[PATCH] sched/pelt: use shift operation instead of division operation

[PATCH] sched/pelt: use shift operation instead of division operation

[qiwuchen55]

From: chenqiwu <chenqiwu@xiaomi.com>

Use shift operation to calculate the periods instead of division,
since shift operation is more efficient than division operation.

Signed-off-by: chenqiwu <chenqiwu@xiaomi.com>
---
 kernel/sched/pelt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/pelt.c b/kernel/sched/pelt.c
index bd006b7..ac79f8e 100644
--- a/kernel/sched/pelt.c
+++ b/kernel/sched/pelt.c
@@ -114,7 +114,7 @@ static u32 __accumulate_pelt_segments(u64 periods, u32 d1, u32 d3)
 	u64 periods;
 
 	delta += sa->period_contrib;
-	periods = delta / 1024; /* A period is 1024us (~1ms) */
+	periods = delta >> 10; /* A period is 1024us (~1ms) */
 
 	/*
 	 * Step 1: decay old *_sum if we crossed period boundaries.
-- 
1.9.1

Re: [PATCH] sched/pelt: use shift operation instead of division operation

[Peter Zijlstra]

On Mon, Feb 24, 2020 at 11:30:55AM +0800, qiwuchen55@gmail.com wrote:
> From: chenqiwu <chenqiwu@xiaomi.com>
> 
> Use shift operation to calculate the periods instead of division,
> since shift operation is more efficient than division operation.
> 
> Signed-off-by: chenqiwu <chenqiwu@xiaomi.com>
> ---
>  kernel/sched/pelt.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/sched/pelt.c b/kernel/sched/pelt.c
> index bd006b7..ac79f8e 100644
> --- a/kernel/sched/pelt.c
> +++ b/kernel/sched/pelt.c
> @@ -114,7 +114,7 @@ static u32 __accumulate_pelt_segments(u64 periods, u32 d1, u32 d3)
>  	u64 periods;
>  
>  	delta += sa->period_contrib;
> -	periods = delta / 1024; /* A period is 1024us (~1ms) */
> +	periods = delta >> 10; /* A period is 1024us (~1ms) */

Find me a compiler that stupid.