[PATCH 00/20] iio: make masklength __private (round 1)

[PATCH 00/20] iio: make masklength __private (round 1)

[Nuno Sa via B4 Relay]

Hi Jonathan,

Here it goes the first round of what we discussed in [1] about
annotating .masklength as __private. Patches should be fairly simple so
hopefully I did not messed up much...

Note that in core, there's one place where we're still directly writing
into .masklength (when it's effectively set). The plan is then to
directly use ACCESS_PRIVATE() in there in the same patch where we mark
.masklength as __private.

Note this was only compile tested for arm with allyesconfig.

[1]: https://lore.kernel.org/linux-iio/20240612-dev-iio-scan-private-v1-0-7c75c8e3d30b@analog.com/

---
Nuno Sa (20):
      iio: core: add accessors  'masklength'
      iio: core: make use of iio_get_masklength()
      iio: buffer: make use of iio_get_masklength()
      iio: accel: adxl367: make use of iio_get_masklength()
      iio: accel: adx1372: make use of iio_get_masklength()
      iio: accel: bma180: make use of iio_for_each_active_channel()
      iio: accel: bmc150-accel-core: make use of iio_for_each_active_channel()
      iio: accel: cros_ec_accel_legacy: make use of iio_get_masklength()
      iio: accel: fxls8962af-core: make use of iio_for_each_active_channel()
      iio: accel: msa311: make use of iio_for_each_active_channel()
      iio: accel: sca3300: make use of iio_for_each_active_channel()
      iio: accel: stk8312: make use of iio_for_each_active_channel()
      iio: accel: stk8ba50: make use of iio_for_each_active_channel()
      iio: adc: ad7266: make use of iio_get_masklength()
      iio: adc: ad7298: make use of iio_get_masklength()
      iio: adc: ad799x: make use of iio_get_masklength()
      iio: adc: ad_sigma_delta: use new '.masklength' accessors
      iio: adc: at91_adc: make use of iio_for_each_active_channel()
      iio: imu: adis16475: make use of iio_for_each_active_channel()
      iio: imu: adis16480: make use of iio_for_each_active_channel()

 drivers/iio/accel/adxl367.c                   |  2 +-
 drivers/iio/accel/adxl372.c                   |  2 +-
 drivers/iio/accel/bma180.c                    |  3 +-
 drivers/iio/accel/bmc150-accel-core.c         |  3 +-
 drivers/iio/accel/cros_ec_accel_legacy.c      |  2 +-
 drivers/iio/accel/fxls8962af-core.c           |  3 +-
 drivers/iio/accel/msa311.c                    |  3 +-
 drivers/iio/accel/sca3300.c                   |  3 +-
 drivers/iio/accel/stk8312.c                   |  3 +-
 drivers/iio/accel/stk8ba50.c                  |  3 +-
 drivers/iio/adc/ad7266.c                      |  3 +-
 drivers/iio/adc/ad7298.c                      |  3 +-
 drivers/iio/adc/ad799x.c                      |  3 +-
 drivers/iio/adc/ad_sigma_delta.c              |  6 ++--
 drivers/iio/adc/at91_adc.c                    |  4 +--
 drivers/iio/buffer/industrialio-buffer-cb.c   |  2 +-
 drivers/iio/buffer/industrialio-hw-consumer.c |  4 +--
 drivers/iio/imu/adis16475.c                   |  3 +-
 drivers/iio/imu/adis16480.c                   |  2 +-
 drivers/iio/industrialio-buffer.c             | 50 ++++++++++++---------------
 drivers/iio/industrialio-core.c               |  2 +-
 include/linux/iio/iio.h                       | 18 ++++++++++
 22 files changed, 67 insertions(+), 60 deletions(-)
---
base-commit: 21ab030b8f8b4704c299435e167824ec90c8bab5
change-id: 20240617-dev-iio-masklength-private-023cb295a786
--

Thanks!
- Nuno Sá

[PATCH 01/20] iio: core: add accessors 'masklength'

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

'masklength' is supposed to be an IIO private member. However, drivers
(often in trigger handlers) need to access it to iterate over the
enabled channels for example (there are other reasons). Hence, a couple
of new accessors are being added:

 * iio_for_each_active_channel() - Iterates over the active channels;
 * iio_get_masklength() - Get length of the channels mask.

The goal of these new accessors is to annotate 'masklength' as private
as soon as all drivers accessing it are converted to use the new
helpers.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 include/linux/iio/iio.h | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/include/linux/iio/iio.h b/include/linux/iio/iio.h
index 894309294182d..dd6bbc468283e 100644
--- a/include/linux/iio/iio.h
+++ b/include/linux/iio/iio.h
@@ -855,6 +855,24 @@ static inline const struct iio_scan_type
 	return &chan->scan_type;
 }
 
+/**
+ * iio_get_masklength - Get length of the channels mask
+ * @indio_dev: the IIO device to get the masklength for
+ */
+static inline unsigned int iio_get_masklength(const struct iio_dev *indio_dev)
+{
+	return indio_dev->masklength;
+}
+
+/**
+ * iio_for_each_active_channel - Iterated over active channels
+ * @indio_dev: the IIO device
+ * @chan: Holds the index of the enabled channel
+ */
+#define iio_for_each_active_channel(indio_dev, chan) \
+	for_each_set_bit((chan), (indio_dev)->active_scan_mask, \
+			 iio_get_masklength(indio_dev))
+
 ssize_t iio_format_value(char *buf, unsigned int type, int size, int *vals);
 
 int iio_str_to_fixpoint(const char *str, int fract_mult, int *integer,

-- 
2.45.2

Re: [PATCH 01/20] iio: core: add accessors 'masklength'

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> 'masklength' is supposed to be an IIO private member. However, drivers
> (often in trigger handlers) need to access it to iterate over the
> enabled channels for example (there are other reasons). Hence, a couple
> of new accessors are being added:
>
>  * iio_for_each_active_channel() - Iterates over the active channels;
>  * iio_get_masklength() - Get length of the channels mask.
>
> The goal of these new accessors is to annotate 'masklength' as private
> as soon as all drivers accessing it are converted to use the new
> helpers.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  include/linux/iio/iio.h | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
>
> diff --git a/include/linux/iio/iio.h b/include/linux/iio/iio.h
> index 894309294182d..dd6bbc468283e 100644
> --- a/include/linux/iio/iio.h
> +++ b/include/linux/iio/iio.h
> @@ -855,6 +855,24 @@ static inline const struct iio_scan_type
>         return &chan->scan_type;
>  }
>
> +/**
> + * iio_get_masklength - Get length of the channels mask
> + * @indio_dev: the IIO device to get the masklength for
> + */
> +static inline unsigned int iio_get_masklength(const struct iio_dev *indio_dev)
> +{
> +       return indio_dev->masklength;
> +}
> +
> +/**
> + * iio_for_each_active_channel - Iterated over active channels
> + * @indio_dev: the IIO device
> + * @chan: Holds the index of the enabled channel
> + */
> +#define iio_for_each_active_channel(indio_dev, chan) \
> +       for_each_set_bit((chan), (indio_dev)->active_scan_mask, \
> +                        iio_get_masklength(indio_dev))
> +
>  ssize_t iio_format_value(char *buf, unsigned int type, int size, int *vals);
>
>  int iio_str_to_fixpoint(const char *str, int fract_mult, int *integer,
>
> --
> 2.45.2
>
>
>

[PATCH 02/20] iio: core: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/industrialio-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iio/industrialio-core.c b/drivers/iio/industrialio-core.c
index 2f185b3869495..e392a10aaf320 100644
--- a/drivers/iio/industrialio-core.c
+++ b/drivers/iio/industrialio-core.c
@@ -1907,7 +1907,7 @@ static void iio_sanity_check_avail_scan_masks(struct iio_dev *indio_dev)
 	int i;
 
 	av_masks = indio_dev->available_scan_masks;
-	masklength = indio_dev->masklength;
+	masklength = iio_get_masklength(indio_dev);
 	longs_per_mask = BITS_TO_LONGS(masklength);
 
 	/*

-- 
2.45.2

Re: [PATCH 02/20] iio: core: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

Well, in `industrialio-core.c` it mostly helps to use the accessor in
order to make grepping easier.

