2021-07-14 11:15:37

by Christophe JAILLET

Subject: [PATCH 1/2] iio: buffer: Save a few cycles in 'iio_scan_mask_set()'

Use 'bitmap_alloc()' instead of 'bitmap_zalloc()' because the bitmap
is fully overridden by a 'bitmap_copy()' call just after its allocation.

While at it, fix the style of a NULL check.

Signed-off-by: Christophe JAILLET <[email protected]>
---
drivers/iio/industrialio-buffer.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
index fdd623407b96..6d4776a7f002 100644
--- a/drivers/iio/industrialio-buffer.c
+++ b/drivers/iio/industrialio-buffer.c
@@ -354,8 +354,8 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
const unsigned long *mask;
unsigned long *trialmask;

- trialmask = bitmap_zalloc(indio_dev->masklength, GFP_KERNEL);
- if (trialmask == NULL)
+ trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
+ if (!trialmask)
return -ENOMEM;
if (!indio_dev->masklength) {
WARN(1, "Trying to set scanmask prior to registering buffer\n");
--
2.30.2

2021-07-14 11:16:04

by Christophe JAILLET

Subject: [PATCH 2/2] iio: buffer: Move a sanity check at the beginning of 'iio_scan_mask_set()'

This is more standard to have sanity checks at the entry of a function,
instead of allocating some memory first and having to free it if a
condition is not met.

Shuffle code a bit to check 'masklength' before calling 'bitmap_alloc()'

Signed-off-by: Christophe JAILLET <[email protected]>
---
drivers/iio/industrialio-buffer.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
index 6d4776a7f002..a95cc2da56be 100644
--- a/drivers/iio/industrialio-buffer.c
+++ b/drivers/iio/industrialio-buffer.c
@@ -354,13 +354,14 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
const unsigned long *mask;
unsigned long *trialmask;

- trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
- if (!trialmask)
- return -ENOMEM;
if (!indio_dev->masklength) {
WARN(1, "Trying to set scanmask prior to registering buffer\n");
- goto err_invalid_mask;
+ return -EINVAL;
}
+
+ trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
+ if (!trialmask)
+ return -ENOMEM;
bitmap_copy(trialmask, buffer->scan_mask, indio_dev->masklength);
set_bit(bit, trialmask);

--
2.30.2

2021-07-17 16:57:36

by Jonathan Cameron

Subject: Re: [PATCH 1/2] iio: buffer: Save a few cycles in 'iio_scan_mask_set()'

On Wed, 14 Jul 2021 13:14:41 +0200
Christophe JAILLET <[email protected]> wrote:

> Use 'bitmap_alloc()' instead of 'bitmap_zalloc()' because the bitmap
> is fully overridden by a 'bitmap_copy()' call just after its allocation.
>
> While at it, fix the style of a NULL check.
>
> Signed-off-by: Christophe JAILLET <[email protected]>
Applied to the togreg branch of iio.git and pushed out as testing for the autobuilders
to poke at it.

Thanks,

Jonathan

> ---
> drivers/iio/industrialio-buffer.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
> index fdd623407b96..6d4776a7f002 100644
> --- a/drivers/iio/industrialio-buffer.c
> +++ b/drivers/iio/industrialio-buffer.c
> @@ -354,8 +354,8 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
> const unsigned long *mask;
> unsigned long *trialmask;
>
> - trialmask = bitmap_zalloc(indio_dev->masklength, GFP_KERNEL);
> - if (trialmask == NULL)
> + trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
> + if (!trialmask)
> return -ENOMEM;
> if (!indio_dev->masklength) {
> WARN(1, "Trying to set scanmask prior to registering buffer\n");

2021-07-17 16:58:06

by Jonathan Cameron

Subject: Re: [PATCH 2/2] iio: buffer: Move a sanity check at the beginning of 'iio_scan_mask_set()'

On Wed, 14 Jul 2021 13:14:51 +0200
Christophe JAILLET <[email protected]> wrote:

> This is more standard to have sanity checks at the entry of a function,
> instead of allocating some memory first and having to free it if a
> condition is not met.

Indeed this is silly. I'd guess result of code evolution, but perhaps
I was crazy in the first place :)

Applied,

Thanks,

Jonathan

>
> Shuffle code a bit to check 'masklength' before calling 'bitmap_alloc()'
>
> Signed-off-by: Christophe JAILLET <[email protected]>
> ---
> drivers/iio/industrialio-buffer.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/iio/industrialio-buffer.c b/drivers/iio/industrialio-buffer.c
> index 6d4776a7f002..a95cc2da56be 100644
> --- a/drivers/iio/industrialio-buffer.c
> +++ b/drivers/iio/industrialio-buffer.c
> @@ -354,13 +354,14 @@ static int iio_scan_mask_set(struct iio_dev *indio_dev,
> const unsigned long *mask;
> unsigned long *trialmask;
>
> - trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
> - if (!trialmask)
> - return -ENOMEM;
> if (!indio_dev->masklength) {
> WARN(1, "Trying to set scanmask prior to registering buffer\n");
> - goto err_invalid_mask;
> + return -EINVAL;
> }
> +
> + trialmask = bitmap_alloc(indio_dev->masklength, GFP_KERNEL);
> + if (!trialmask)
> + return -ENOMEM;
> bitmap_copy(trialmask, buffer->scan_mask, indio_dev->masklength);
> set_bit(bit, trialmask);
>