
@@ -609,7 +609,7 @@ int st_lsm6dsx_read_tagged_fifo(struct st_lsm6dsx_hw *hw)* must be passed a buffer that is aligned to 8 bytes so * as to allow insertion of a naturally aligned timestamp. */ -u8 iio_buff[ST_LSM6DSX_IIO_BUFF_SIZE ]__aligned(8) ; +u8 iio_buff[ST_LSM6DSX_IIO_BUFF_SIZE ]__aligned(8) = { };u8 tag;bool reset_ts = false;int i, err, read_len;
Signature · 1 line
2.53.0
Re: [PATCH 2/3] iio: imu: st_lsm6dsx: fix stack leak in tagged FIFO buffer
Quoted 6 lines · Greg Kroah-Hartman
The tagged FIFO path declares
iio_buffon the stack with__aligned(8)but no initializer, but there is a hole in the structure, which will then leak to userspace asST_LSM6DSX_SAMPLE_SIZEbytes (6) will be copied, but the space between that and the timestamp are not initialized.
Quoted 7 lines
Commit
c14edb4d0bdc("iio:imu:st_lsm6dsx Fix alignment and data leak issues") moved the untagged FIFO path to a kzalloc'd buffer inhw->scan, but for the tagged path it only added the alignment qualifier and not the initializer :(Fix this by just zero-initializing the structure on the stack.
drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c int st_lsm6dsx_read_tagged_fifo(struct st_lsm6dsx_hw *hw) · line 612
Quoted 11 lines
drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c +1
-1@@ -609,7 +609,7 @@ int st_lsm6dsx_read_tagged_fifo(struct st_lsm6dsx_hw *hw)* must be passed a buffer that is aligned to 8 bytes so * as to allow insertion of a naturally aligned timestamp. */ -u8 iio_buff[ST_LSM6DSX_IIO_BUFF_SIZE ]__aligned(8) ; +u8 iio_buff[ST_LSM6DSX_IIO_BUFF_SIZE ]__aligned(8) = { };
Looks like a case where we could follow this up with a patch to use IIO_DECLARE_BUF_WITH_TS().
Quoted 3 lines
drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_buffer.c +0
-0u8 tag;bool reset_ts = false;int i, err, read_len;
PATCH 3/3 iio: imu: adis16550: fix stack leak in trigger handler
In reply to Greg Kroah-Hartmanadis16550_trigger_handler() declares the scan data array on the stack without initializing it. The memcpy() at the bottom fills only the first 28 bytes (TEMP + 6 channels of GYRO/ACCEL data), and iio_push_to_buffers_with_timestamp() writes the s64 timestamp at the 8-byte-aligned offset 32. Bytes 28-31 remain uninitialized stack data which leaks to userspace on ever trigger.
Fix this all by just zero-initializing the structure on the stack.
--- drivers/iio/imu/adis16550.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/imu/adis16550.c +1 -1
@@ -836,7 +836,7 @@ static irqreturn_t adis16550_trigger_handler(int irq, void *p)u16 dummy;bool valid;struct iio_poll_func *pf = p; -__be32 data[ADIS16550_MAX_SCAN_DATA ]__aligned(8) ; +__be32 data[ADIS16550_MAX_SCAN_DATA ]__aligned(8) = { };struct iio_dev *indio_dev = pf->indio_dev;struct adis16550 *st =iio_priv (indio_dev);struct adis *adis =iio_device_get_drvdata (indio_dev);
Signature · 1 line
2.53.0
Re: [PATCH 3/3] iio: imu: adis16550: fix stack leak in trigger handler
Quoted 9 lines · Greg Kroah-Hartman
adis16550_trigger_handler()declares the scan data array on the stack without initializing it. Thememcpy()at the bottom fills only the first 28 bytes (TEMP + 6 channels of GYRO/ACCEL data), andiio_push_to_buffers_with_timestamp()writes the s64 timestamp at the 8-byte-aligned offset 32. Bytes 28-31 remain uninitialized stack data which leaks to userspace on ever trigger.Fix this all by just zero-initializing the structure on the stack.
drivers/iio/imu/adis16550.c static irqreturn_t adis16550_trigger_handler(int irq, void *p) · line 839
Quoted 11 lines
drivers/iio/imu/adis16550.c +1
-1@@ -836,7 +836,7 @@ static irqreturn_t adis16550_trigger_handler(int irq, void *p)u16 dummy;bool valid;struct iio_poll_func *pf = p; -__be32 data[ADIS16550_MAX_SCAN_DATA ]__aligned(8) ; +__be32 data[ADIS16550_MAX_SCAN_DATA ]__aligned(8) = { };
And another case where a followup patch to use IIO_DECLARE_BUFFER_WITH_TS() would be appropriate.
Quoted 3 lines
drivers/iio/imu/adis16550.c +0
-0struct iio_dev *indio_dev = pf->indio_dev;struct adis16550 *st =iio_priv (indio_dev);struct adis *adis =iio_device_get_drvdata (indio_dev);
Quoted 6 lines · Greg Kroah-Hartman
bmp580_trigger_handler()declares its scan buffer on the stack without an initializer and then memcpy()s 3 bytes of 24-bit sensor data into each 4-byte__le32field. The high byte ofcomp_tempandcomp_pressis left uninitialized, and the channel storagebits is 32, so two bytes of stack are pushed to userspace per scan.
Quoted 8 lines
This is a regression from when the buffer lived in the private data, the move to a stack-local struct dropped the implicit zeroing.
bme280_trigger_handler()was fixed up to handle this bug, but this driver was not fixed because there was no padding hole, but rather a short-fill issue.Fix this all by just zero-initializing the structure on the stack.
Jonathan Cameron
maintainerQuoted 13 lines · On Thu, 9 Apr 2026 10:01:43 -0500 David Lechner <[email protected]> wrote:
Quoted 12 lines · Greg Kroah-Hartman
bmp580_trigger_handler()declares its scan buffer on the stack without an initializer and then memcpy()s 3 bytes of 24-bit sensor data into each 4-byte__le32field. The high byte ofcomp_tempandcomp_pressis left uninitialized, and the channel storagebits is 32, so two bytes of stack are pushed to userspace per scan.This is a regression from when the buffer lived in the private data, the move to a stack-local struct dropped the implicit zeroing.
bme280_trigger_handler()was fixed up to handle this bug, but this driver was not fixed because there was no padding hole, but rather a short-fill issue.
Quoted 4 lines
Quoted 2 lines
Fix this all by just zero-initializing the structure on the stack.
Reviewed-by David Lechner <[email protected]>
Series applied to the fixes-togreg branch of iio.git.
I'll be rebase on rc1 once available before sending out a pull request but in the meantime it can get some build coverage
Thanks,
Jonathan