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/industrialio-core.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iio/industrialio-core.c b/drivers/iio/industrialio-core.c
> index 2f185b3869495..e392a10aaf320 100644
> --- a/drivers/iio/industrialio-core.c
> +++ b/drivers/iio/industrialio-core.c
> @@ -1907,7 +1907,7 @@ static void iio_sanity_check_avail_scan_masks(struct iio_dev *indio_dev)
>         int i;
>
>         av_masks = indio_dev->available_scan_masks;
> -       masklength = indio_dev->masklength;
> +       masklength = iio_get_masklength(indio_dev);
>         longs_per_mask = BITS_TO_LONGS(masklength);
>
>         /*
>
> --
> 2.45.2
>
>
>

[PATCH 03/20] iio: buffer: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

While at it, remove some unneeded line breaks.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/buffer/industrialio-buffer-cb.c   |  2 +-
 drivers/iio/buffer/industrialio-hw-consumer.c |  4 +--
 drivers/iio/industrialio-buffer.c             | 50 ++++++++++++---------------
 3 files changed, 26 insertions(+), 30 deletions(-)

diff --git a/drivers/iio/buffer/industrialio-buffer-cb.c b/drivers/iio/buffer/industrialio-buffer-cb.c
index 4c12b7a94af59..4befc9f55201e 100644
--- a/drivers/iio/buffer/industrialio-buffer-cb.c
+++ b/drivers/iio/buffer/industrialio-buffer-cb.c
@@ -77,7 +77,7 @@ struct iio_cb_buffer *iio_channel_get_all_cb(struct device *dev,
 	}
 
 	cb_buff->indio_dev = cb_buff->channels[0].indio_dev;
-	cb_buff->buffer.scan_mask = bitmap_zalloc(cb_buff->indio_dev->masklength,
+	cb_buff->buffer.scan_mask = bitmap_zalloc(iio_get_masklength(cb_buff->indio_dev),
 						  GFP_KERNEL);
 	if (cb_buff->buffer.scan_mask == NULL) {
 		ret = -ENOMEM;
diff --git a/drivers/iio/buffer/industrialio-hw-consumer.c b/drivers/iio/buffer/industrialio-hw-consumer.c
index fb58f599a80b1..526b2a8d725d1 100644
--- a/drivers/iio/buffer/industrialio-hw-consumer.c
+++ b/drivers/iio/buffer/industrialio-hw-consumer.c
@@ -52,6 +52,7 @@ static const struct iio_buffer_access_funcs iio_hw_buf_access = {
 static struct hw_consumer_buffer *iio_hw_consumer_get_buffer(
 	struct iio_hw_consumer *hwc, struct iio_dev *indio_dev)
 {
+	unsigned int mask_longs = BITS_TO_LONGS(iio_get_masklength(indio_dev));
 	struct hw_consumer_buffer *buf;
 
 	list_for_each_entry(buf, &hwc->buffers, head) {
@@ -59,8 +60,7 @@ static struct hw_consumer_buffer *iio_hw_consumer_get_buffer(
 			return buf;
 	}
 
-	buf = kzalloc(struct_size(buf, scan_mask, BITS_TO_LONGS(indio_dev->masklength)),
-		      GFP_KERNEL);
+	buf = kzalloc(struct_size(buf, scan_mask, mask_longs), GFP_KERNEL);
 	if (!buf)
 		return NULL;
 
diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
index 0138b21b244f0..389bfb238b510 100644
--- a/drivers/iio/industrialio-buffer.c
+++ b/drivers/iio/industrialio-buffer.c
@@ -473,18 +473,19 @@ static bool iio_validate_scan_mask(struct iio_dev *indio_dev,
 static int iio_scan_mask_set(struct iio_dev *indio_dev,
 			     struct iio_buffer *buffer, int bit)
 {
+	unsigned int masklength = iio_get_masklength(indio_dev);
 	const unsigned long *mask;
 	unsigned long *trialmask;
 
-	if (!indio_dev->masklength) {
+	if (!masklength) {
 		WARN(1, "Trying to set scanmask prior to registering buffer\n");
 		return -EINVAL;
 	}
 
-	trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
+	trialmask = bitmap_alloc(masklength, GFP_KERNEL);
 	if (!trialmask)
 		return -ENOMEM;
-	bitmap_copy(trialmask, buffer->scan_mask, indio_dev->masklength);
+	bitmap_copy(trialmask, buffer->scan_mask, masklength);
 	set_bit(bit, trialmask);
 
 	if (!iio_validate_scan_mask(indio_dev, trialmask))
@@ -492,12 +493,11 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
 
 	if (indio_dev->available_scan_masks) {
 		mask = iio_scan_mask_match(indio_dev->available_scan_masks,
-					   indio_dev->masklength,
-					   trialmask, false);
+					   masklength, trialmask, false);
 		if (!mask)
 			goto err_invalid_mask;
 	}
-	bitmap_copy(buffer->scan_mask, trialmask, indio_dev->masklength);
+	bitmap_copy(buffer->scan_mask, trialmask, masklength);
 
 	bitmap_free(trialmask);
 
@@ -517,7 +517,7 @@ static int iio_scan_mask_clear(struct iio_buffer *buffer, int bit)
 static int iio_scan_mask_query(struct iio_dev *indio_dev,
 			       struct iio_buffer *buffer, int bit)
 {
-	if (bit > indio_dev->masklength)
+	if (bit > iio_get_masklength(indio_dev))
 		return -EINVAL;
 
 	if (!buffer->scan_mask)
@@ -733,8 +733,7 @@ static int iio_compute_scan_bytes(struct iio_dev *indio_dev,
 	int length, i, largest = 0;
 
 	/* How much space will the demuxed element take? */
-	for_each_set_bit(i, mask,
-			 indio_dev->masklength) {
+	for_each_set_bit(i, mask, iio_get_masklength(indio_dev)) {
 		length = iio_storage_bytes_for_si(indio_dev, i);
 		if (length < 0)
 			return length;
@@ -855,6 +854,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
 			     struct iio_device_config *config)
 {
 	struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
+	unsigned int masklength = iio_get_masklength(indio_dev);
 	unsigned long *compound_mask;
 	const unsigned long *scan_mask;
 	bool strict_scanmask = false;
@@ -863,7 +863,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
 	unsigned int modes;
 
 	if (insert_buffer &&
-	    bitmap_empty(insert_buffer->scan_mask, indio_dev->masklength)) {
+	    bitmap_empty(insert_buffer->scan_mask, masklength)) {
 		dev_dbg(&indio_dev->dev,
 			"At least one scan element must be enabled first\n");
 		return -EINVAL;
@@ -917,7 +917,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
 	}
 
 	/* What scan mask do we actually have? */
-	compound_mask = bitmap_zalloc(indio_dev->masklength, GFP_KERNEL);
+	compound_mask = bitmap_zalloc(masklength, GFP_KERNEL);
 	if (!compound_mask)
 		return -ENOMEM;
 
@@ -927,20 +927,19 @@ static int iio_verify_update(struct iio_dev *indio_dev,
 		if (buffer == remove_buffer)
 			continue;
 		bitmap_or(compound_mask, compound_mask, buffer->scan_mask,
-			  indio_dev->masklength);
+			  masklength);
 		scan_timestamp |= buffer->scan_timestamp;
 	}
 
 	if (insert_buffer) {
 		bitmap_or(compound_mask, compound_mask,
-			  insert_buffer->scan_mask, indio_dev->masklength);
+			  insert_buffer->scan_mask, masklength);
 		scan_timestamp |= insert_buffer->scan_timestamp;
 	}
 
 	if (indio_dev->available_scan_masks) {
 		scan_mask = iio_scan_mask_match(indio_dev->available_scan_masks,
-						indio_dev->masklength,
-						compound_mask,
+						masklength, compound_mask,
 						strict_scanmask);
 		bitmap_free(compound_mask);
 		if (!scan_mask)
@@ -1005,6 +1004,7 @@ static int iio_buffer_add_demux(struct iio_buffer *buffer,
 static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 				   struct iio_buffer *buffer)
 {
+	unsigned int masklength = iio_get_masklength(indio_dev);
 	int ret, in_ind = -1, out_ind, length;
 	unsigned int in_loc = 0, out_loc = 0;
 	struct iio_demux_table *p = NULL;
@@ -1016,17 +1016,13 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 
 	/* First work out which scan mode we will actually have */
 	if (bitmap_equal(indio_dev->active_scan_mask,
-			 buffer->scan_mask,
-			 indio_dev->masklength))
+			 buffer->scan_mask, masklength))
 		return 0;
 
