linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] blk-iolatency:postpone ktime_get() execution until blk_iolatency_enabled() return true
@ 2020-07-01  6:04 Hongnan Li
  2020-07-01  7:52 ` Johannes Thumshirn
  0 siblings, 1 reply; 2+ messages in thread
From: Hongnan Li @ 2020-07-01  6:04 UTC (permalink / raw)
  To: linux-block; +Cc: axboe, Hongnan Li

ktime_to_ns(ktime_get()) which is expensive do not need to be executed if
blk_iolatency_enabled() return false in blkcg_iolatency_done_bio().
Postponing ktime_to_ns(ktime_get()) execution can reduce CPU usage
when blk_iolatency was disabled.

Signed-off-by: Hongnan Li <hongnan.li@linux.alibaba.com>
---
 block/blk-iolatency.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/block/blk-iolatency.c b/block/blk-iolatency.c
index c128d50..8c1487e 100644
--- a/block/blk-iolatency.c
+++ b/block/blk-iolatency.c
@@ -591,7 +591,7 @@ static void blkcg_iolatency_done_bio(struct rq_qos *rqos, struct bio *bio)
 	struct rq_wait *rqw;
 	struct iolatency_grp *iolat;
 	u64 window_start;
-	u64 now = ktime_to_ns(ktime_get());
+
 	bool issue_as_root = bio_issue_as_root_blkg(bio);
 	bool enabled = false;
 	int inflight = 0;
@@ -608,6 +608,7 @@ static void blkcg_iolatency_done_bio(struct rq_qos *rqos, struct bio *bio)
 	if (!enabled)
 		return;
 
+	u64 now = ktime_to_ns(ktime_get());
 	while (blkg && blkg->parent) {
 		iolat = blkg_to_lat(blkg);
 		if (!iolat) {
-- 
1.8.3.1


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] blk-iolatency:postpone ktime_get() execution until blk_iolatency_enabled() return true
  2020-07-01  6:04 [PATCH] blk-iolatency:postpone ktime_get() execution until blk_iolatency_enabled() return true Hongnan Li
@ 2020-07-01  7:52 ` Johannes Thumshirn
  0 siblings, 0 replies; 2+ messages in thread
From: Johannes Thumshirn @ 2020-07-01  7:52 UTC (permalink / raw)
  To: Hongnan Li, linux-block; +Cc: axboe

On 01/07/2020 08:05, Hongnan Li wrote:
> ktime_to_ns(ktime_get()) which is expensive do not need to be executed if
> blk_iolatency_enabled() return false in blkcg_iolatency_done_bio().
> Postponing ktime_to_ns(ktime_get()) execution can reduce CPU usage
> when blk_iolatency was disabled.
> 
> Signed-off-by: Hongnan Li <hongnan.li@linux.alibaba.com>
> ---
>  block/blk-iolatency.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/block/blk-iolatency.c b/block/blk-iolatency.c
> index c128d50..8c1487e 100644
> --- a/block/blk-iolatency.c
> +++ b/block/blk-iolatency.c
> @@ -591,7 +591,7 @@ static void blkcg_iolatency_done_bio(struct rq_qos *rqos, struct bio *bio)
>  	struct rq_wait *rqw;
>  	struct iolatency_grp *iolat;
>  	u64 window_start;
> -	u64 now = ktime_to_ns(ktime_get());
> +
>  	bool issue_as_root = bio_issue_as_root_blkg(bio);
>  	bool enabled = false;
>  	int inflight = 0;
> @@ -608,6 +608,7 @@ static void blkcg_iolatency_done_bio(struct rq_qos *rqos, struct bio *bio)
>  	if (!enabled)
>  		return;
>  
> +	u64 now = ktime_to_ns(ktime_get());
>  	while (blkg && blkg->parent) {
>  		iolat = blkg_to_lat(blkg);
>  		if (!iolat) {
> 

You're mixing declarations and code.

please do
-	u64 now = ktime_to_ns(ktime_get());
+	u64 now;

[...]
+	now = ktime_to_ns(ktime_get());

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-07-01  7:52 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-07-01  6:04 [PATCH] blk-iolatency:postpone ktime_get() execution until blk_iolatency_enabled() return true Hongnan Li
2020-07-01  7:52 ` Johannes Thumshirn

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).