[PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[Jakob Koschel]

In preparation to limit the scope of the list iterator variable to the
list traversal loop, use a dedicated pointer to iterate through the
list [1].

Since that variable should not be used past the loop iteration, a
separate variable is used to 'remember the current location within the
loop'.

To either continue iterating from that position or start a new
iteration (if the previous iteration was complete) list_prepare_entry()
is used.

Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/ [1]
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
---
 drivers/iio/industrialio-buffer.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
index 208b5193c621..151a77c2affd 100644
--- a/drivers/iio/industrialio-buffer.c
+++ b/drivers/iio/industrialio-buffer.c
@@ -1059,7 +1059,7 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,
 	struct iio_device_config *config)
 {
 	struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
-	struct iio_buffer *buffer;
+	struct iio_buffer *buffer, *tmp = NULL;
 	int ret;

 	indio_dev->active_scan_mask = config->scan_mask;
@@ -1097,8 +1097,10 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,

 	list_for_each_entry(buffer, &iio_dev_opaque->buffer_list, buffer_list) {
 		ret = iio_buffer_enable(buffer, indio_dev);
-		if (ret)
+		if (ret) {
+			tmp = buffer;
 			goto err_disable_buffers;
+		}
 	}

 	if (indio_dev->currentmode == INDIO_BUFFER_TRIGGERED) {
@@ -1125,6 +1127,7 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,
 					     indio_dev->pollfunc);
 	}
 err_disable_buffers:
+	buffer = list_prepare_entry(tmp, &iio_dev_opaque->buffer_list, buffer_list);
 	list_for_each_entry_continue_reverse(buffer, &iio_dev_opaque->buffer_list,
 					     buffer_list)
 		iio_buffer_disable(buffer, indio_dev);

base-commit: f82da161ea75dc4db21b2499e4b1facd36dab275
--
2.25.1

[PATCH 2/3] iio: ssp_sensors: replace usage of found with dedicated list iterator variable

[Jakob Koschel]

To move the list iterator variable into the list_for_each_entry_*()
macro in the future it should be avoided to use the list iterator
variable after the loop body.

To *never* use the list iterator variable after the loop it was
concluded to use a separate iterator variable instead of a
found boolean [1].

This removes the need to use a found variable and simply checking if
the variable was set, can determine if the break/goto was hit.

Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/ [1]
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
---
 drivers/iio/common/ssp_sensors/ssp_spi.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/iio/common/ssp_sensors/ssp_spi.c b/drivers/iio/common/ssp_sensors/ssp_spi.c
index 769bd9280524..f32b04b63ea1 100644
--- a/drivers/iio/common/ssp_sensors/ssp_spi.c
+++ b/drivers/iio/common/ssp_sensors/ssp_spi.c
@@ -331,12 +331,11 @@ static int ssp_parse_dataframe(struct ssp_data *data, char *dataframe, int len)
 /* threaded irq */
 int ssp_irq_msg(struct ssp_data *data)
 {
-	bool found = false;
 	char *buffer;
 	u8 msg_type;
 	int ret;
 	u16 length, msg_options;
-	struct ssp_msg *msg, *n;
+	struct ssp_msg *msg = NULL, *iter, *n;
 
 	ret = spi_read(data->spi, data->header_buffer, SSP_HEADER_BUFFER_SIZE);
 	if (ret < 0) {
@@ -362,15 +361,15 @@ int ssp_irq_msg(struct ssp_data *data)
 		 * received with no order
 		 */
 		mutex_lock(&data->pending_lock);
-		list_for_each_entry_safe(msg, n, &data->pending_list, list) {
-			if (msg->options == msg_options) {
-				list_del(&msg->list);
-				found = true;
+		list_for_each_entry_safe(iter, n, &data->pending_list, list) {
+			if (iter->options == msg_options) {
+				list_del(&iter->list);
+				msg = iter;
 				break;
 			}
 		}
 