 	/* Now we have the two masks, work from least sig and build up sizes */
-	for_each_set_bit(out_ind,
-			 buffer->scan_mask,
-			 indio_dev->masklength) {
+	for_each_set_bit(out_ind, buffer->scan_mask, masklength) {
 		in_ind = find_next_bit(indio_dev->active_scan_mask,
-				       indio_dev->masklength,
-				       in_ind + 1);
+				       masklength, in_ind + 1);
 		while (in_ind != out_ind) {
 			ret = iio_storage_bytes_for_si(indio_dev, in_ind);
 			if (ret < 0)
@@ -1036,8 +1032,7 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
 			/* Make sure we are aligned */
 			in_loc = roundup(in_loc, length) + length;
 			in_ind = find_next_bit(indio_dev->active_scan_mask,
-					       indio_dev->masklength,
-					       in_ind + 1);
+					       masklength, in_ind + 1);
 		}
 		ret = iio_storage_bytes_for_si(indio_dev, in_ind);
 		if (ret < 0)
@@ -1646,6 +1641,7 @@ static int __iio_buffer_alloc_sysfs_and_mask(struct iio_buffer *buffer,
 					     int index)
 {
 	struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
+	unsigned int masklength = iio_get_masklength(indio_dev);
 	struct iio_dev_attr *p;
 	const struct iio_dev_attr *id_attr;
 	struct attribute **attr;
@@ -1708,8 +1704,8 @@ static int __iio_buffer_alloc_sysfs_and_mask(struct iio_buffer *buffer,
 				iio_dev_opaque->scan_index_timestamp =
 					channels[i].scan_index;
 		}
-		if (indio_dev->masklength && !buffer->scan_mask) {
-			buffer->scan_mask = bitmap_zalloc(indio_dev->masklength,
+		if (masklength && !buffer->scan_mask) {
+			buffer->scan_mask = bitmap_zalloc(masklength,
 							  GFP_KERNEL);
 			if (!buffer->scan_mask) {
 				ret = -ENOMEM;
@@ -1879,7 +1875,7 @@ void iio_buffers_free_sysfs_and_mask(struct iio_dev *indio_dev)
 bool iio_validate_scan_mask_onehot(struct iio_dev *indio_dev,
 				   const unsigned long *mask)
 {
-	return bitmap_weight(mask, indio_dev->masklength) == 1;
+	return bitmap_weight(mask, iio_get_masklength(indio_dev)) == 1;
 }
 EXPORT_SYMBOL_GPL(iio_validate_scan_mask_onehot);
 

-- 
2.45.2

Re: [PATCH 03/20] iio: buffer: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>
> While at it, remove some unneeded line breaks.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/buffer/industrialio-buffer-cb.c   |  2 +-
>  drivers/iio/buffer/industrialio-hw-consumer.c |  4 +--
>  drivers/iio/industrialio-buffer.c             | 50 ++++++++++++---------------
>  3 files changed, 26 insertions(+), 30 deletions(-)
>
> diff --git a/drivers/iio/buffer/industrialio-buffer-cb.c b/drivers/iio/buffer/industrialio-buffer-cb.c
> index 4c12b7a94af59..4befc9f55201e 100644
> --- a/drivers/iio/buffer/industrialio-buffer-cb.c
> +++ b/drivers/iio/buffer/industrialio-buffer-cb.c
> @@ -77,7 +77,7 @@ struct iio_cb_buffer *iio_channel_get_all_cb(struct device *dev,
>         }
>
>         cb_buff->indio_dev = cb_buff->channels[0].indio_dev;
> -       cb_buff->buffer.scan_mask = bitmap_zalloc(cb_buff->indio_dev->masklength,
> +       cb_buff->buffer.scan_mask = bitmap_zalloc(iio_get_masklength(cb_buff->indio_dev),
>                                                   GFP_KERNEL);
>         if (cb_buff->buffer.scan_mask == NULL) {
>                 ret = -ENOMEM;
> diff --git a/drivers/iio/buffer/industrialio-hw-consumer.c b/drivers/iio/buffer/industrialio-hw-consumer.c
> index fb58f599a80b1..526b2a8d725d1 100644
> --- a/drivers/iio/buffer/industrialio-hw-consumer.c
> +++ b/drivers/iio/buffer/industrialio-hw-consumer.c
> @@ -52,6 +52,7 @@ static const struct iio_buffer_access_funcs iio_hw_buf_access = {
>  static struct hw_consumer_buffer *iio_hw_consumer_get_buffer(
>         struct iio_hw_consumer *hwc, struct iio_dev *indio_dev)
>  {
> +       unsigned int mask_longs = BITS_TO_LONGS(iio_get_masklength(indio_dev));
>         struct hw_consumer_buffer *buf;
>
>         list_for_each_entry(buf, &hwc->buffers, head) {
> @@ -59,8 +60,7 @@ static struct hw_consumer_buffer *iio_hw_consumer_get_buffer(
>                         return buf;
>         }
>
> -       buf = kzalloc(struct_size(buf, scan_mask, BITS_TO_LONGS(indio_dev->masklength)),
> -                     GFP_KERNEL);
> +       buf = kzalloc(struct_size(buf, scan_mask, mask_longs), GFP_KERNEL);
>         if (!buf)
>                 return NULL;
>
> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
> index 0138b21b244f0..389bfb238b510 100644
> --- a/drivers/iio/industrialio-buffer.c
> +++ b/drivers/iio/industrialio-buffer.c
> @@ -473,18 +473,19 @@ static bool iio_validate_scan_mask(struct iio_dev *indio_dev,
>  static int iio_scan_mask_set(struct iio_dev *indio_dev,
>                              struct iio_buffer *buffer, int bit)
>  {
> +       unsigned int masklength = iio_get_masklength(indio_dev);
>         const unsigned long *mask;
>         unsigned long *trialmask;
>
> -       if (!indio_dev->masklength) {
> +       if (!masklength) {
>                 WARN(1, "Trying to set scanmask prior to registering buffer\n");
>                 return -EINVAL;
>         }
>
> -       trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
> +       trialmask = bitmap_alloc(masklength, GFP_KERNEL);
>         if (!trialmask)
>                 return -ENOMEM;
> -       bitmap_copy(trialmask, buffer->scan_mask, indio_dev->masklength);
> +       bitmap_copy(trialmask, buffer->scan_mask, masklength);
>         set_bit(bit, trialmask);
>
>         if (!iio_validate_scan_mask(indio_dev, trialmask))
> @@ -492,12 +493,11 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
>
>         if (indio_dev->available_scan_masks) {
>                 mask = iio_scan_mask_match(indio_dev->available_scan_masks,
> -                                          indio_dev->masklength,
> -                                          trialmask, false);
> +                                          masklength, trialmask, false);
>                 if (!mask)
>                         goto err_invalid_mask;
>         }
> -       bitmap_copy(buffer->scan_mask, trialmask, indio_dev->masklength);
> +       bitmap_copy(buffer->scan_mask, trialmask, masklength);
>
>         bitmap_free(trialmask);
>
> @@ -517,7 +517,7 @@ static int iio_scan_mask_clear(struct iio_buffer *buffer, int bit)
>  static int iio_scan_mask_query(struct iio_dev *indio_dev,
>                                struct iio_buffer *buffer, int bit)
>  {
> -       if (bit > indio_dev->masklength)
> +       if (bit > iio_get_masklength(indio_dev))
>                 return -EINVAL;
>
>         if (!buffer->scan_mask)
> @@ -733,8 +733,7 @@ static int iio_compute_scan_bytes(struct iio_dev *indio_dev,
>         int length, i, largest = 0;
>
>         /* How much space will the demuxed element take? */
> -       for_each_set_bit(i, mask,
> -                        indio_dev->masklength) {
> +       for_each_set_bit(i, mask, iio_get_masklength(indio_dev)) {
>                 length = iio_storage_bytes_for_si(indio_dev, i);
>                 if (length < 0)
>                         return length;
> @@ -855,6 +854,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
>                              struct iio_device_config *config)
>  {
>         struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
> +       unsigned int masklength = iio_get_masklength(indio_dev);
>         unsigned long *compound_mask;
>         const unsigned long *scan_mask;
>         bool strict_scanmask = false;
> @@ -863,7 +863,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
>         unsigned int modes;
>
>         if (insert_buffer &&
> -           bitmap_empty(insert_buffer->scan_mask, indio_dev->masklength)) {
> +           bitmap_empty(insert_buffer->scan_mask, masklength)) {
>                 dev_dbg(&indio_dev->dev,
>                         "At least one scan element must be enabled first\n");
>                 return -EINVAL;
> @@ -917,7 +917,7 @@ static int iio_verify_update(struct iio_dev *indio_dev,
>         }
>
>         /* What scan mask do we actually have? */
> -       compound_mask = bitmap_zalloc(indio_dev->masklength, GFP_KERNEL);
> +       compound_mask = bitmap_zalloc(masklength, GFP_KERNEL);
>         if (!compound_mask)
>                 return -ENOMEM;
>
> @@ -927,20 +927,19 @@ static int iio_verify_update(struct iio_dev *indio_dev,
>                 if (buffer == remove_buffer)
>                         continue;
>                 bitmap_or(compound_mask, compound_mask, buffer->scan_mask,
> -                         indio_dev->masklength);
> +                         masklength);
>                 scan_timestamp |= buffer->scan_timestamp;
>         }
>
>         if (insert_buffer) {
>                 bitmap_or(compound_mask, compound_mask,
> -                         insert_buffer->scan_mask, indio_dev->masklength);
> +                         insert_buffer->scan_mask, masklength);
>                 scan_timestamp |= insert_buffer->scan_timestamp;
>         }
>
>         if (indio_dev->available_scan_masks) {
>                 scan_mask = iio_scan_mask_match(indio_dev->available_scan_masks,
> -                                               indio_dev->masklength,
> -                                               compound_mask,
> +                                               masklength, compound_mask,
>                                                 strict_scanmask);
>                 bitmap_free(compound_mask);
>                 if (!scan_mask)
> @@ -1005,6 +1004,7 @@ static int iio_buffer_add_demux(struct iio_buffer *buffer,
>  static int iio_buffer_update_demux(struct iio_dev *indio_dev,
>                                    struct iio_buffer *buffer)
>  {
> +       unsigned int masklength = iio_get_masklength(indio_dev);
>         int ret, in_ind = -1, out_ind, length;
>         unsigned int in_loc = 0, out_loc = 0;
>         struct iio_demux_table *p = NULL;
> @@ -1016,17 +1016,13 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
>
>         /* First work out which scan mode we will actually have */
>         if (bitmap_equal(indio_dev->active_scan_mask,
> -                        buffer->scan_mask,
> -                        indio_dev->masklength))
> +                        buffer->scan_mask, masklength))
>                 return 0;
>
>         /* Now we have the two masks, work from least sig and build up sizes */
> -       for_each_set_bit(out_ind,
> -                        buffer->scan_mask,
> -                        indio_dev->masklength) {
> +       for_each_set_bit(out_ind, buffer->scan_mask, masklength) {
>                 in_ind = find_next_bit(indio_dev->active_scan_mask,
> -                                      indio_dev->masklength,
> -                                      in_ind + 1);
> +                                      masklength, in_ind + 1);
>                 while (in_ind != out_ind) {
>                         ret = iio_storage_bytes_for_si(indio_dev, in_ind);
>                         if (ret < 0)
> @@ -1036,8 +1032,7 @@ static int iio_buffer_update_demux(struct iio_dev *indio_dev,
>                         /* Make sure we are aligned */
>                         in_loc = roundup(in_loc, length) + length;
>                         in_ind = find_next_bit(indio_dev->active_scan_mask,
> -                                              indio_dev->masklength,
> -                                              in_ind + 1);
> +                                              masklength, in_ind + 1);
>                 }
>                 ret = iio_storage_bytes_for_si(indio_dev, in_ind);
>                 if (ret < 0)
> @@ -1646,6 +1641,7 @@ static int __iio_buffer_alloc_sysfs_and_mask(struct iio_buffer *buffer,
>                                              int index)
>  {
>         struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
> +       unsigned int masklength = iio_get_masklength(indio_dev);
>         struct iio_dev_attr *p;
>         const struct iio_dev_attr *id_attr;
>         struct attribute **attr;
> @@ -1708,8 +1704,8 @@ static int __iio_buffer_alloc_sysfs_and_mask(struct iio_buffer *buffer,
>                                 iio_dev_opaque->scan_index_timestamp =
>                                         channels[i].scan_index;
>                 }
> -               if (indio_dev->masklength && !buffer->scan_mask) {
> -                       buffer->scan_mask = bitmap_zalloc(indio_dev->masklength,
> +               if (masklength && !buffer->scan_mask) {
> +                       buffer->scan_mask = bitmap_zalloc(masklength,
>                                                           GFP_KERNEL);
>                         if (!buffer->scan_mask) {
>                                 ret = -ENOMEM;
> @@ -1879,7 +1875,7 @@ void iio_buffers_free_sysfs_and_mask(struct iio_dev *indio_dev)
>  bool iio_validate_scan_mask_onehot(struct iio_dev *indio_dev,
>                                    const unsigned long *mask)
>  {
> -       return bitmap_weight(mask, indio_dev->masklength) == 1;
> +       return bitmap_weight(mask, iio_get_masklength(indio_dev)) == 1;
>  }
>  EXPORT_SYMBOL_GPL(iio_validate_scan_mask_onehot);
>
>
> --
> 2.45.2
>
>
>

