Skip to contentSkip to conversation

gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

PATCH Applied Andy Shevchenko 3 participants 7 messages Sep 4, 2024 – Sep 5, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

Use the mem_is_zero() helper where possible.

---
 drivers/gpio/gpiolib-cdev.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpio/gpiolib-cdev.c +4 -4
index 5aac59de0d76..e98d75dd8acd 100644
@@ -16,7 +16,6 @@
 #include <linux/hte.h>
 #include <linux/interrupt.h>
 #include <linux/irqreturn.h>
-#include <linux/kernel.h>
 #include <linux/kfifo.h>
 #include <linux/module.h>
 #include <linux/mutex.h>
@@ -26,6 +25,7 @@
 #include <linux/rbtree.h>
 #include <linux/seq_file.h>
 #include <linux/spinlock.h>
+#include <linux/string.h>
 #include <linux/timekeeping.h>
 #include <linux/uaccess.h>
 #include <linux/workqueue.h>
@@ -1331,7 +1331,7 @@ static int gpio_v2_line_config_validate(struct gpio_v2_line_config *lc,
 	if (lc->num_attrs > GPIO_V2_LINE_NUM_ATTRS_MAX)
 		return -EINVAL;
 
-	if (memchr_inv(lc->padding, 0, sizeof(lc->padding)))
+	if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
 		return -EINVAL;
 
 	for (i = 0; i < num_lines; i++) {
@@ -1746,7 +1746,7 @@ static int linereq_create(struct gpio_device *gdev, void __user *ip)
 	if ((ulr.num_lines == 0) || (ulr.num_lines > GPIO_V2_LINES_MAX))
 		return -EINVAL;
 
-	if (memchr_inv(ulr.padding, 0, sizeof(ulr.padding)))
+	if (!mem_is_zero(ulr.padding, sizeof(ulr.padding)))
 		return -EINVAL;
 
 	lc = &ulr.config;
@@ -2516,7 +2516,7 @@ static int lineinfo_get(struct gpio_chardev_data *cdev, void __user *ip,
 	if (copy_from_user(&lineinfo, ip, sizeof(lineinfo)))
 		return -EFAULT;
 
-	if (memchr_inv(lineinfo.padding, 0, sizeof(lineinfo.padding)))
+	if (!mem_is_zero(lineinfo.padding, sizeof(lineinfo.padding)))
 		return -EINVAL;
 
 	desc = gpio_device_get_desc(cdev->gdev, lineinfo.offset);
Signature · 1 line
2.46.0
reviewer Acked-by on v1 1/1
#1 lore raw
Quoted 3 lines · Andy Shevchenko

Use the mem_is_zero() helper where possible.

Somehow I still don't receive mail directly from you, so I only picked this up via the list. Weird. It should be gmail to gmail, right - WTH? Unless it isn't - and that is the problem.

I'm ok with the change, though mem_is_zero() hasn't hit any of the trees I have in front of me, and I had to find the corresponding patch on lore. Given this is explicitly the use case for it, I would've embedded the negation in the function and the name, so mem_not_zero() rather than !mem_is_zero(), as that reads better for me, but ok.

Quoted 54 lines
---
 drivers/gpio/gpiolib-cdev.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpio/gpiolib-cdev.c +4 -5
index 5aac59de0d76..e98d75dd8acd 100644
@@ -16,7 +16,6 @@
 #include <linux/hte.h>
 #include <linux/interrupt.h>
 #include <linux/irqreturn.h>
-#include <linux/kernel.h>
 #include <linux/kfifo.h>
 #include <linux/module.h>
 #include <linux/mutex.h>
@@ -26,6 +25,7 @@
 #include <linux/rbtree.h>
 #include <linux/seq_file.h>
 #include <linux/spinlock.h>
+#include <linux/string.h>
 #include <linux/timekeeping.h>
 #include <linux/uaccess.h>
 #include <linux/workqueue.h>
@@ -1331,7 +1331,7 @@ static int gpio_v2_line_config_validate(struct gpio_v2_line_config *lc,
 	if (lc->num_attrs > GPIO_V2_LINE_NUM_ATTRS_MAX)
 		return -EINVAL;
 
-	if (memchr_inv(lc->padding, 0, sizeof(lc->padding)))
+	if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
 		return -EINVAL;
 
 	for (i = 0; i < num_lines; i++) {
@@ -1746,7 +1746,7 @@ static int linereq_create(struct gpio_device *gdev, void __user *ip)
 	if ((ulr.num_lines == 0) || (ulr.num_lines > GPIO_V2_LINES_MAX))
 		return -EINVAL;
 
-	if (memchr_inv(ulr.padding, 0, sizeof(ulr.padding)))
+	if (!mem_is_zero(ulr.padding, sizeof(ulr.padding)))
 		return -EINVAL;
 
 	lc = &ulr.config;
@@ -2516,7 +2516,7 @@ static int lineinfo_get(struct gpio_chardev_data *cdev, void __user *ip,
 	if (copy_from_user(&lineinfo, ip, sizeof(lineinfo)))
 		return -EFAULT;
 
-	if (memchr_inv(lineinfo.padding, 0, sizeof(lineinfo.padding)))
+	if (!mem_is_zero(lineinfo.padding, sizeof(lineinfo.padding)))
 		return -EINVAL;
 
 	desc = gpio_device_get_desc(cdev->gdev, lineinfo.offset);
--

2.46.0

reviewer
#3 lore raw
Quoted 4 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

On Wed, 04 Sep 2024 12:07:43 +0300, Andy Shevchenko wrote:

Quoted 9 lines
Quoted 3 lines

Use the mem_is_zero() helper where possible.