-		if (!found) {
+		if (!msg) {
 			/*
 			 * here can be implemented dead messages handling
 			 * but the slave should not send such ones - it is to
-- 
2.25.1

RE: [PATCH 2/3] iio: ssp_sensors: replace usage of found with dedicated list iterator variable

[Sa, Nuno]

> -----Original Message-----
> From: Jakob Koschel <jakobkoschel@gmail.com>
> Sent: Friday, April 1, 2022 1:07 AM
> To: Jonathan Cameron <jic23@kernel.org>
> Cc: Lars-Peter Clausen <lars@metafoo.de>; Dan Carpenter
> <dan.carpenter@oracle.com>; Jakob Koschel
> <jakobkoschel@gmail.com>; linux-iio@vger.kernel.org; linux-
> kernel@vger.kernel.org; Mike Rapoport <rppt@kernel.org>; Brian
> Johannesmeyer <bjohannesmeyer@gmail.com>; Cristiano Giuffrida
> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
> Subject: [PATCH 2/3] iio: ssp_sensors: replace usage of found with
> dedicated list iterator variable
> 
> [External]
> 
> To move the list iterator variable into the list_for_each_entry_*()
> macro in the future it should be avoided to use the list iterator
> variable after the loop body.
> 
> To *never* use the list iterator variable after the loop it was
> concluded to use a separate iterator variable instead of a
> found boolean [1].
> 
> This removes the need to use a found variable and simply checking if
> the variable was set, can determine if the break/goto was hit.
> 
> Link: https://urldefense.com/v3/__https://lore.kernel.org/all/CAHk-
> =wgRr_D8CB-D9Kg-
> c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/__;!!A3Ni8CS0y
> 2Y!tqcYepOSYys02RcvKD7LiZk3mHwPmSLa_1S0QglWPRZSCoPnXk3ccjY
> _XzsDDA$  [1]
> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
> ---

Reviewed-by: Nuno Sá <nuno.sa@analog.com>

[PATCH 3/3] iio: sysfs-trigger: replace usage of found with dedicated list iterator variable

[Jakob Koschel]

To move the list iterator variable into the list_for_each_entry_*()
macro in the future it should be avoided to use the list iterator
variable after the loop body.

To *never* use the list iterator variable after the loop it was
concluded to use a separate iterator variable instead of a
found boolean [1].

This removes the need to use a found variable and simply checking if
the variable was set, can determine if the break/goto was hit.

Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/ [1]
Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
---
 drivers/iio/trigger/iio-trig-sysfs.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/iio/trigger/iio-trig-sysfs.c b/drivers/iio/trigger/iio-trig-sysfs.c
index 2a4b75897910..f1a8704e6cc1 100644
--- a/drivers/iio/trigger/iio-trig-sysfs.c
+++ b/drivers/iio/trigger/iio-trig-sysfs.c
@@ -176,16 +176,15 @@ static int iio_sysfs_trigger_probe(int id)
 
 static int iio_sysfs_trigger_remove(int id)
 {
-	bool foundit = false;
-	struct iio_sysfs_trig *t;
+	struct iio_sysfs_trig *t = NULL, *iter;
 
 	mutex_lock(&iio_sysfs_trig_list_mut);
-	list_for_each_entry(t, &iio_sysfs_trig_list, l)
-		if (id == t->id) {
-			foundit = true;
+	list_for_each_entry(iter, &iio_sysfs_trig_list, l)
+		if (id == iter->id) {
+			t = iter;
 			break;
 		}
-	if (!foundit) {
+	if (!t) {
 		mutex_unlock(&iio_sysfs_trig_list_mut);
 		return -EINVAL;
 	}
-- 
2.25.1

RE: [PATCH 3/3] iio: sysfs-trigger: replace usage of found with dedicated list iterator variable

[Sa, Nuno]