[PATCH 04/20] iio: accel: adxl367: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/adxl367.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iio/accel/adxl367.c b/drivers/iio/accel/adxl367.c
index 5cf4828a5eb5..1c046e96aef9 100644
--- a/drivers/iio/accel/adxl367.c
+++ b/drivers/iio/accel/adxl367.c
@@ -1220,7 +1220,7 @@ static int adxl367_update_scan_mode(struct iio_dev *indio_dev,
 		return ret;
 
 	st->fifo_set_size = bitmap_weight(active_scan_mask,
-					  indio_dev->masklength);
+					  iio_get_masklength(indio_dev));
 
 	return 0;
 }

-- 
2.45.2

Re: [PATCH 04/20] iio: accel: adxl367: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/adxl367.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iio/accel/adxl367.c b/drivers/iio/accel/adxl367.c
> index 5cf4828a5eb5..1c046e96aef9 100644
> --- a/drivers/iio/accel/adxl367.c
> +++ b/drivers/iio/accel/adxl367.c
> @@ -1220,7 +1220,7 @@ static int adxl367_update_scan_mode(struct iio_dev *indio_dev,
>                 return ret;
>
>         st->fifo_set_size = bitmap_weight(active_scan_mask,
> -                                         indio_dev->masklength);
> +                                         iio_get_masklength(indio_dev));
>
>         return 0;
>  }
>
> --
> 2.45.2
>
>
>

[PATCH 05/20] iio: accel: adx1372: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/adxl372.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iio/accel/adxl372.c b/drivers/iio/accel/adxl372.c
index c4193286eb05..ef8dd557877b 100644
--- a/drivers/iio/accel/adxl372.c
+++ b/drivers/iio/accel/adxl372.c
@@ -1050,7 +1050,7 @@ static int adxl372_buffer_postenable(struct iio_dev *indio_dev)
 	st->fifo_format = adxl372_axis_lookup_table[i].fifo_format;
 	st->fifo_axis_mask = adxl372_axis_lookup_table[i].bits;
 	st->fifo_set_size = bitmap_weight(indio_dev->active_scan_mask,
-					  indio_dev->masklength);
+					  iio_get_masklength(indio_dev));
 
 	/* Configure the FIFO to store sets of impact event peak. */
 	if (st->peak_fifo_mode_en) {

-- 
2.45.2

Re: [PATCH 05/20] iio: accel: adx1372: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

Typo in commit title   adx1372  ->  adxl372

With that addressed:

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/adxl372.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iio/accel/adxl372.c b/drivers/iio/accel/adxl372.c
> index c4193286eb05..ef8dd557877b 100644
> --- a/drivers/iio/accel/adxl372.c
> +++ b/drivers/iio/accel/adxl372.c
> @@ -1050,7 +1050,7 @@ static int adxl372_buffer_postenable(struct iio_dev *indio_dev)
>         st->fifo_format = adxl372_axis_lookup_table[i].fifo_format;
>         st->fifo_axis_mask = adxl372_axis_lookup_table[i].bits;
>         st->fifo_set_size = bitmap_weight(indio_dev->active_scan_mask,
> -                                         indio_dev->masklength);
> +                                         iio_get_masklength(indio_dev));
>
>         /* Configure the FIFO to store sets of impact event peak. */
>         if (st->peak_fifo_mode_en) {
>
> --
> 2.45.2
>
>
>

[PATCH 06/20] iio: accel: bma180: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/bma180.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/bma180.c b/drivers/iio/accel/bma180.c
index 6581772cb0c46..2445a0f7bc2ba 100644
--- a/drivers/iio/accel/bma180.c
+++ b/drivers/iio/accel/bma180.c
@@ -876,8 +876,7 @@ static irqreturn_t bma180_trigger_handler(int irq, void *p)
 
 	mutex_lock(&data->mutex);
 
-	for_each_set_bit(bit, indio_dev->active_scan_mask,
-			 indio_dev->masklength) {
+	iio_for_each_active_channel(indio_dev, bit) {
 		ret = bma180_get_data_reg(data, bit);
 		if (ret < 0) {
 			mutex_unlock(&data->mutex);

-- 
2.45.2

Re: [PATCH 06/20] iio: accel: bma180: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/bma180.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/bma180.c b/drivers/iio/accel/bma180.c
> index 6581772cb0c46..2445a0f7bc2ba 100644
> --- a/drivers/iio/accel/bma180.c
> +++ b/drivers/iio/accel/bma180.c
> @@ -876,8 +876,7 @@ static irqreturn_t bma180_trigger_handler(int irq, void *p)
>
>         mutex_lock(&data->mutex);
>
> -       for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                        indio_dev->masklength) {
> +       iio_for_each_active_channel(indio_dev, bit) {
>                 ret = bma180_get_data_reg(data, bit);
>                 if (ret < 0) {
>                         mutex_unlock(&data->mutex);
>
> --
> 2.45.2
>
>
>

