Depending on the value of `speed` and how fast our backends are,
delay_ns might be 0 very, very often. This creates some warning
messages that spook users, but it's also pretty inefficient.

Use block_job_throttle instead to yield a little more intelligently.

Signed-off-by: John Snow <jsnow@redhat.com>
---
 block/commit.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/block/commit.c b/block/commit.c
index 873e749d50..567064215b 100644
--- a/block/commit.c
+++ b/block/commit.c
@@ -172,7 +172,8 @@ static void coroutine_fn commit_run(void *opaque)
         /* Note that even when no rate limit is applied we need to yield
          * with no pending I/O here so that bdrv_drain_all() returns.
          */
-        block_job_sleep_ns(&s->common, delay_ns);
+        block_job_throttle(&s->common, delay_ns);
+
         if (block_job_is_cancelled(&s->common)) {
             break;
         }
-- 
2.14.3

On 14/12/2017 01:59, John Snow wrote:
> Depending on the value of `speed` and how fast our backends are,
> delay_ns might be 0 very, very often. This creates some warning
> messages that spook users, but it's also pretty inefficient.
> 
> Use block_job_throttle instead to yield a little more intelligently.
> 
> Signed-off-by: John Snow <jsnow@redhat.com>
> ---
>  block/commit.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/block/commit.c b/block/commit.c
> index 873e749d50..567064215b 100644
> --- a/block/commit.c
> +++ b/block/commit.c
> @@ -172,7 +172,8 @@ static void coroutine_fn commit_run(void *opaque)
>          /* Note that even when no rate limit is applied we need to yield
>           * with no pending I/O here so that bdrv_drain_all() returns.
>           */
> -        block_job_sleep_ns(&s->common, delay_ns);
> +        block_job_throttle(&s->common, delay_ns);
> +
>          if (block_job_is_cancelled(&s->common)) {
>              break;
>          }
> 

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

On Wed, Dec 13, 2017 at 07:59:51PM -0500, John Snow wrote:
> Depending on the value of `speed` and how fast our backends are,
> delay_ns might be 0 very, very often. This creates some warning
> messages that spook users, but it's also pretty inefficient.
> 
> Use block_job_throttle instead to yield a little more intelligently.
> 
> Signed-off-by: John Snow <jsnow@redhat.com>
> ---
>  block/commit.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>