> -----Original Message-----
> From: Jakob Koschel <jakobkoschel@gmail.com>
> Sent: Friday, April 1, 2022 1:07 AM
> To: Jonathan Cameron <jic23@kernel.org>
> Cc: Lars-Peter Clausen <lars@metafoo.de>; Dan Carpenter
> <dan.carpenter@oracle.com>; Jakob Koschel
> <jakobkoschel@gmail.com>; linux-iio@vger.kernel.org; linux-
> kernel@vger.kernel.org; Mike Rapoport <rppt@kernel.org>; Brian
> Johannesmeyer <bjohannesmeyer@gmail.com>; Cristiano Giuffrida
> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
> Subject: [PATCH 3/3] iio: sysfs-trigger: replace usage of found with
> dedicated list iterator variable
> 
> [External]
> 
> To move the list iterator variable into the list_for_each_entry_*()
> macro in the future it should be avoided to use the list iterator
> variable after the loop body.
> 
> To *never* use the list iterator variable after the loop it was
> concluded to use a separate iterator variable instead of a
> found boolean [1].
> 
> This removes the need to use a found variable and simply checking if
> the variable was set, can determine if the break/goto was hit.
> 
> Link: https://urldefense.com/v3/__https://lore.kernel.org/all/CAHk-
> =wgRr_D8CB-D9Kg-
> c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/__;!!A3Ni8CS0y
> 2Y!q8mEciLiCAWjMGEwW9jHDD2DoeaKzrMwtUcRCNAZ0gW9DPBVtJk
> xR5FMPPCJNw$  [1]
> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
> ---

Reviewed-by: Nuno Sá <nuno.sa@analog.com>

RE: [PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[Sa, Nuno]

Hi Jakob,

> -----Original Message-----
> From: Jakob Koschel <jakobkoschel@gmail.com>
> Sent: Friday, April 1, 2022 1:07 AM
> To: Jonathan Cameron <jic23@kernel.org>
> Cc: Lars-Peter Clausen <lars@metafoo.de>; Dan Carpenter
> <dan.carpenter@oracle.com>; Jakob Koschel
> <jakobkoschel@gmail.com>; linux-iio@vger.kernel.org; linux-
> kernel@vger.kernel.org; Mike Rapoport <rppt@kernel.org>; Brian
> Johannesmeyer <bjohannesmeyer@gmail.com>; Cristiano Giuffrida
> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
> Subject: [PATCH 1/3] iio: buffer: remove usage of list iterator variable
> for list_for_each_entry_continue_reverse()
> 
> [External]
> 
> In preparation to limit the scope of the list iterator variable to the
> list traversal loop, use a dedicated pointer to iterate through the
> list [1].
> 
> Since that variable should not be used past the loop iteration, a
> separate variable is used to 'remember the current location within the
> loop'.
> 
> To either continue iterating from that position or start a new
> iteration (if the previous iteration was complete) list_prepare_entry()
> is used.
> 
> Link: https://urldefense.com/v3/__https://lore.kernel.org/all/CAHk-
> =wgRr_D8CB-D9Kg-
> c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/__;!!A3Ni8CS0y
> 2Y!q8llw5UCaMIsAU7tPtPDhwVor0wy032I7FJHv0VxBZksNuRJF04HjWe
> 0XYG7OQ$  [1]
> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
> ---
>  drivers/iio/industrialio-buffer.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-
> buffer.c
> index 208b5193c621..151a77c2affd 100644
> --- a/drivers/iio/industrialio-buffer.c
> +++ b/drivers/iio/industrialio-buffer.c
> @@ -1059,7 +1059,7 @@ static int iio_enable_buffers(struct iio_dev
> *indio_dev,
>  	struct iio_device_config *config)
>  {
>  	struct iio_dev_opaque *iio_dev_opaque =
> to_iio_dev_opaque(indio_dev);
> -	struct iio_buffer *buffer;
> +	struct iio_buffer *buffer, *tmp = NULL;
>  	int ret;
> 
>  	indio_dev->active_scan_mask = config->scan_mask;
> @@ -1097,8 +1097,10 @@ static int iio_enable_buffers(struct iio_dev
> *indio_dev,
> 
>  	list_for_each_entry(buffer, &iio_dev_opaque->buffer_list,
> buffer_list) {
>  		ret = iio_buffer_enable(buffer, indio_dev);
> -		if (ret)
> +		if (ret) {
> +			tmp = buffer;
>  			goto err_disable_buffers;
> +		}
>  	}
> 
>  	if (indio_dev->currentmode == INDIO_BUFFER_TRIGGERED) {
> @@ -1125,6 +1127,7 @@ static int iio_enable_buffers(struct iio_dev
> *indio_dev,
>  					     indio_dev->pollfunc);
>  	}
>  err_disable_buffers:
> +	buffer = list_prepare_entry(tmp, &iio_dev_opaque-
> >buffer_list, buffer_list);