[PATCH 07/20] iio: accel: bmc150-accel-core: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/bmc150-accel-core.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/bmc150-accel-core.c b/drivers/iio/accel/bmc150-accel-core.c
index ae0cd48a3e293..03121d020470e 100644
--- a/drivers/iio/accel/bmc150-accel-core.c
+++ b/drivers/iio/accel/bmc150-accel-core.c
@@ -1007,8 +1007,7 @@ static int __bmc150_accel_fifo_flush(struct iio_dev *indio_dev,
 		int j, bit;
 
 		j = 0;
-		for_each_set_bit(bit, indio_dev->active_scan_mask,
-				 indio_dev->masklength)
+		iio_for_each_active_channel(indio_dev, bit)
 			memcpy(&data->scan.channels[j++], &buffer[i * 3 + bit],
 			       sizeof(data->scan.channels[0]));
 

-- 
2.45.2

Re: [PATCH 07/20] iio: accel: bmc150-accel-core: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/bmc150-accel-core.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/bmc150-accel-core.c b/drivers/iio/accel/bmc150-accel-core.c
> index ae0cd48a3e293..03121d020470e 100644
> --- a/drivers/iio/accel/bmc150-accel-core.c
> +++ b/drivers/iio/accel/bmc150-accel-core.c
> @@ -1007,8 +1007,7 @@ static int __bmc150_accel_fifo_flush(struct iio_dev *indio_dev,
>                 int j, bit;
>
>                 j = 0;
> -               for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                                indio_dev->masklength)
> +               iio_for_each_active_channel(indio_dev, bit)
>                         memcpy(&data->scan.channels[j++], &buffer[i * 3 + bit],
>                                sizeof(data->scan.channels[0]));
>
>
> --
> 2.45.2
>
>
>

[PATCH 08/20] iio: accel: cros_ec_accel_legacy: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Ue iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/cros_ec_accel_legacy.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iio/accel/cros_ec_accel_legacy.c b/drivers/iio/accel/cros_ec_accel_legacy.c
index 0f403342b1fc..f7e4dc02b34d 100644
--- a/drivers/iio/accel/cros_ec_accel_legacy.c
+++ b/drivers/iio/accel/cros_ec_accel_legacy.c
@@ -62,7 +62,7 @@ static int cros_ec_accel_legacy_read_cmd(struct iio_dev *indio_dev,
 		return ret;
 	}
 
-	for_each_set_bit(i, &scan_mask, indio_dev->masklength) {
+	for_each_set_bit(i, &scan_mask, iio_get_masklength(indio_dev)) {
 		*data = st->resp->dump.sensor[sensor_num].data[i] *
 			st->sign[i];
 		data++;

-- 
2.45.2

Re: [PATCH 08/20] iio: accel: cros_ec_accel_legacy: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Ue iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

I would have suggested maybe a "iio_for_each_active_channel()" here
instead, but after looking into the driver a bit more, it looks like
it's part of a broader rework.

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/cros_ec_accel_legacy.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iio/accel/cros_ec_accel_legacy.c b/drivers/iio/accel/cros_ec_accel_legacy.c
> index 0f403342b1fc..f7e4dc02b34d 100644
> --- a/drivers/iio/accel/cros_ec_accel_legacy.c
> +++ b/drivers/iio/accel/cros_ec_accel_legacy.c
> @@ -62,7 +62,7 @@ static int cros_ec_accel_legacy_read_cmd(struct iio_dev *indio_dev,
>                 return ret;
>         }
>
> -       for_each_set_bit(i, &scan_mask, indio_dev->masklength) {
> +       for_each_set_bit(i, &scan_mask, iio_get_masklength(indio_dev)) {
>                 *data = st->resp->dump.sensor[sensor_num].data[i] *
>                         st->sign[i];
>                 data++;
>
> --
> 2.45.2
>
>
>

Re: [PATCH 08/20] iio: accel: cros_ec_accel_legacy: make use of iio_get_masklength()

[Nuno Sá]

On Wed, 2024-07-03 at 09:37 +0300, Alexandru Ardelean wrote:
> On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
> <devnull+nuno.sa.analog.com@kernel.org> wrote:
> > 
> > From: Nuno Sa <nuno.sa@analog.com>
> > 
> > Ue iio_get_masklength() to access '.masklength' so it can be annotated
> > as __private when there are no more direct users of it.
> > 
> 
> I would have suggested maybe a "iio_for_each_active_channel()" here
> instead, but after looking into the driver a bit more, it looks like
> it's part of a broader rework.
> 

Yeah, we have a scan_mask that I don't think it's guaranteed to be indio_dev-
>active_scan_mask. Also don't think (at this point) we need another
iio_for_each_active_channel() helper which takes mask as most of usecases use
indio_dev->active_scan_mask

- Nuno Sá
> 

[PATCH 09/20] iio: accel: fxls8962af-core: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there
are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/fxls8962af-core.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/fxls8962af-core.c b/drivers/iio/accel/fxls8962af-core.c
index 4fbc01bda62e4..7adb8937cb4b6 100644
--- a/drivers/iio/accel/fxls8962af-core.c
+++ b/drivers/iio/accel/fxls8962af-core.c
@@ -968,8 +968,7 @@ static int fxls8962af_fifo_flush(struct iio_dev *indio_dev)
 		int j, bit;
 
 		j = 0;
-		for_each_set_bit(bit, indio_dev->active_scan_mask,
-				 indio_dev->masklength) {
+		iio_for_each_active_channel(indio_dev, bit) {
 			memcpy(&data->scan.channels[j++], &buffer[i * 3 + bit],
 			       sizeof(data->scan.channels[0]));
 		}

-- 
2.45.2

Re: [PATCH 09/20] iio: accel: fxls8962af-core: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there
> are no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/fxls8962af-core.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/fxls8962af-core.c b/drivers/iio/accel/fxls8962af-core.c
> index 4fbc01bda62e4..7adb8937cb4b6 100644
> --- a/drivers/iio/accel/fxls8962af-core.c
> +++ b/drivers/iio/accel/fxls8962af-core.c
> @@ -968,8 +968,7 @@ static int fxls8962af_fifo_flush(struct iio_dev *indio_dev)
>                 int j, bit;
>
>                 j = 0;
> -               for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                                indio_dev->masklength) {
> +               iio_for_each_active_channel(indio_dev, bit) {
>                         memcpy(&data->scan.channels[j++], &buffer[i * 3 + bit],
>                                sizeof(data->scan.channels[0]));
>                 }
>
> --
> 2.45.2
>
>
>

[PATCH 10/20] iio: accel: msa311: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/msa311.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/msa311.c b/drivers/iio/accel/msa311.c
index b8ddbfd98f112..8861010256f2e 100644
--- a/drivers/iio/accel/msa311.c
+++ b/drivers/iio/accel/msa311.c
@@ -900,8 +900,7 @@ static irqreturn_t msa311_buffer_thread(int irq, void *p)
 
 	mutex_lock(&msa311->lock);
 
-	for_each_set_bit(bit, indio_dev->active_scan_mask,
-			 indio_dev->masklength) {
+	iio_for_each_active_channel(indio_dev, bit) {
 		chan = &msa311_channels[bit];
 
 		err = msa311_get_axis(msa311, chan, &axis);

-- 
2.45.2

Re: [PATCH 10/20] iio: accel: msa311: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:02 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/msa311.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/msa311.c b/drivers/iio/accel/msa311.c
> index b8ddbfd98f112..8861010256f2e 100644
> --- a/drivers/iio/accel/msa311.c
> +++ b/drivers/iio/accel/msa311.c
> @@ -900,8 +900,7 @@ static irqreturn_t msa311_buffer_thread(int irq, void *p)
>
>         mutex_lock(&msa311->lock);
>
> -       for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                        indio_dev->masklength) {
> +       iio_for_each_active_channel(indio_dev, bit) {
>                 chan = &msa311_channels[bit];
>
>                 err = msa311_get_axis(msa311, chan, &axis);
>
> --
> 2.45.2
>
>
>

[PATCH 11/20] iio: accel: sca3300: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/sca3300.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/sca3300.c b/drivers/iio/accel/sca3300.c
index 306482b70fad..fca77d660625 100644
--- a/drivers/iio/accel/sca3300.c
+++ b/drivers/iio/accel/sca3300.c
@@ -494,8 +494,7 @@ static irqreturn_t sca3300_trigger_handler(int irq, void *p)
 	int bit, ret, val, i = 0;
 	s16 *channels = (s16 *)data->buffer;
 
-	for_each_set_bit(bit, indio_dev->active_scan_mask,
-			 indio_dev->masklength) {
+	iio_for_each_active_channel(indio_dev, bit) {
 		ret = sca3300_read_reg(data, indio_dev->channels[bit].address, &val);
 		if (ret) {
 			dev_err_ratelimited(&data->spi->dev,

-- 
2.45.2

Re: [PATCH 11/20] iio: accel: sca3300: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/sca3300.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/sca3300.c b/drivers/iio/accel/sca3300.c
> index 306482b70fad..fca77d660625 100644
> --- a/drivers/iio/accel/sca3300.c
> +++ b/drivers/iio/accel/sca3300.c
> @@ -494,8 +494,7 @@ static irqreturn_t sca3300_trigger_handler(int irq, void *p)
>         int bit, ret, val, i = 0;
>         s16 *channels = (s16 *)data->buffer;
>
> -       for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                        indio_dev->masklength) {
> +       iio_for_each_active_channel(indio_dev, bit) {
>                 ret = sca3300_read_reg(data, indio_dev->channels[bit].address, &val);
>                 if (ret) {
>                         dev_err_ratelimited(&data->spi->dev,
>
> --
> 2.45.2
>
>
>

