1. Patchew
  2. linux
  3. View series

[PATCH] rcuscale: using kmalloc_array() to relpace kmalloc()

Su Hui posted 1 patch 7 months, 3 weeks ago 
kernel/rcu/rcuscale.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
[PATCH] rcuscale: using kmalloc_array() to relpace kmalloc()
Posted by Su Hui 7 months, 3 weeks ago 
It's safer to using kmalloc_array() because it can prevent overflow
problem.

Signed-off-by: Su Hui <suhui@nfschina.com>
---
 kernel/rcu/rcuscale.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
index 0f3059b1b80d..cbe2195f08d6 100644
--- a/kernel/rcu/rcuscale.c
+++ b/kernel/rcu/rcuscale.c
@@ -762,7 +762,7 @@ kfree_scale_thread(void *arg)
 		}
 
 		for (i = 0; i < kfree_alloc_num; i++) {
-			alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
+			alloc_ptr = kmalloc_array(kfree_mult, sizeof(struct kfree_obj), GFP_KERNEL);
 			if (!alloc_ptr)
 				return -ENOMEM;
 
-- 
2.30.2
Re: [PATCH] rcuscale: using kmalloc_array() to relpace kmalloc()
Posted by Paul E. McKenney 7 months, 3 weeks ago 
On Mon, Apr 21, 2025 at 02:15:09PM +0800, Su Hui wrote:
> It's safer to using kmalloc_array() because it can prevent overflow
> problem.
> 
> Signed-off-by: Su Hui <suhui@nfschina.com>

Thank you!

But isn't kcalloc just a wrapper around kmalloc_array() anyway?

							Thanx, Paul

> ---
>  kernel/rcu/rcuscale.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
> index 0f3059b1b80d..cbe2195f08d6 100644
> --- a/kernel/rcu/rcuscale.c
> +++ b/kernel/rcu/rcuscale.c
> @@ -762,7 +762,7 @@ kfree_scale_thread(void *arg)
>  		}
>  
>  		for (i = 0; i < kfree_alloc_num; i++) {
> -			alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
> +			alloc_ptr = kmalloc_array(kfree_mult, sizeof(struct kfree_obj), GFP_KERNEL);
>  			if (!alloc_ptr)
>  				return -ENOMEM;
>  
> -- 
> 2.30.2
> 
>
Re: [PATCH] rcuscale: using kmalloc_array() to relpace kmalloc()
Posted by Su Hui 7 months, 3 weeks ago 
On 2025/4/21 21:49, Paul E. McKenney wrote:
> On Mon, Apr 21, 2025 at 02:15:09PM +0800, Su Hui wrote:
>> It's safer to using kmalloc_array() because it can prevent overflow
>> problem.
>>
>> Signed-off-by: Su Hui <suhui@nfschina.com>
> Thank you!
>
> But isn't kcalloc just a wrapper around kmalloc_array() anyway?
Yes, and kcalloc() add the _GFP_ZERO flag. I can send a v2 patch with 
kcalloc().
Thanks for your suggestion.

Su Hui

>> ---
>>   kernel/rcu/rcuscale.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
>> index 0f3059b1b80d..cbe2195f08d6 100644
>> --- a/kernel/rcu/rcuscale.c
>> +++ b/kernel/rcu/rcuscale.c
>> @@ -762,7 +762,7 @@ kfree_scale_thread(void *arg)
>>   		}
>>   
>>   		for (i = 0; i < kfree_alloc_num; i++) {
>> -			alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
>> +			alloc_ptr = kmalloc_array(kfree_mult, sizeof(struct kfree_obj), GFP_KERNEL);
>>   			if (!alloc_ptr)
>>   				return -ENOMEM;
>>   
>> -- 
>> 2.30.2
>>
>>
[PATCH v2] rcuscale: using kcalloc() to relpace kmalloc()
Posted by Su Hui 7 months, 3 weeks ago 
It's safer to using kcalloc() because it can prevent overflow
problem.

Signed-off-by: Su Hui <suhui@nfschina.com>
---
v2: 
 - using kcalloc rather than kmalloc_array().
v1:
 - https://lore.kernel.org/all/20250421061508.718752-1-suhui@nfschina.com/

 kernel/rcu/rcuscale.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
index 0f3059b1b80d..b521d0455992 100644
--- a/kernel/rcu/rcuscale.c
+++ b/kernel/rcu/rcuscale.c
@@ -762,7 +762,7 @@ kfree_scale_thread(void *arg)
 		}
 
 		for (i = 0; i < kfree_alloc_num; i++) {
-			alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
+			alloc_ptr = kcalloc(kfree_mult, sizeof(struct kfree_obj), GFP_KERNEL);
 			if (!alloc_ptr)
 				return -ENOMEM;
 
-- 
2.30.2
Re: [PATCH v2] rcuscale: using kcalloc() to relpace kmalloc()
Posted by Paul E. McKenney 7 months, 3 weeks ago 
On Tue, Apr 22, 2025 at 09:51:45AM +0800, Su Hui wrote:
> It's safer to using kcalloc() because it can prevent overflow
> problem.
> 
> Signed-off-by: Su Hui <suhui@nfschina.com>

Reviewed-by: Paul E. McKenney <paulmck@kernel.org>

> ---
> v2: 
>  - using kcalloc rather than kmalloc_array().
> v1:
>  - https://lore.kernel.org/all/20250421061508.718752-1-suhui@nfschina.com/
> 
>  kernel/rcu/rcuscale.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
> index 0f3059b1b80d..b521d0455992 100644
> --- a/kernel/rcu/rcuscale.c
> +++ b/kernel/rcu/rcuscale.c
> @@ -762,7 +762,7 @@ kfree_scale_thread(void *arg)
>  		}
>  
>  		for (i = 0; i < kfree_alloc_num; i++) {
> -			alloc_ptr = kmalloc(kfree_mult * sizeof(struct kfree_obj), GFP_KERNEL);
> +			alloc_ptr = kcalloc(kfree_mult, sizeof(struct kfree_obj), GFP_KERNEL);
>  			if (!alloc_ptr)
>  				return -ENOMEM;
>  
> -- 
> 2.30.2
>
Re: [PATCH v2] rcuscale: using kcalloc() to relpace kmalloc()
Posted by Joel Fernandes 7 months, 3 weeks ago 

On 4/22/2025 1:58 PM, Paul E. McKenney wrote:
> On Tue, Apr 22, 2025 at 09:51:45AM +0800, Su Hui wrote:
>> It's safer to using kcalloc() because it can prevent overflow
>> problem.
>>
>> Signed-off-by: Su Hui <suhui@nfschina.com>
> 
> Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Applied, thanks.

Patchew 2.1-devel-b67e4c2

© 2016 - 2025 Red Hat, Inc.