Ok, it's Friday so I might be seeing ghosts... But looking at 'list_prepare_entry()'...
If tmp != NULL, then all goes well but if tmp == NULL, then we get

list_entry(&iio_dev_opaque->buffer_list, struct iio_buffer, buffer_list) which
would require 'struct iio_dev_opaque'. It looks like like 'list_prepare_entry()'
assumes that pos and head are of the same type... 

Am I missing something?

- Nuno Sá

Re: [PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[Jakob Koschel]

> On 1. Apr 2022, at 14:40, Sa, Nuno <Nuno.Sa@analog.com> wrote:
> 
> Hi Jakob,
> 
>> -----Original Message-----
>> From: Jakob Koschel <jakobkoschel@gmail.com>
>> Sent: Friday, April 1, 2022 1:07 AM
>> To: Jonathan Cameron <jic23@kernel.org>
>> Cc: Lars-Peter Clausen <lars@metafoo.de>; Dan Carpenter
>> <dan.carpenter@oracle.com>; Jakob Koschel
>> <jakobkoschel@gmail.com>; linux-iio@vger.kernel.org; linux-
>> kernel@vger.kernel.org; Mike Rapoport <rppt@kernel.org>; Brian
>> Johannesmeyer <bjohannesmeyer@gmail.com>; Cristiano Giuffrida
>> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
>> Subject: [PATCH 1/3] iio: buffer: remove usage of list iterator variable
>> for list_for_each_entry_continue_reverse()
>> 
>> [External]
>> 
>> In preparation to limit the scope of the list iterator variable to the
>> list traversal loop, use a dedicated pointer to iterate through the
>> list [1].
>> 
>> Since that variable should not be used past the loop iteration, a
>> separate variable is used to 'remember the current location within the
>> loop'.
>> 
>> To either continue iterating from that position or start a new
>> iteration (if the previous iteration was complete) list_prepare_entry()
>> is used.
>> 
>> Link: https://urldefense.com/v3/__https://lore.kernel.org/all/CAHk-
>> =wgRr_D8CB-D9Kg-
>> c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/__;!!A3Ni8CS0y
>> 2Y!q8llw5UCaMIsAU7tPtPDhwVor0wy032I7FJHv0VxBZksNuRJF04HjWe
>> 0XYG7OQ$  [1]
>> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
>> ---
>> drivers/iio/industrialio-buffer.c | 7 +++++--
>> 1 file changed, 5 insertions(+), 2 deletions(-)
>> 
>> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-
>> buffer.c
>> index 208b5193c621..151a77c2affd 100644
>> --- a/drivers/iio/industrialio-buffer.c
>> +++ b/drivers/iio/industrialio-buffer.c
>> @@ -1059,7 +1059,7 @@ static int iio_enable_buffers(struct iio_dev
>> *indio_dev,
>> 	struct iio_device_config *config)
>> {
>> 	struct iio_dev_opaque *iio_dev_opaque =
>> to_iio_dev_opaque(indio_dev);
>> -	struct iio_buffer *buffer;
>> +	struct iio_buffer *buffer, *tmp = NULL;
>> 	int ret;
>> 
>> 	indio_dev->active_scan_mask = config->scan_mask;
>> @@ -1097,8 +1097,10 @@ static int iio_enable_buffers(struct iio_dev
>> *indio_dev,
>> 
>> 	list_for_each_entry(buffer, &iio_dev_opaque->buffer_list,
>> buffer_list) {
>> 		ret = iio_buffer_enable(buffer, indio_dev);
>> -		if (ret)
>> +		if (ret) {
>> +			tmp = buffer;
>> 			goto err_disable_buffers;
>> +		}
>> 	}
>> 
>> 	if (indio_dev->currentmode == INDIO_BUFFER_TRIGGERED) {
>> @@ -1125,6 +1127,7 @@ static int iio_enable_buffers(struct iio_dev
>> *indio_dev,
>> 					     indio_dev->pollfunc);
>> 	}
>> err_disable_buffers:
>> +	buffer = list_prepare_entry(tmp, &iio_dev_opaque-
>>> buffer_list, buffer_list);
> 
> Ok, it's Friday so I might be seeing ghosts... But looking at 'list_prepare_entry()'...
> If tmp != NULL, then all goes well but if tmp == NULL, then we get
> 
> list_entry(&iio_dev_opaque->buffer_list, struct iio_buffer, buffer_list) which
> would require 'struct iio_dev_opaque'. It looks like like 'list_prepare_entry()'
> assumes that pos and head are of the same type... 
> 
> Am I missing something?