[PATCH 12/20] iio: accel: stk8312: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/stk8312.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/stk8312.c b/drivers/iio/accel/stk8312.c
index b3534d5751b95..abead190254b5 100644
--- a/drivers/iio/accel/stk8312.c
+++ b/drivers/iio/accel/stk8312.c
@@ -448,8 +448,7 @@ static irqreturn_t stk8312_trigger_handler(int irq, void *p)
 			goto err;
 		}
 	} else {
-		for_each_set_bit(bit, indio_dev->active_scan_mask,
-				 indio_dev->masklength) {
+		iio_for_each_active_channel(indio_dev, bit) {
 			ret = stk8312_read_accel(data, bit);
 			if (ret < 0) {
 				mutex_unlock(&data->lock);

-- 
2.45.2

Re: [PATCH 12/20] iio: accel: stk8312: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/stk8312.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/stk8312.c b/drivers/iio/accel/stk8312.c
> index b3534d5751b95..abead190254b5 100644
> --- a/drivers/iio/accel/stk8312.c
> +++ b/drivers/iio/accel/stk8312.c
> @@ -448,8 +448,7 @@ static irqreturn_t stk8312_trigger_handler(int irq, void *p)
>                         goto err;
>                 }
>         } else {
> -               for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                                indio_dev->masklength) {
> +               iio_for_each_active_channel(indio_dev, bit) {
>                         ret = stk8312_read_accel(data, bit);
>                         if (ret < 0) {
>                                 mutex_unlock(&data->lock);
>
> --
> 2.45.2
>
>
>

[PATCH 13/20] iio: accel: stk8ba50: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/accel/stk8ba50.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/accel/stk8ba50.c b/drivers/iio/accel/stk8ba50.c
index 6d3c7f444d21d..a32a77324e92d 100644
--- a/drivers/iio/accel/stk8ba50.c
+++ b/drivers/iio/accel/stk8ba50.c
@@ -330,8 +330,7 @@ static irqreturn_t stk8ba50_trigger_handler(int irq, void *p)
 			goto err;
 		}
 	} else {
-		for_each_set_bit(bit, indio_dev->active_scan_mask,
-				 indio_dev->masklength) {
+		iio_for_each_active_channel(indio_dev, bit) {
 			ret = stk8ba50_read_accel(data,
 						  stk8ba50_channel_table[bit]);
 			if (ret < 0)

-- 
2.45.2

Re: [PATCH 13/20] iio: accel: stk8ba50: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/accel/stk8ba50.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/accel/stk8ba50.c b/drivers/iio/accel/stk8ba50.c
> index 6d3c7f444d21d..a32a77324e92d 100644
> --- a/drivers/iio/accel/stk8ba50.c
> +++ b/drivers/iio/accel/stk8ba50.c
> @@ -330,8 +330,7 @@ static irqreturn_t stk8ba50_trigger_handler(int irq, void *p)
>                         goto err;
>                 }
>         } else {
> -               for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                                indio_dev->masklength) {
> +               iio_for_each_active_channel(indio_dev, bit) {
>                         ret = stk8ba50_read_accel(data,
>                                                   stk8ba50_channel_table[bit]);
>                         if (ret < 0)
>
> --
> 2.45.2
>
>
>

[PATCH 14/20] iio: adc: ad7266: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/adc/ad7266.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/iio/adc/ad7266.c b/drivers/iio/adc/ad7266.c
index 874d2dc34f928..accb78d4c3086 100644
--- a/drivers/iio/adc/ad7266.c
+++ b/drivers/iio/adc/ad7266.c
@@ -123,7 +123,8 @@ static int ad7266_update_scan_mode(struct iio_dev *indio_dev,
 	const unsigned long *scan_mask)
 {
 	struct ad7266_state *st = iio_priv(indio_dev);
-	unsigned int nr = find_first_bit(scan_mask, indio_dev->masklength);
+	unsigned int nr = find_first_bit(scan_mask,
+					 iio_get_masklength(indio_dev));
 
 	ad7266_select_input(st, nr);
 

-- 
2.45.2

Re: [PATCH 14/20] iio: adc: ad7266: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/adc/ad7266.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/iio/adc/ad7266.c b/drivers/iio/adc/ad7266.c
> index 874d2dc34f928..accb78d4c3086 100644
> --- a/drivers/iio/adc/ad7266.c
> +++ b/drivers/iio/adc/ad7266.c
> @@ -123,7 +123,8 @@ static int ad7266_update_scan_mode(struct iio_dev *indio_dev,
>         const unsigned long *scan_mask)
>  {
>         struct ad7266_state *st = iio_priv(indio_dev);
> -       unsigned int nr = find_first_bit(scan_mask, indio_dev->masklength);
> +       unsigned int nr = find_first_bit(scan_mask,
> +                                        iio_get_masklength(indio_dev));
>
>         ad7266_select_input(st, nr);
>
>
> --
> 2.45.2
>
>
>

[PATCH 15/20] iio: adc: ad7298: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/adc/ad7298.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/iio/adc/ad7298.c b/drivers/iio/adc/ad7298.c
index c0430f71f592..0128436367db 100644
--- a/drivers/iio/adc/ad7298.c
+++ b/drivers/iio/adc/ad7298.c
@@ -109,7 +109,8 @@ static int ad7298_update_scan_mode(struct iio_dev *indio_dev,
 	int scan_count;
 
 	/* Now compute overall size */
-	scan_count = bitmap_weight(active_scan_mask, indio_dev->masklength);
+	scan_count = bitmap_weight(active_scan_mask,
+				   iio_get_masklength(indio_dev));
 
 	command = AD7298_WRITE | st->ext_ref;
 

-- 
2.45.2

Re: [PATCH 15/20] iio: adc: ad7298: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/adc/ad7298.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/iio/adc/ad7298.c b/drivers/iio/adc/ad7298.c
> index c0430f71f592..0128436367db 100644
> --- a/drivers/iio/adc/ad7298.c
> +++ b/drivers/iio/adc/ad7298.c
> @@ -109,7 +109,8 @@ static int ad7298_update_scan_mode(struct iio_dev *indio_dev,
>         int scan_count;
>
>         /* Now compute overall size */
> -       scan_count = bitmap_weight(active_scan_mask, indio_dev->masklength);
> +       scan_count = bitmap_weight(active_scan_mask,
> +                                  iio_get_masklength(indio_dev));
>
>         command = AD7298_WRITE | st->ext_ref;
>
>
> --
> 2.45.2
>
>
>

[PATCH 16/20] iio: adc: ad799x: make use of iio_get_masklength()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_get_masklength() to access '.masklength' so it can be annotated
as __private when there are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/adc/ad799x.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/iio/adc/ad799x.c b/drivers/iio/adc/ad799x.c
index 0f0dcd9ca6b60..0f107e3fc2c85 100644
--- a/drivers/iio/adc/ad799x.c
+++ b/drivers/iio/adc/ad799x.c
@@ -237,7 +237,8 @@ static int ad799x_update_scan_mode(struct iio_dev *indio_dev,
 	if (!st->rx_buf)
 		return -ENOMEM;
 
-	st->transfer_size = bitmap_weight(scan_mask, indio_dev->masklength) * 2;
+	st->transfer_size = bitmap_weight(scan_mask,
+					  iio_get_masklength(indio_dev)) * 2;
 
 	switch (st->id) {
 	case ad7992:

-- 
2.45.2

Re: [PATCH 16/20] iio: adc: ad799x: make use of iio_get_masklength()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_get_masklength() to access '.masklength' so it can be annotated
> as __private when there are no more direct users of it.
>

At this point, I am wondering if it would make sense to go a little
further with the abstraction and find a wrapper for the
"bitmap_weight(indio_dev->active_scan_mask, indio_dev->masklength)"

Maybe it's doable at a later point.

