by making sure that vcpus only run on the pcpu(s) they are allowed to
run on based on their hard affinity cpu masks.
Signed-off-by: Justin T. Weaver
---
Changes in v4:
* Renamed scratch_mask to _scratch_mask
* Renamed csched2_cpumask to scratch_mask
* Removed "else continue" i
Hello,
The credit2 vcpu scheduler currently ignores per-vcpu hard and soft affinity
masks.
In the v3 review I was asked by George Dunlap to split up the soft affinity
patch into multiple patches by function and only resend the changes to two of
them, get_fallback_cpu and runq_tickle, so this seri
Move affinity balancing related functions and defines from sched_credit.c to
sched-if.h so other schedulers can use them. Change name prefixes from csched
to sched since they are no longer specific to the credit scheduler. No
functional changes intended.
Signed-off-by: Justin T. Weaver
Reviewed
s soft affinity. If the
vcpu on the found cpu has less credit than the given vcpu, then that cpu is
chosen. Finally, if no cpu was picked yet, the second step looks for the
non-idle, non-tickled cpu with the lowest credit in the vcpu's hard affinity.
Signed-off-by: Justin T. Weaver
---
Chang
Move macro VCPU2ONLINE from schedule.c to sched.h so it can be used by other
source files.
Signed-off-by: Justin T. Weaver
---
xen/common/schedule.c |1 -
xen/include/xen/sched.h |2 ++
2 files changed, 2 insertions(+), 1 deletion(-)
diff --git a/xen/common/schedule.c b/xen/common
x27;s cpupool, and in the vcpu's soft
affinity
hard affinity step...
4) another pcpu from the vcpu's current runq in the vcpu's hard affinity
3) an online pcpu in the vcpu's domain's cpupool, and in the vcpu's hard
affinity
Signed-off-by: Justin T. Weaver
---
C
-by: Justin T. Weaver
---
Changes in v3: First introduced in patch series version 3
---
xen/common/sched_credit2.c | 152 ++--
1 file changed, 76 insertions(+), 76 deletions(-)
diff --git a/xen/common/sched_credit2.c b/xen/common/sched_credit2.c
index
Hello,
The credit2 vcpu scheduler currently ignores per-vcpu hard and soft affinity
masks.
The first patch updates the scheduler to ensure that vcpus only run
on pcpus on which they are allowed to run (hard affinity). I tested it using
xl vcpu-pin and xl vcpu-list. I changed the affinity in diffe
by making sure that vcpus only run on the pcpu(s) they are allowed to
run on based on their hard affinity cpu masks.
Signed-off-by: Justin T. Weaver
---
Changes in v3:
(all v3 changes are based on v2 review comments)
* Renamed cpumask to scratch_mask
* Renamed function get_safe_pcpu to
Signed-off-by: Justin T. Weaver
---
Changes in v3:
* get_fallback_cpu: added balance loop to try to find a soft affinity cpu
* runq_tickle: replaced use of local var mask with csched2_cpumask
* runq_tickle: added two balance loops, one for finding idle, but not
tickled, and other for findin
Move affinity balancing related functions and defines from sched_credit.c to
sched-if.h so other schedulers can use them. Change name prefixes from csched
to sched since they are no longer specific to the credit scheduler.
Signed-off-by: Justin T. Weaver
---
Changes in v3: First introduced in
From: "Justin T. Weaver"
by making sure that vcpus only run on the pcpu(s) they are allowed to
run on based on their hard affinity cpu masks.
Signed-off-by: Justin T. Weaver
---
Changes in v2:
* Added dynamically allocated cpu masks to avoid putting them on the stack;
replaced
by making sure that vcpus only run on the pcpu(s) they are allowed to
run on based on their hard affinity cpu masks.
Signed-off-by: Justin T. Weaver
---
xen/common/sched_credit2.c | 199 +---
1 file changed, 171 insertions(+), 28 deletions(-)
diff --git
Hello,
The credit2 scheduler currently ignores per-vcpu hard and soft affinity
masks.
The first patch updates the scheduler to ensure that vcpus only run
on pcpus on which they are allowed to run (hard affinity).
The second patch updates the scheduler to make more informed run queue
assignment d
that order.
Second, in function balance_load, if moving a vcpu with soft affinity from
one run queue to another means moving it away from its soft affinity to hard
affinity only, then that move is not considered for load balancing.
Signed-off-by: Justin T. Weaver
---
xen/common/sched_credit2.c
15 matches
Mail list logo