The list iterators are weird in this perspective...

If you look at the original code, list_for_each_entry_continue_reverse() is called on 'buffer'.

'buffer' would be a valid struct element of &iio_dev_opaque->buffer_list if the break is hit,
but if no break is hit in the earlier list_for_each_entry() buffer is not a valid entry.

Before the terminating condition of list_for_each_entry() is met, it essentially does:

	buffer = list_entry(&iio_dev_opaque->buffer_list, typeof(*buffer), buffer_list);

the buffer returned here is not a valid pointer to struct however.
But since list_for_each_entry_continue_reverse() immediately calls list_prev_entry(buffer, buffer_list)
on it you end up with the last entry of the list again and start iterating with that one.

It's a very weird design choice but since list_for_each_entry_continue_reverse() expects
a pointer to the element struct and not the list_head struct, you need to pass in this 'bogus'
pointer if you want it to start on the head element.


Keep in mind that the code here is just a more explicit version of this 'type confusion' whereas
with the original code it was just hidden within the list_for_each_entry() macro and far less obvious.
The functionality is exactly the same.

PS: list_prepare_entry() was made for this use case, so basically it is expected to craft this bogus pointer from
the head if pos == NULL.

> 
> - Nuno Sá

Thanks,
Jakob

RE: [PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[Sa, Nuno]