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/adc/ad799x.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/iio/adc/ad799x.c b/drivers/iio/adc/ad799x.c
> index 0f0dcd9ca6b60..0f107e3fc2c85 100644
> --- a/drivers/iio/adc/ad799x.c
> +++ b/drivers/iio/adc/ad799x.c
> @@ -237,7 +237,8 @@ static int ad799x_update_scan_mode(struct iio_dev *indio_dev,
>         if (!st->rx_buf)
>                 return -ENOMEM;
>
> -       st->transfer_size = bitmap_weight(scan_mask, indio_dev->masklength) * 2;
> +       st->transfer_size = bitmap_weight(scan_mask,
> +                                         iio_get_masklength(indio_dev)) * 2;
>
>         switch (st->id) {
>         case ad7992:
>
> --
> 2.45.2
>
>
>

[PATCH 17/20] iio: adc: ad_sigma_delta: use new '.masklength' accessors

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Make use of iio_get_masklength) and iio_for_each_active_channel() to
access '.masklength' so it can be annotated as __private when there
are no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/adc/ad_sigma_delta.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/iio/adc/ad_sigma_delta.c b/drivers/iio/adc/ad_sigma_delta.c
index a2b87f6b7a07..62653d7fe39b 100644
--- a/drivers/iio/adc/ad_sigma_delta.c
+++ b/drivers/iio/adc/ad_sigma_delta.c
@@ -350,7 +350,7 @@ static int ad_sd_buffer_postenable(struct iio_dev *indio_dev)
 
 	if (sigma_delta->num_slots == 1) {
 		channel = find_first_bit(indio_dev->active_scan_mask,
-					 indio_dev->masklength);
+					 iio_get_masklength(indio_dev));
 		ret = ad_sigma_delta_set_channel(sigma_delta,
 						 indio_dev->channels[channel].address);
 		if (ret)
@@ -363,7 +363,7 @@ static int ad_sd_buffer_postenable(struct iio_dev *indio_dev)
 		 * implementation is mandatory.
 		 */
 		slot = 0;
-		for_each_set_bit(i, indio_dev->active_scan_mask, indio_dev->masklength) {
+		iio_for_each_active_channel(indio_dev, i) {
 			sigma_delta->slots[slot] = indio_dev->channels[i].address;
 			slot++;
 		}
@@ -525,7 +525,7 @@ static bool ad_sd_validate_scan_mask(struct iio_dev *indio_dev, const unsigned l
 {
 	struct ad_sigma_delta *sigma_delta = iio_device_get_drvdata(indio_dev);
 
-	return bitmap_weight(mask, indio_dev->masklength) <= sigma_delta->num_slots;
+	return bitmap_weight(mask, iio_get_masklength(indio_dev)) <= sigma_delta->num_slots;
 }
 
 static const struct iio_buffer_setup_ops ad_sd_buffer_setup_ops = {

-- 
2.45.2

Re: [PATCH 17/20] iio: adc: ad_sigma_delta: use new '.masklength' accessors

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Make use of iio_get_masklength) and iio_for_each_active_channel() to
> access '.masklength' so it can be annotated as __private when there
> are no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/adc/ad_sigma_delta.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/iio/adc/ad_sigma_delta.c b/drivers/iio/adc/ad_sigma_delta.c
> index a2b87f6b7a07..62653d7fe39b 100644
> --- a/drivers/iio/adc/ad_sigma_delta.c
> +++ b/drivers/iio/adc/ad_sigma_delta.c
> @@ -350,7 +350,7 @@ static int ad_sd_buffer_postenable(struct iio_dev *indio_dev)
>
>         if (sigma_delta->num_slots == 1) {
>                 channel = find_first_bit(indio_dev->active_scan_mask,
> -                                        indio_dev->masklength);
> +                                        iio_get_masklength(indio_dev));
>                 ret = ad_sigma_delta_set_channel(sigma_delta,
>                                                  indio_dev->channels[channel].address);
>                 if (ret)
> @@ -363,7 +363,7 @@ static int ad_sd_buffer_postenable(struct iio_dev *indio_dev)
>                  * implementation is mandatory.
>                  */
>                 slot = 0;
> -               for_each_set_bit(i, indio_dev->active_scan_mask, indio_dev->masklength) {
> +               iio_for_each_active_channel(indio_dev, i) {
>                         sigma_delta->slots[slot] = indio_dev->channels[i].address;
>                         slot++;
>                 }
> @@ -525,7 +525,7 @@ static bool ad_sd_validate_scan_mask(struct iio_dev *indio_dev, const unsigned l
>  {
>         struct ad_sigma_delta *sigma_delta = iio_device_get_drvdata(indio_dev);
>
> -       return bitmap_weight(mask, indio_dev->masklength) <= sigma_delta->num_slots;
> +       return bitmap_weight(mask, iio_get_masklength(indio_dev)) <= sigma_delta->num_slots;
>  }
>
>  static const struct iio_buffer_setup_ops ad_sd_buffer_setup_ops = {
>
> --
> 2.45.2
>
>
>

[PATCH 18/20] iio: adc: at91_adc: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/adc/at91_adc.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/iio/adc/at91_adc.c b/drivers/iio/adc/at91_adc.c
index eb501e3c86a5..af4b0224076f 100644
--- a/drivers/iio/adc/at91_adc.c
+++ b/drivers/iio/adc/at91_adc.c
@@ -268,9 +268,7 @@ static irqreturn_t at91_adc_trigger_handler(int irq, void *p)
 	struct iio_chan_spec const *chan;
 	int i, j = 0;
 
-	for (i = 0; i < idev->masklength; i++) {
-		if (!test_bit(i, idev->active_scan_mask))
-			continue;
+	iio_for_each_active_channel(idev, i) {
 		chan = idev->channels + i;
 		st->buffer[j] = at91_adc_readl(st, AT91_ADC_CHAN(st, chan->channel));
 		j++;

-- 
2.45.2

Re: [PATCH 18/20] iio: adc: at91_adc: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/adc/at91_adc.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/iio/adc/at91_adc.c b/drivers/iio/adc/at91_adc.c
> index eb501e3c86a5..af4b0224076f 100644
> --- a/drivers/iio/adc/at91_adc.c
> +++ b/drivers/iio/adc/at91_adc.c
> @@ -268,9 +268,7 @@ static irqreturn_t at91_adc_trigger_handler(int irq, void *p)
>         struct iio_chan_spec const *chan;
>         int i, j = 0;
>
> -       for (i = 0; i < idev->masklength; i++) {
> -               if (!test_bit(i, idev->active_scan_mask))
> -                       continue;
> +       iio_for_each_active_channel(idev, i) {
>                 chan = idev->channels + i;
>                 st->buffer[j] = at91_adc_readl(st, AT91_ADC_CHAN(st, chan->channel));
>                 j++;
>
> --
> 2.45.2
>
>
>

[PATCH 19/20] iio: imu: adis16475: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/imu/adis16475.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/iio/imu/adis16475.c b/drivers/iio/imu/adis16475.c
index cf73c6f46c79d..1cc2a66a78ff3 100644
--- a/drivers/iio/imu/adis16475.c
+++ b/drivers/iio/imu/adis16475.c
@@ -1608,8 +1608,7 @@ static int adis16475_push_single_sample(struct iio_poll_func *pf)
 		return -EINVAL;
 	}
 
-	for_each_set_bit(bit, indio_dev->active_scan_mask,
-			 indio_dev->masklength) {
+	iio_for_each_active_channel(indio_dev, bit) {
 		/*
 		 * When burst mode is used, system flags is the first data
 		 * channel in the sequence, but the scan index is 7.

-- 
2.45.2

Re: [PATCH 19/20] iio: imu: adis16475: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/imu/adis16475.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/drivers/iio/imu/adis16475.c b/drivers/iio/imu/adis16475.c
> index cf73c6f46c79d..1cc2a66a78ff3 100644
> --- a/drivers/iio/imu/adis16475.c
> +++ b/drivers/iio/imu/adis16475.c
> @@ -1608,8 +1608,7 @@ static int adis16475_push_single_sample(struct iio_poll_func *pf)
>                 return -EINVAL;
>         }
>
> -       for_each_set_bit(bit, indio_dev->active_scan_mask,
> -                        indio_dev->masklength) {
> +       iio_for_each_active_channel(indio_dev, bit) {
>                 /*
>                  * When burst mode is used, system flags is the first data
>                  * channel in the sequence, but the scan index is 7.
>
> --
> 2.45.2
>
>
>

[PATCH 20/20] iio: imu: adis16480: make use of iio_for_each_active_channel()

[Nuno Sa via B4 Relay]

From: Nuno Sa <nuno.sa@analog.com>

Use iio_for_each_active_channel() to iterate over active channels
accessing '.masklength' so it can be annotated as __private when there are
no more direct users of it.