Applied, thanks!

[1/1] gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

      commit: b1da870ba36b3f525aee9be35b2f08a1feec61a7

As per my other mail, mem_is_zero() is not defined in gpio/for-next yet. So how does this work?

If I build for-next I get:

drivers/gpio/gpiolib-cdev.c: In function ‘gpio_v2_line_config_validate’:
drivers/gpio/gpiolib-cdev.c:1334:14: error: implicit declaration of function ‘mem_is_zero’; did you mean ‘xa_is_zero’? [-Werror=implicit-function-declaration]

 1334 |         if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
      |              ^~~~~~~~~~~
      |              xa_is_zero

Cheers,
Kent.

#4 lore raw
Quoted 23 lines · Kent Gibson
Quoted 4 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

On Wed, 04 Sep 2024 12:07:43 +0300, Andy Shevchenko wrote:

Quoted 9 lines
Quoted 3 lines

Use the mem_is_zero() helper where possible.

Applied, thanks!

[1/1] gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

      commit: b1da870ba36b3f525aee9be35b2f08a1feec61a7

As per my other mail, mem_is_zero() is not defined in gpio/for-next yet. So how does this work?

If I build for-next I get:

drivers/gpio/gpiolib-cdev.c: In function ‘gpio_v2_line_config_validate’:
drivers/gpio/gpiolib-cdev.c:1334:14: error: implicit declaration of function ‘mem_is_zero’; did you mean ‘xa_is_zero’? [-Werror=implicit-function-declaration]

Quoted 7 lines
 1334 |         if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
      |              ^~~~~~~~~~~
      |              xa_is_zero

Cheers,
Kent.

Ah, it'll have to wait until the next release cycle then. Sorry for vacuuming patches too eagerly today. Dropping it.

Bart

reviewer
#5 lore raw
Quoted 29 lines · Bartosz Golaszewski
Quoted 28 lines · Kent Gibson
Quoted 4 lines · Bartosz Golaszewski

From: Bartosz Golaszewski <[email protected]>

On Wed, 04 Sep 2024 12:07:43 +0300, Andy Shevchenko wrote:

Quoted 9 lines
Quoted 3 lines

Use the mem_is_zero() helper where possible.

Applied, thanks!

[1/1] gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

      commit: b1da870ba36b3f525aee9be35b2f08a1feec61a7

As per my other mail, mem_is_zero() is not defined in gpio/for-next yet. So how does this work?

If I build for-next I get:

drivers/gpio/gpiolib-cdev.c: In function ‘gpio_v2_line_config_validate’:
drivers/gpio/gpiolib-cdev.c:1334:14: error: implicit declaration of function ‘mem_is_zero’; did you mean ‘xa_is_zero’? [-Werror=implicit-function-declaration]

 1334 |         if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
      |              ^~~~~~~~~~~
      |              xa_is_zero
Quoted 6 lines
Quoted 2 lines

Cheers,
Kent.

Ah, it'll have to wait until the next release cycle then. Sorry for vacuuming patches too eagerly today. Dropping it.

My bad for not being clearer in my initial mail - that was why I acked rather than giving it a Rb - it still needs you to work out how to deal with the other patch that it requires.

Well that and I would flip the sense of the function but, assuming that ship has sailed, that is just grumbling.

Cheers,
Kent.

#6 lore raw
In reply to Bartosz Golaszewski
Quoted 15 lines · Bartosz Golaszewski
Quoted 14 lines · Kent Gibson
Quoted 7 lines · Bartosz Golaszewski
Quoted 1 line · Andy Shevchenko

Use the mem_is_zero() helper where possible.

Applied, thanks!

[1/1] gpiolib: cdev: use !mem_is_zero() instead of memchr_inv(s, 0, n)

      commit: b1da870ba36b3f525aee9be35b2f08a1feec61a7

As per my other mail, mem_is_zero() is not defined in gpio/for-next yet. So how does this work?

If I build for-next I get:

Quoted 8 lines
Quoted 5 lines

drivers/gpio/gpiolib-cdev.c: In function ‘gpio_v2_line_config_validate’:
drivers/gpio/gpiolib-cdev.c:1334:14: error: implicit declaration of function ‘mem_is_zero’; did you mean ‘xa_is_zero’? [-Werror=implicit-function-declaration]

 1334 |         if (!mem_is_zero(lc->padding, sizeof(lc->padding)))
      |              ^~~~~~~~~~~
      |              xa_is_zero

Ah, it'll have to wait until the next release cycle then. Sorry for vacuuming patches too eagerly today. Dropping it.

Hmm... Somehow I missed that it appeared only on Aug 19 in Linux Next.

Signature · 2 lines
With Best Regards,
Andy Shevchenko