> From: Jakob Koschel <jakobkoschel@gmail.com>
> Sent: Friday, April 1, 2022 3:55 PM
> To: Sa, Nuno <Nuno.Sa@analog.com>
> Cc: Jonathan Cameron <jic23@kernel.org>; Lars-Peter Clausen
> <lars@metafoo.de>; Dan Carpenter <dan.carpenter@oracle.com>;
> linux-iio@vger.kernel.org; linux-kernel@vger.kernel.org; Mike
> Rapoport <rppt@kernel.org>; Brian Johannesmeyer
> <bjohannesmeyer@gmail.com>; Cristiano Giuffrida
> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
> Subject: Re: [PATCH 1/3] iio: buffer: remove usage of list iterator
> variable for list_for_each_entry_continue_reverse()
> 
> [External]
> 
> 
> 
> > On 1. Apr 2022, at 14:40, Sa, Nuno <Nuno.Sa@analog.com> wrote:
> >
> > Hi Jakob,
> >
> >> -----Original Message-----
> >> From: Jakob Koschel <jakobkoschel@gmail.com>
> >> Sent: Friday, April 1, 2022 1:07 AM
> >> To: Jonathan Cameron <jic23@kernel.org>
> >> Cc: Lars-Peter Clausen <lars@metafoo.de>; Dan Carpenter
> >> <dan.carpenter@oracle.com>; Jakob Koschel
> >> <jakobkoschel@gmail.com>; linux-iio@vger.kernel.org; linux-
> >> kernel@vger.kernel.org; Mike Rapoport <rppt@kernel.org>; Brian
> >> Johannesmeyer <bjohannesmeyer@gmail.com>; Cristiano
> Giuffrida
> >> <c.giuffrida@vu.nl>; Bos, H.J. <h.j.bos@vu.nl>
> >> Subject: [PATCH 1/3] iio: buffer: remove usage of list iterator
> variable
> >> for list_for_each_entry_continue_reverse()
> >>
> >> [External]
> >>
> >> In preparation to limit the scope of the list iterator variable to the
> >> list traversal loop, use a dedicated pointer to iterate through the
> >> list [1].
> >>
> >> Since that variable should not be used past the loop iteration, a
> >> separate variable is used to 'remember the current location within
> the
> >> loop'.
> >>
> >> To either continue iterating from that position or start a new
> >> iteration (if the previous iteration was complete)
> list_prepare_entry()
> >> is used.
> >>
> >> Link:
> https://urldefense.com/v3/__https://lore.kernel.org/all/CAHk-
> >> =wgRr_D8CB-D9Kg-
> >>
> c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/__;!!A3Ni8CS0y
> >>
> 2Y!q8llw5UCaMIsAU7tPtPDhwVor0wy032I7FJHv0VxBZksNuRJF04HjWe
> >> 0XYG7OQ$  [1]
> >> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
> >> ---
> >> drivers/iio/industrialio-buffer.c | 7 +++++--
> >> 1 file changed, 5 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-
> >> buffer.c
> >> index 208b5193c621..151a77c2affd 100644
> >> --- a/drivers/iio/industrialio-buffer.c
> >> +++ b/drivers/iio/industrialio-buffer.c
> >> @@ -1059,7 +1059,7 @@ static int iio_enable_buffers(struct
> iio_dev
> >> *indio_dev,
> >> 	struct iio_device_config *config)
> >> {
> >> 	struct iio_dev_opaque *iio_dev_opaque =
> >> to_iio_dev_opaque(indio_dev);
> >> -	struct iio_buffer *buffer;
> >> +	struct iio_buffer *buffer, *tmp = NULL;
> >> 	int ret;
> >>
> >> 	indio_dev->active_scan_mask = config->scan_mask;
> >> @@ -1097,8 +1097,10 @@ static int iio_enable_buffers(struct
> iio_dev
> >> *indio_dev,
> >>
> >> 	list_for_each_entry(buffer, &iio_dev_opaque->buffer_list,
> >> buffer_list) {
> >> 		ret = iio_buffer_enable(buffer, indio_dev);
> >> -		if (ret)
> >> +		if (ret) {
> >> +			tmp = buffer;
> >> 			goto err_disable_buffers;
> >> +		}
> >> 	}
> >>
> >> 	if (indio_dev->currentmode == INDIO_BUFFER_TRIGGERED) {
> >> @@ -1125,6 +1127,7 @@ static int iio_enable_buffers(struct
> iio_dev
> >> *indio_dev,
> >> 					     indio_dev->pollfunc);
> >> 	}
> >> err_disable_buffers:
> >> +	buffer = list_prepare_entry(tmp, &iio_dev_opaque-
> >>> buffer_list, buffer_list);
> >
> > Ok, it's Friday so I might be seeing ghosts... But looking at
> 'list_prepare_entry()'...
> > If tmp != NULL, then all goes well but if tmp == NULL, then we get
> >
> > list_entry(&iio_dev_opaque->buffer_list, struct iio_buffer,
> buffer_list) which
> > would require 'struct iio_dev_opaque'. It looks like like
> 'list_prepare_entry()'
> > assumes that pos and head are of the same type...
> >
> > Am I missing something?
> 
> The list iterators are weird in this perspective...
> 
> If you look at the original code,
> list_for_each_entry_continue_reverse() is called on 'buffer'.
> 
> 'buffer' would be a valid struct element of &iio_dev_opaque-
> >buffer_list if the break is hit,
> but if no break is hit in the earlier list_for_each_entry() buffer is not a
> valid entry.
> 
> Before the terminating condition of list_for_each_entry() is met, it
> essentially does:
> 
> 	buffer = list_entry(&iio_dev_opaque->buffer_list,
> typeof(*buffer), buffer_list);
> 
> the buffer returned here is not a valid pointer to struct however.
> But since list_for_each_entry_continue_reverse() immediately calls
> list_prev_entry(buffer, buffer_list)
> on it you end up with the last entry of the list again and start iterating
> with that one.

