AutoPage Migration - V0.1 - 6/8 hook sched migrate to memory migration
AutoPage Migration - V0.1 - 6/8 hook sched migrate to memory migration
Add check for internode migration to scheduler -- in most places where a new cpu is assigned via set_task_cpu(). If MIGRATION is configured, and sched_migrate_memory is enabled [and this is a user space task], the check will set "migration pending" for the task if the destination cpu is on a different cpu from the last cpu to which the task was assigned. Migration of affected pages [those with default policy] will occur when the task returns to user space.
kernel/sched.c +5 -0
@@ -52,6 +52,7 @@#include <asm/tlb.h> #include <asm/unistd.h> +#include <linux/auto-migrate.h> /* * Convert user-nice values [ -20 ... 0 ... 19 ]
@@ -880,6 +881,7 @@ static int migrate_task(task_t *p, int d* it is sufficient to simply update the task's cpu field. */ if (!p->array && !task_running (rq, p)) { +check_internode_migration (p, dest_cpu);set_task_cpu (p, dest_cpu);return 0 ; }
@@ -1260,6 +1262,7 @@ static int try_to_wake_up(task_t *p, uns out_set_cpu: new_cpu =wake_idle (new_cpu, p);if (new_cpu != cpu) { +check_internode_migration (p, new_cpu);set_task_cpu (p, new_cpu);task_rq_unlock (rq, &flags);/* might preempt at this point */
@@ -1778,6 +1781,7 @@ void pull_task(runqueue_t *src_rq, prio_ {dequeue_task (p, src_array); src_rq->nr_running--; +check_internode_migration (p, this_cpu);set_task_cpu (p, this_cpu); this_rq->nr_running++;enqueue_task (p, this_array);
@@ -4452,6 +4456,7 @@ static void __migrate_task(struct task_sif (!cpu_isset (dest_cpu, p->cpus_allowed)) goto out; +check_internode_migration (p, dest_cpu);set_task_cpu (p, dest_cpu);if (p->array) {/*
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 7 lines · Lee Schermerhorn
AutoPage Migration - V0.1 - 6/8 hook sched migrate to memory migration
Add check for internode migration to scheduler -- in most places where a new cpu is assigned via
set_task_cpu(). If MIGRATION is configured, andsched_migrate_memoryis enabled [and this is a user space task], the check will set "migration pending" for the task if the destination cpu is on a different cpu from the last
:-( ^-node
Quoted 5 lines
cpu to which the task was assigned. Migration of affected pages [those with default policy] will occur when the task returns to user space.
Signed-off-by Lee Schermerhorn <[email protected]>
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>