Replace the manual sequence of cpumask_next() and cpumask_first()
with a single call to cpumask_next_wrap() in get_next_cpu().

Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
---
 kernel/bpf/bpf_lru_list.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/kernel/bpf/bpf_lru_list.c b/kernel/bpf/bpf_lru_list.c
index 2d6e1c98d8ad..34881f4da8ae 100644
--- a/kernel/bpf/bpf_lru_list.c
+++ b/kernel/bpf/bpf_lru_list.c
@@ -21,10 +21,7 @@
 
 static int get_next_cpu(int cpu)
 {
-	cpu = cpumask_next(cpu, cpu_possible_mask);
-	if (cpu >= nr_cpu_ids)
-		cpu = cpumask_first(cpu_possible_mask);
-	return cpu;
+	return cpumask_next_wrap(cpu, cpu_possible_mask);
 }
 
 /* Local list helpers */
-- 
2.36.1

On 8/7/25 4:48 AM, Fushuai Wang wrote:
> Replace the manual sequence of cpumask_next() and cpumask_first()
> with a single call to cpumask_next_wrap() in get_next_cpu().
> 
> Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
> ---
>   kernel/bpf/bpf_lru_list.c | 5 +----
>   1 file changed, 1 insertion(+), 4 deletions(-)
> 
> diff --git a/kernel/bpf/bpf_lru_list.c b/kernel/bpf/bpf_lru_list.c
> index 2d6e1c98d8ad..34881f4da8ae 100644
> --- a/kernel/bpf/bpf_lru_list.c
> +++ b/kernel/bpf/bpf_lru_list.c
> @@ -21,10 +21,7 @@
>   
>   static int get_next_cpu(int cpu)
>   {
> -	cpu = cpumask_next(cpu, cpu_possible_mask);
> -	if (cpu >= nr_cpu_ids)
> -		cpu = cpumask_first(cpu_possible_mask);
> -	return cpu;
> +	return cpumask_next_wrap(cpu, cpu_possible_mask);
>   }

Lets then get rid of the get_next_cpu() function since its only used
once, and just use the cpumask_next_wrap() at call site ?

[...]
                 raw_spin_unlock_irqrestore(&steal_loc_l->lock, flags);

                 steal = cpumask_next_wrap(steal, cpu_possible_mask);
         } while (!node && steal != first_steal);
[...]

Btw, in $subj please target [PATCH bpf-next] given its a cleanup,
not a fix.

Thanks,
Daniel

>> Replace the manual sequence of cpumask_next() and cpumask_first()
>> with a single call to cpumask_next_wrap() in get_next_cpu().
>> 
>> Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
>> ---
>>   kernel/bpf/bpf_lru_list.c | 5 +----
>>   1 file changed, 1 insertion(+), 4 deletions(-)
>> 
>> diff --git a/kernel/bpf/bpf_lru_list.c b/kernel/bpf/bpf_lru_list.c
>> index 2d6e1c98d8ad..34881f4da8ae 100644
>> --- a/kernel/bpf/bpf_lru_list.c
>> +++ b/kernel/bpf/bpf_lru_list.c
>> @@ -21,10 +21,7 @@
>>   
>>   static int get_next_cpu(int cpu)
>>   {
>> -	cpu = cpumask_next(cpu, cpu_possible_mask);
>> -	if (cpu >= nr_cpu_ids)
>> -		cpu = cpumask_first(cpu_possible_mask);
>> -	return cpu;
>> +	return cpumask_next_wrap(cpu, cpu_possible_mask);
>>   }
> 
> Lets then get rid of the get_next_cpu() function since its only used
> once, and just use the cpumask_next_wrap() at call site ?
> 
> [...]
>                  raw_spin_unlock_irqrestore(&steal_loc_l->lock, flags);
> 
>                  steal = cpumask_next_wrap(steal, cpu_possible_mask);
>          } while (!node && steal != first_steal);
> [...]
>
 
Thank you for your suggestion.

> Btw, in $subj please target [PATCH bpf-next] given its a cleanup,
> not a fix.

I will send a v2 shortly.

Regards,
Wang.

On 8/6/25 7:48 PM, Fushuai Wang wrote:
> Replace the manual sequence of cpumask_next() and cpumask_first()
> with a single call to cpumask_next_wrap() in get_next_cpu().
>
> Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>

Acked-by: Yonghong Song <yonghong.song@linux.dev>