Ahh I see it now... I really never had noticed this dance but in fact the
same dance is also used to detect the end of list_for_each_entry()...

Ok, then:

Reviewed-by: Nuno Sá <nuno.sa@analog.com>

Re: [PATCH 1/3] iio: buffer: remove usage of list iterator variable for list_for_each_entry_continue_reverse()

[Jonathan Cameron]

On Fri,  1 Apr 2022 01:06:30 +0200
Jakob Koschel <jakobkoschel@gmail.com> wrote:

> In preparation to limit the scope of the list iterator variable to the
> list traversal loop, use a dedicated pointer to iterate through the
> list [1].
> 
> Since that variable should not be used past the loop iteration, a
> separate variable is used to 'remember the current location within the
> loop'.
> 
> To either continue iterating from that position or start a new
> iteration (if the previous iteration was complete) list_prepare_entry()
> is used.
> 
> Link: https://lore.kernel.org/all/CAHk-=wgRr_D8CB-D9Kg-c=EHreAsk5SqXPwr9Y7k9sA6cWXJ6w@mail.gmail.com/ [1]
> Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>

Hi Jakob,

Series applied to the togreg branch of iio.git and pushed out as testing to let
0-day take a poke at it. Note I'll be rebasing that branch on rc1 once available
before pushing it out for linux-next to pick up.

Note I'm happy to add additional tags if anyone else takes a look at it.

Thanks

Jonathan

> ---
>  drivers/iio/industrialio-buffer.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
> index 208b5193c621..151a77c2affd 100644
> --- a/drivers/iio/industrialio-buffer.c
> +++ b/drivers/iio/industrialio-buffer.c
> @@ -1059,7 +1059,7 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,
>  	struct iio_device_config *config)
>  {
>  	struct iio_dev_opaque *iio_dev_opaque = to_iio_dev_opaque(indio_dev);
> -	struct iio_buffer *buffer;
> +	struct iio_buffer *buffer, *tmp = NULL;
>  	int ret;
> 
>  	indio_dev->active_scan_mask = config->scan_mask;
> @@ -1097,8 +1097,10 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,
> 
>  	list_for_each_entry(buffer, &iio_dev_opaque->buffer_list, buffer_list) {
>  		ret = iio_buffer_enable(buffer, indio_dev);
> -		if (ret)
> +		if (ret) {
> +			tmp = buffer;
>  			goto err_disable_buffers;
> +		}
>  	}
> 
>  	if (indio_dev->currentmode == INDIO_BUFFER_TRIGGERED) {
> @@ -1125,6 +1127,7 @@ static int iio_enable_buffers(struct iio_dev *indio_dev,
>  					     indio_dev->pollfunc);
>  	}
>  err_disable_buffers:
> +	buffer = list_prepare_entry(tmp, &iio_dev_opaque->buffer_list, buffer_list);
>  	list_for_each_entry_continue_reverse(buffer, &iio_dev_opaque->buffer_list,
>  					     buffer_list)
>  		iio_buffer_disable(buffer, indio_dev);
> 
> base-commit: f82da161ea75dc4db21b2499e4b1facd36dab275
> --
> 2.25.1
>