Commit 70a27d6d authored by Ingo Molnar's avatar Ingo Molnar
Browse files

sched/balancing: Rename run_rebalance_domains() => sched_balance_softirq() 



run_rebalance_domains() is a misnomer, as it doesn't only
run rebalance_domains(), but since the introduction of the
NOHZ code it also runs nohz_idle_balance().

Rename it to sched_balance_softirq(), reflecting its more
generic purpose and that it's a softirq handler.

Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
Reviewed-by: default avatarValentin Schneider <vschneid@redhat.com>
Reviewed-by: default avatarShrikanth Hegde <sshegde@linux.ibm.com>
Link: https://lore.kernel.org/r/20240308111819.1101550-2-mingo@kernel.org
parent 33928ed8

Documentation/scheduler/sched-domains.rst

+1 −1
Original line number Diff line number Diff line
@@ -34,7 +34,7 @@ out of balance are tasks moved between groups. 
In kernel/sched/core.c, trigger_load_balance() is run periodically on each CPU

through scheduler_tick(). It raises a softirq after the next regularly scheduled

rebalancing event for the current runqueue has arrived. The actual load

balancing workhorse, run_rebalance_domains()->rebalance_domains(), is then run

balancing workhorse, sched_balance_softirq()->rebalance_domains(), is then run

in softirq context (SCHED_SOFTIRQ).



The latter function takes two arguments: the runqueue of current CPU and whether

Documentation/translations/zh_CN/scheduler/sched-domains.rst

+1 −1
Original line number Diff line number Diff line
@@ -36,7 +36,7 @@ CPU共享。任意两个组的CPU掩码的交集不一定为空，如果是这 


在kernel/sched/core.c中，trigger_load_balance()在每个CPU上通过scheduler_tick()

周期执行。在当前运行队列下一个定期调度再平衡事件到达后，它引发一个软中断。负载均衡真正

的工作由run_rebalance_domains()->rebalance_domains()完成，在软中断上下文中执行

的工作由sched_balance_softirq()->rebalance_domains()完成，在软中断上下文中执行

（SCHED_SOFTIRQ）。



后一个函数有两个入参：当前CPU的运行队列、它在scheduler_tick()调用时是否空闲。函数会从

kernel/sched/fair.c

+2 −2
Original line number Diff line number Diff line
@@ -12415,7 +12415,7 @@ static int newidle_balance(struct rq *this_rq, struct rq_flags *rf) 
 * - indirectly from a remote scheduler_tick() for NOHZ idle balancing

 *   through the SMP cross-call nohz_csd_func()

 */

static __latent_entropy void run_rebalance_domains(struct softirq_action *h)

static __latent_entropy void sched_balance_softirq(struct softirq_action *h)

{

	struct rq *this_rq = this_rq();

	enum cpu_idle_type idle = this_rq->idle_balance;
@@ -13216,7 +13216,7 @@ __init void init_sched_fair_class(void) 
#endif

	}



	open_softirq(SCHED_SOFTIRQ, run_rebalance_domains);

	open_softirq(SCHED_SOFTIRQ, sched_balance_softirq);



#ifdef CONFIG_NO_HZ_COMMON

	nohz.next_balance = jiffies;