linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store()
@ 2025-10-20 11:53 Leon Hwang
  2025-10-20 13:53 ` Lorenzo Stoakes
                   ` (2 more replies)
  0 siblings, 3 replies; 5+ messages in thread
From: Leon Hwang @ 2025-10-20 11:53 UTC (permalink / raw)
  To: akpm, david, lorenzo.stoakes
  Cc: ziy, baolin.wang, Liam.Howlett, npache, ryan.roberts, dev.jain,
	baohua, lance.yang, linux-mm, linux-kernel, Leon Hwang

Both scan_sleep_millisecs_store() and alloc_sleep_millisecs_store()
perform the same operations: parse the input value, update their
respective sleep interval, reset khugepaged_sleep_expire, and wake up
the khugepaged thread.

Factor out this duplicated logic into a helper function
__sleep_millisecs_store(), and simplify both store functions.

No functional change intended.

Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
---
 mm/khugepaged.c | 29 ++++++++++++++---------------
 1 file changed, 14 insertions(+), 15 deletions(-)

diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index 0341c3d13e9e..0b7915015e9e 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -129,9 +129,8 @@ static ssize_t scan_sleep_millisecs_show(struct kobject *kobj,
 	return sysfs_emit(buf, "%u\n", khugepaged_scan_sleep_millisecs);
 }
 
-static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
-					  struct kobj_attribute *attr,
-					  const char *buf, size_t count)
+static ssize_t __sleep_millisecs_store(const char *buf, size_t count,
+				       unsigned int *millisecs)
 {
 	unsigned int msecs;
 	int err;
@@ -140,12 +139,21 @@ static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
 	if (err)
 		return -EINVAL;
 
-	khugepaged_scan_sleep_millisecs = msecs;
+	*millisecs = msecs;
 	khugepaged_sleep_expire = 0;
 	wake_up_interruptible(&khugepaged_wait);
 
 	return count;
 }
+
+static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
+					  struct kobj_attribute *attr,
+					  const char *buf, size_t count)
+{
+	unsigned int *millisecs = &khugepaged_scan_sleep_millisecs;
+
+	return __sleep_millisecs_store(buf, count, millisecs);
+}
 static struct kobj_attribute scan_sleep_millisecs_attr =
 	__ATTR_RW(scan_sleep_millisecs);
 
@@ -160,18 +168,9 @@ static ssize_t alloc_sleep_millisecs_store(struct kobject *kobj,
 					   struct kobj_attribute *attr,
 					   const char *buf, size_t count)
 {
-	unsigned int msecs;
-	int err;
-
-	err = kstrtouint(buf, 10, &msecs);
-	if (err)
-		return -EINVAL;
+	unsigned int *millisecs = &khugepaged_alloc_sleep_millisecs;
 
-	khugepaged_alloc_sleep_millisecs = msecs;
-	khugepaged_sleep_expire = 0;
-	wake_up_interruptible(&khugepaged_wait);
-
-	return count;
+	return __sleep_millisecs_store(buf, count, millisecs);
 }
 static struct kobj_attribute alloc_sleep_millisecs_attr =
 	__ATTR_RW(alloc_sleep_millisecs);
-- 
2.51.0



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store()
  2025-10-20 11:53 [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store() Leon Hwang
@ 2025-10-20 13:53 ` Lorenzo Stoakes
  2025-10-20 15:32   ` Leon Hwang
  2025-10-20 20:01 ` Zi Yan
  2025-10-21  2:14 ` Lance Yang
  2 siblings, 1 reply; 5+ messages in thread
From: Lorenzo Stoakes @ 2025-10-20 13:53 UTC (permalink / raw)
  To: Leon Hwang
  Cc: akpm, david, ziy, baolin.wang, Liam.Howlett, npache,
	ryan.roberts, dev.jain, baohua, lance.yang, linux-mm,
	linux-kernel

On Mon, Oct 20, 2025 at 07:53:50PM +0800, Leon Hwang wrote:
> Both scan_sleep_millisecs_store() and alloc_sleep_millisecs_store()
> perform the same operations: parse the input value, update their
> respective sleep interval, reset khugepaged_sleep_expire, and wake up
> the khugepaged thread.
>
> Factor out this duplicated logic into a helper function
> __sleep_millisecs_store(), and simplify both store functions.
>
> No functional change intended.
>
> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>

Thanks this is a decent cleanup, with the nit(s0 below addressed LGTM, so:

Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>

> ---
>  mm/khugepaged.c | 29 ++++++++++++++---------------
>  1 file changed, 14 insertions(+), 15 deletions(-)
>
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index 0341c3d13e9e..0b7915015e9e 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -129,9 +129,8 @@ static ssize_t scan_sleep_millisecs_show(struct kobject *kobj,
>  	return sysfs_emit(buf, "%u\n", khugepaged_scan_sleep_millisecs);
>  }
>
> -static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
> -					  struct kobj_attribute *attr,
> -					  const char *buf, size_t count)
> +static ssize_t __sleep_millisecs_store(const char *buf, size_t count,
> +				       unsigned int *millisecs)
>  {
>  	unsigned int msecs;
>  	int err;
> @@ -140,12 +139,21 @@ static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
>  	if (err)
>  		return -EINVAL;
>
> -	khugepaged_scan_sleep_millisecs = msecs;
> +	*millisecs = msecs;
>  	khugepaged_sleep_expire = 0;
>  	wake_up_interruptible(&khugepaged_wait);
>
>  	return count;
>  }
> +
> +static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
> +					  struct kobj_attribute *attr,
> +					  const char *buf, size_t count)
> +{
> +	unsigned int *millisecs = &khugepaged_scan_sleep_millisecs;
> +
> +	return __sleep_millisecs_store(buf, count, millisecs);

I think this'd be much clearer as:

return __sleep_millisecs_store(buf, count, &khugepaged_alloc_scan_millisecs);

> +}
>  static struct kobj_attribute scan_sleep_millisecs_attr =
>  	__ATTR_RW(scan_sleep_millisecs);
>
> @@ -160,18 +168,9 @@ static ssize_t alloc_sleep_millisecs_store(struct kobject *kobj,
>  					   struct kobj_attribute *attr,
>  					   const char *buf, size_t count)
>  {
> -	unsigned int msecs;
> -	int err;
> -
> -	err = kstrtouint(buf, 10, &msecs);
> -	if (err)
> -		return -EINVAL;
> +	unsigned int *millisecs = &khugepaged_alloc_sleep_millisecs;
>
> -	khugepaged_alloc_sleep_millisecs = msecs;
> -	khugepaged_sleep_expire = 0;
> -	wake_up_interruptible(&khugepaged_wait);
> -
> -	return count;
> +	return __sleep_millisecs_store(buf, count, millisecs);
>  }

I think this'd be much clearer as:

return __sleep_millisecs_store(buf, count, &khugepaged_alloc_sleep_millisecs);

>  static struct kobj_attribute alloc_sleep_millisecs_attr =
>  	__ATTR_RW(alloc_sleep_millisecs);
> --
> 2.51.0
>


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store()
  2025-10-20 13:53 ` Lorenzo Stoakes
@ 2025-10-20 15:32   ` Leon Hwang
  0 siblings, 0 replies; 5+ messages in thread
From: Leon Hwang @ 2025-10-20 15:32 UTC (permalink / raw)
  To: Lorenzo Stoakes
  Cc: akpm, david, ziy, baolin.wang, Liam.Howlett, npache,
	ryan.roberts, dev.jain, baohua, lance.yang, linux-mm,
	linux-kernel



On 2025/10/20 21:53, Lorenzo Stoakes wrote:
> On Mon, Oct 20, 2025 at 07:53:50PM +0800, Leon Hwang wrote:
>> Both scan_sleep_millisecs_store() and alloc_sleep_millisecs_store()
>> perform the same operations: parse the input value, update their
>> respective sleep interval, reset khugepaged_sleep_expire, and wake up
>> the khugepaged thread.
>>
>> Factor out this duplicated logic into a helper function
>> __sleep_millisecs_store(), and simplify both store functions.
>>
>> No functional change intended.
>>
>> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> 
> Thanks this is a decent cleanup, with the nit(s0 below addressed LGTM, so:
> 
> Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
> 

Thanks for the review, Lorenzo.