Signed-off-by: Nuno Sa <nuno.sa@analog.com>
---
 drivers/iio/imu/adis16480.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iio/imu/adis16480.c b/drivers/iio/imu/adis16480.c
index 56ca5a09fbbfc..2f08268bd724a 100644
--- a/drivers/iio/imu/adis16480.c
+++ b/drivers/iio/imu/adis16480.c
@@ -1410,7 +1410,7 @@ static irqreturn_t adis16480_trigger_handler(int irq, void *p)
 		goto irq_done;
 	}
 
-	for_each_set_bit(bit, indio_dev->active_scan_mask, indio_dev->masklength) {
+	iio_for_each_active_channel(indio_dev, bit) {
 		/*
 		 * When burst mode is used, temperature is the first data
 		 * channel in the sequence, but the temperature scan index

-- 
2.45.2

Re: [PATCH 20/20] iio: imu: adis16480: make use of iio_for_each_active_channel()

[Alexandru Ardelean]

On Tue, Jul 2, 2024 at 7:03 PM Nuno Sa via B4 Relay
<devnull+nuno.sa.analog.com@kernel.org> wrote:
>
> From: Nuno Sa <nuno.sa@analog.com>
>
> Use iio_for_each_active_channel() to iterate over active channels
> accessing '.masklength' so it can be annotated as __private when there are
> no more direct users of it.
>

Reviewed-by: Alexandru Ardelean <aardelean@baylibre.com>

> Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> ---
>  drivers/iio/imu/adis16480.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iio/imu/adis16480.c b/drivers/iio/imu/adis16480.c
> index 56ca5a09fbbfc..2f08268bd724a 100644
> --- a/drivers/iio/imu/adis16480.c
> +++ b/drivers/iio/imu/adis16480.c
> @@ -1410,7 +1410,7 @@ static irqreturn_t adis16480_trigger_handler(int irq, void *p)
>                 goto irq_done;
>         }
>
> -       for_each_set_bit(bit, indio_dev->active_scan_mask, indio_dev->masklength) {
> +       iio_for_each_active_channel(indio_dev, bit) {
>                 /*
>                  * When burst mode is used, temperature is the first data
>                  * channel in the sequence, but the temperature scan index
>
> --
> 2.45.2
>
>
>

Re: [PATCH 00/20] iio: make masklength __private (round 1)

[Jonathan Cameron]

On Tue, 02 Jul 2024 18:02:32 +0200
Nuno Sa via B4 Relay <devnull+nuno.sa.analog.com@kernel.org> wrote:

> Hi Jonathan,
> 
> Here it goes the first round of what we discussed in [1] about
> annotating .masklength as __private. Patches should be fairly simple so
> hopefully I did not messed up much...
> 
> Note that in core, there's one place where we're still directly writing
> into .masklength (when it's effectively set). The plan is then to
> directly use ACCESS_PRIVATE() in there in the same patch where we mark
> .masklength as __private.

Makes sense. Series looks good to me, but I'll not pick it up just yet.
I've tagged a like second 6.11 pull request, so this is probably 6.12 material
unless things go particularly smoothly and Greg takes an additional last
minute one as there is one set waiting for a fix that is in char-misc
but not yet iio-togreg.

Jonathan

> 
> Note this was only compile tested for arm with allyesconfig.
> 
> [1]: https://lore.kernel.org/linux-iio/20240612-dev-iio-scan-private-v1-0-7c75c8e3d30b@analog.com/
> 
> ---
> Nuno Sa (20):
>       iio: core: add accessors  'masklength'
>       iio: core: make use of iio_get_masklength()
>       iio: buffer: make use of iio_get_masklength()
>       iio: accel: adxl367: make use of iio_get_masklength()
>       iio: accel: adx1372: make use of iio_get_masklength()
>       iio: accel: bma180: make use of iio_for_each_active_channel()
>       iio: accel: bmc150-accel-core: make use of iio_for_each_active_channel()
>       iio: accel: cros_ec_accel_legacy: make use of iio_get_masklength()
>       iio: accel: fxls8962af-core: make use of iio_for_each_active_channel()
>       iio: accel: msa311: make use of iio_for_each_active_channel()
>       iio: accel: sca3300: make use of iio_for_each_active_channel()
>       iio: accel: stk8312: make use of iio_for_each_active_channel()
>       iio: accel: stk8ba50: make use of iio_for_each_active_channel()
>       iio: adc: ad7266: make use of iio_get_masklength()
>       iio: adc: ad7298: make use of iio_get_masklength()
>       iio: adc: ad799x: make use of iio_get_masklength()
>       iio: adc: ad_sigma_delta: use new '.masklength' accessors
>       iio: adc: at91_adc: make use of iio_for_each_active_channel()
>       iio: imu: adis16475: make use of iio_for_each_active_channel()
>       iio: imu: adis16480: make use of iio_for_each_active_channel()
> 
>  drivers/iio/accel/adxl367.c                   |  2 +-
>  drivers/iio/accel/adxl372.c                   |  2 +-
>  drivers/iio/accel/bma180.c                    |  3 +-
>  drivers/iio/accel/bmc150-accel-core.c         |  3 +-
>  drivers/iio/accel/cros_ec_accel_legacy.c      |  2 +-
>  drivers/iio/accel/fxls8962af-core.c           |  3 +-
>  drivers/iio/accel/msa311.c                    |  3 +-
>  drivers/iio/accel/sca3300.c                   |  3 +-
>  drivers/iio/accel/stk8312.c                   |  3 +-
>  drivers/iio/accel/stk8ba50.c                  |  3 +-
>  drivers/iio/adc/ad7266.c                      |  3 +-
>  drivers/iio/adc/ad7298.c                      |  3 +-
>  drivers/iio/adc/ad799x.c                      |  3 +-
>  drivers/iio/adc/ad_sigma_delta.c              |  6 ++--
>  drivers/iio/adc/at91_adc.c                    |  4 +--
>  drivers/iio/buffer/industrialio-buffer-cb.c   |  2 +-
>  drivers/iio/buffer/industrialio-hw-consumer.c |  4 +--
>  drivers/iio/imu/adis16475.c                   |  3 +-
>  drivers/iio/imu/adis16480.c                   |  2 +-
>  drivers/iio/industrialio-buffer.c             | 50 ++++++++++++---------------
>  drivers/iio/industrialio-core.c               |  2 +-
>  include/linux/iio/iio.h                       | 18 ++++++++++
>  22 files changed, 67 insertions(+), 60 deletions(-)
> ---
> base-commit: 21ab030b8f8b4704c299435e167824ec90c8bab5
> change-id: 20240617-dev-iio-masklength-private-023cb295a786
> --
> 
> Thanks!
> - Nuno Sá
> 
> 

Re: [PATCH 00/20] iio: make masklength __private (round 1)

[Nuno Sá]

On Tue, 2024-07-02 at 21:26 +0100, Jonathan Cameron wrote:
> On Tue, 02 Jul 2024 18:02:32 +0200
> Nuno Sa via B4 Relay <devnull+nuno.sa.analog.com@kernel.org> wrote:
> 
> > Hi Jonathan,
> > 
> > Here it goes the first round of what we discussed in [1] about
> > annotating .masklength as __private. Patches should be fairly simple so
> > hopefully I did not messed up much...
> > 
> > Note that in core, there's one place where we're still directly writing
> > into .masklength (when it's effectively set). The plan is then to
> > directly use ACCESS_PRIVATE() in there in the same patch where we mark
> > .masklength as __private.
> 
> Makes sense. Series looks good to me, but I'll not pick it up just yet.
> I've tagged a like second 6.11 pull request, so this is probably 6.12 material
> unless things go particularly smoothly and Greg takes an additional last
> minute one as there is one set waiting for a fix that is in char-misc
> but not yet iio-togreg.
> 
> Jonathan
> 

Yeah, looks sane to me. Advantage of going in 6.12 is that (likely) we'' get all
the conversions done.

- Nuno Sá
> 

Re: [PATCH 00/20] iio: make masklength __private (round 1)

[Jonathan Cameron]

On Wed, 03 Jul 2024 09:10:58 +0200
Nuno Sá <noname.nuno@gmail.com> wrote:

> On Tue, 2024-07-02 at 21:26 +0100, Jonathan Cameron wrote:
> > On Tue, 02 Jul 2024 18:02:32 +0200
> > Nuno Sa via B4 Relay <devnull+nuno.sa.analog.com@kernel.org> wrote:
> >   
> > > Hi Jonathan,
> > > 
> > > Here it goes the first round of what we discussed in [1] about
> > > annotating .masklength as __private. Patches should be fairly simple so
> > > hopefully I did not messed up much...
> > > 
> > > Note that in core, there's one place where we're still directly writing
> > > into .masklength (when it's effectively set). The plan is then to
> > > directly use ACCESS_PRIVATE() in there in the same patch where we mark
> > > .masklength as __private.  
> > 
> > Makes sense. Series looks good to me, but I'll not pick it up just yet.
> > I've tagged a like second 6.11 pull request, so this is probably 6.12 material
> > unless things go particularly smoothly and Greg takes an additional last
> > minute one as there is one set waiting for a fix that is in char-misc
> > but not yet iio-togreg.
> > 
> > Jonathan
> >   
> 
> Yeah, looks sane to me. Advantage of going in 6.12 is that (likely) we'' get all
> the conversions done.

Given I'm not planning to send another pull request (no time for suitable
testing in linux-next etc) for 6.11 I've started queuing up as probable
6.12 material.

Usual process applies at this stage - I won't be pushing out as togreg
until I can rebase on rc1.

Applied and pushed out as testing for now with the typo Alexandru
pointed out in adxl372 fixed up.

Jonathan
> 
> - Nuno Sá
> >   
>