If priority is changed, continue checking workload_expires and service tree
count of previous workload is meaningless. We should always choose the workload
with lowest key of new priority in such case.
Signed-off-by: Shaohua Li <[email protected]>
---
block/cfq-iosched.c | 5 +++++
1 file changed, 5 insertions(+)
Index: linux/block/cfq-iosched.c
===================================================================
--- linux.orig/block/cfq-iosched.c 2010-12-08 19:39:01.000000000 +0800
+++ linux/block/cfq-iosched.c 2010-12-09 10:27:24.000000000 +0800
@@ -2116,6 +2116,7 @@ static void choose_service_tree(struct c
unsigned count;
struct cfq_rb_root *st;
unsigned group_slice;
+ enum wl_prio_t original_prio = cfqd->serving_prio;
if (!cfqg) {
cfqd->serving_prio = IDLE_WORKLOAD;
@@ -2134,6 +2135,9 @@ static void choose_service_tree(struct c
return;
}
+ if (original_prio != cfqd->serving_prio)
+ goto new_workload;
+
/*
* For RT and BE, we have to choose also the type
* (SYNC, SYNC_NOIDLE, ASYNC), and to compute a workload
@@ -2148,6 +2152,7 @@ static void choose_service_tree(struct c
if (count && !time_after(jiffies, cfqd->workload_expires))
return;
+new_workload:
/* otherwise select new workload type */
cfqd->serving_type =
cfq_choose_wl(cfqd, cfqg, cfqd->serving_prio);
Shaohua Li <[email protected]> writes:
> If priority is changed, continue checking workload_expires and service tree
> count of previous workload is meaningless. We should always choose the workload
I'd change that first sentence to:
If priority is changed, continuing to check workload_expires and
service tree count of the previous workload does not make sense.
> with lowest key of new priority in such case.
>
> Signed-off-by: Shaohua Li <[email protected]>
This makes sense to me.
Reviewed-by: Jeff Moyer <[email protected]>
On Fri, 2010-12-10 at 22:56 +0800, Jeff Moyer wrote:
> Shaohua Li <[email protected]> writes:
>
> > If priority is changed, continue checking workload_expires and service tree
> > count of previous workload is meaningless. We should always choose the workload
>
> I'd change that first sentence to:
>
> If priority is changed, continuing to check workload_expires and
> service tree count of the previous workload does not make sense.
>
> > with lowest key of new priority in such case.
> >
> > Signed-off-by: Shaohua Li <[email protected]>
>
> This makes sense to me.
>
> Reviewed-by: Jeff Moyer <[email protected]>
Thanks, modified the the changelog.
If priority is changed, continuing to check workload_expires and service tree
count of the previous workload does not make sense. We should always choose
the workload with lowest key of new priority in such case.
Signed-off-by: Shaohua Li <[email protected]>
Reviewed-by: Jeff Moyer <[email protected]>
---
block/cfq-iosched.c | 5 +++++
1 file changed, 5 insertions(+)
Index: linux/block/cfq-iosched.c
===================================================================
--- linux.orig/block/cfq-iosched.c 2010-12-08 19:39:01.000000000 +0800
+++ linux/block/cfq-iosched.c 2010-12-09 10:27:24.000000000 +0800
@@ -2116,6 +2116,7 @@ static void choose_service_tree(struct c
unsigned count;
struct cfq_rb_root *st;
unsigned group_slice;
+ enum wl_prio_t original_prio = cfqd->serving_prio;
if (!cfqg) {
cfqd->serving_prio = IDLE_WORKLOAD;
@@ -2134,6 +2135,9 @@ static void choose_service_tree(struct c
return;
}
+ if (original_prio != cfqd->serving_prio)
+ goto new_workload;
+
/*
* For RT and BE, we have to choose also the type
* (SYNC, SYNC_NOIDLE, ASYNC), and to compute a workload
@@ -2148,6 +2152,7 @@ static void choose_service_tree(struct c
if (count && !time_after(jiffies, cfqd->workload_expires))
return;
+new_workload:
/* otherwise select new workload type */
cfqd->serving_type =
cfq_choose_wl(cfqd, cfqg, cfqd->serving_prio);
On 2010-12-13 01:39, Shaohua Li wrote:
> On Fri, 2010-12-10 at 22:56 +0800, Jeff Moyer wrote:
>> Shaohua Li <[email protected]> writes:
>>
>>> If priority is changed, continue checking workload_expires and service tree
>>> count of previous workload is meaningless. We should always choose the workload
>>
>> I'd change that first sentence to:
>>
>> If priority is changed, continuing to check workload_expires and
>> service tree count of the previous workload does not make sense.
>>
>>> with lowest key of new priority in such case.
>>>
>>> Signed-off-by: Shaohua Li <[email protected]>
>>
>> This makes sense to me.
>>
>> Reviewed-by: Jeff Moyer <[email protected]>
> Thanks, modified the the changelog.
Thanks, applied to for-2.6.38/core
--
Jens Axboe