>> ---
>>  mm/khugepaged.c | 29 ++++++++++++++---------------
>>  1 file changed, 14 insertions(+), 15 deletions(-)
>>
>> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
>> index 0341c3d13e9e..0b7915015e9e 100644
>> --- a/mm/khugepaged.c
>> +++ b/mm/khugepaged.c
>> @@ -129,9 +129,8 @@ static ssize_t scan_sleep_millisecs_show(struct kobject *kobj,
>>  	return sysfs_emit(buf, "%u\n", khugepaged_scan_sleep_millisecs);
>>  }
>>
>> -static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
>> -					  struct kobj_attribute *attr,
>> -					  const char *buf, size_t count)
>> +static ssize_t __sleep_millisecs_store(const char *buf, size_t count,
>> +				       unsigned int *millisecs)
>>  {
>>  	unsigned int msecs;
>>  	int err;
>> @@ -140,12 +139,21 @@ static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
>>  	if (err)
>>  		return -EINVAL;
>>
>> -	khugepaged_scan_sleep_millisecs = msecs;
>> +	*millisecs = msecs;
>>  	khugepaged_sleep_expire = 0;
>>  	wake_up_interruptible(&khugepaged_wait);
>>
>>  	return count;
>>  }
>> +
>> +static ssize_t scan_sleep_millisecs_store(struct kobject *kobj,
>> +					  struct kobj_attribute *attr,
>> +					  const char *buf, size_t count)
>> +{
>> +	unsigned int *millisecs = &khugepaged_scan_sleep_millisecs;
>> +
>> +	return __sleep_millisecs_store(buf, count, millisecs);
> 
> I think this'd be much clearer as:
> > return __sleep_millisecs_store(buf, count,
&khugepaged_alloc_scan_millisecs);
> 
>> +}
>>  static struct kobj_attribute scan_sleep_millisecs_attr =
>>  	__ATTR_RW(scan_sleep_millisecs);
>>
>> @@ -160,18 +168,9 @@ static ssize_t alloc_sleep_millisecs_store(struct kobject *kobj,
>>  					   struct kobj_attribute *attr,
>>  					   const char *buf, size_t count)
>>  {
>> -	unsigned int msecs;
>> -	int err;
>> -
>> -	err = kstrtouint(buf, 10, &msecs);
>> -	if (err)
>> -		return -EINVAL;
>> +	unsigned int *millisecs = &khugepaged_alloc_sleep_millisecs;
>>
>> -	khugepaged_alloc_sleep_millisecs = msecs;
>> -	khugepaged_sleep_expire = 0;
>> -	wake_up_interruptible(&khugepaged_wait);
>> -
>> -	return count;
>> +	return __sleep_millisecs_store(buf, count, millisecs);
>>  }
> 
> I think this'd be much clearer as:
> 
> return __sleep_millisecs_store(buf, count, &khugepaged_alloc_sleep_millisecs);
> 

Indeed, your suggestion makes the code clearer.

I'll update both occurrences in the next revision.

Thanks,
Leon

>>  static struct kobj_attribute alloc_sleep_millisecs_attr =
>>  	__ATTR_RW(alloc_sleep_millisecs);
>> --
>> 2.51.0
>>



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store()
  2025-10-20 11:53 [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store() Leon Hwang
  2025-10-20 13:53 ` Lorenzo Stoakes
@ 2025-10-20 20:01 ` Zi Yan
  2025-10-21  2:14 ` Lance Yang
  2 siblings, 0 replies; 5+ messages in thread
From: Zi Yan @ 2025-10-20 20:01 UTC (permalink / raw)
  To: Leon Hwang
  Cc: akpm, david, lorenzo.stoakes, baolin.wang, Liam.Howlett, npache,
	ryan.roberts, dev.jain, baohua, lance.yang, linux-mm,
	linux-kernel

On 20 Oct 2025, at 7:53, Leon Hwang wrote:

> Both scan_sleep_millisecs_store() and alloc_sleep_millisecs_store()
> perform the same operations: parse the input value, update their
> respective sleep interval, reset khugepaged_sleep_expire, and wake up
> the khugepaged thread.
>
> Factor out this duplicated logic into a helper function
> __sleep_millisecs_store(), and simplify both store functions.
>
> No functional change intended.
>
> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> ---
>  mm/khugepaged.c | 29 ++++++++++++++---------------
>  1 file changed, 14 insertions(+), 15 deletions(-)
>
With the changes suggested by Lorenzo, feel free to add
Reviewed-by: Zi Yan <ziy@nvidia.com>

--
Best Regards,
Yan, Zi


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store()
  2025-10-20 11:53 [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store() Leon Hwang
  2025-10-20 13:53 ` Lorenzo Stoakes
  2025-10-20 20:01 ` Zi Yan
@ 2025-10-21  2:14 ` Lance Yang
  2 siblings, 0 replies; 5+ messages in thread
From: Lance Yang @ 2025-10-21  2:14 UTC (permalink / raw)
  To: Leon Hwang
  Cc: ziy, baolin.wang, david, Liam.Howlett, npache, ryan.roberts,
	dev.jain, baohua, linux-mm, akpm, linux-kernel, lorenzo.stoakes



On 2025/10/20 19:53, Leon Hwang wrote:
> Both scan_sleep_millisecs_store() and alloc_sleep_millisecs_store()
> perform the same operations: parse the input value, update their
> respective sleep interval, reset khugepaged_sleep_expire, and wake up
> the khugepaged thread.
> 
> Factor out this duplicated logic into a helper function
> __sleep_millisecs_store(), and simplify both store functions.

Thanks, less is more ;)

> 
> No functional change intended.
> 
> Signed-off-by: Leon Hwang <leon.hwang@linux.dev>
> ---

After tweaking as Lorenzo suggested, LGTM.

Reviewed-by: Lance Yang <lance.yang@linux.dev>


^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2025-10-21  2:15 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-10-20 11:53 [PATCH mm-new 1/1] mm/khugepaged: Factor out common logic in [scan,alloc]_sleep_millisecs_store() Leon Hwang
2025-10-20 13:53 ` Lorenzo Stoakes
2025-10-20 15:32   ` Leon Hwang
2025-10-20 20:01 ` Zi Yan
2025-10-21  2:14 ` Lance Yang

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox