[PATCH 21/21] blk: mq: Convert to hotplug state machine

From: Sebastian Andrzej Siewior
Date: Tue Sep 06 2016 - 13:08:12 EST


Install the callbacks via the state machine.

Cc: Jens Axboe <axboe@xxxxxxxxx>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
---
This patch against linux-next to take the block changes into account. It
depends on "blk: mq: reserve hotplug ID states for block"

block/blk-mq.c | 86 ++++++++++++++++++++++++++++-----------------------------
1 file changed, 43 insertions(+), 43 deletions(-)

--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2173,47 +2173,15 @@ static void blk_mq_queue_reinit(struct r
blk_mq_sysfs_register(q);
}

-static int blk_mq_queue_reinit_notify(struct notifier_block *nb,
- unsigned long action, void *hcpu)
+/*
+ * New online cpumask which is going to be set in this hotplug event.
+ * Declare this cpumasks as global as cpu-hotplug operation is invoked
+ * one-by-one and dynamically allocating this could result in a failure.
+ */
+static struct cpumask cpuhp_online_new;
+static int blk_mq_queue_reinit_work(void)
{
struct request_queue *q;
- int cpu = (unsigned long)hcpu;
- /*
- * New online cpumask which is going to be set in this hotplug event.
- * Declare this cpumasks as global as cpu-hotplug operation is invoked
- * one-by-one and dynamically allocating this could result in a failure.
- */
- static struct cpumask online_new;
-
- /*
- * Before hotadded cpu starts handling requests, new mappings must
- * be established. Otherwise, these requests in hw queue might
- * never be dispatched.
- *
- * For example, there is a single hw queue (hctx) and two CPU queues
- * (ctx0 for CPU0, and ctx1 for CPU1).
- *
- * Now CPU1 is just onlined and a request is inserted into
- * ctx1->rq_list and set bit0 in pending bitmap as ctx1->index_hw is
- * still zero.
- *
- * And then while running hw queue, flush_busy_ctxs() finds bit0 is
- * set in pending bitmap and tries to retrieve requests in
- * hctx->ctxs[0]->rq_list. But htx->ctxs[0] is a pointer to ctx0,
- * so the request in ctx1->rq_list is ignored.
- */
- switch (action & ~CPU_TASKS_FROZEN) {
- case CPU_DEAD:
- case CPU_UP_CANCELED:
- cpumask_copy(&online_new, cpu_online_mask);
- break;
- case CPU_UP_PREPARE:
- cpumask_copy(&online_new, cpu_online_mask);
- cpumask_set_cpu(cpu, &online_new);
- break;
- default:
- return NOTIFY_OK;
- }

mutex_lock(&all_q_mutex);

@@ -2237,13 +2205,44 @@ static int blk_mq_queue_reinit_notify(st
}

list_for_each_entry(q, &all_q_list, all_q_node)
- blk_mq_queue_reinit(q, &online_new);
+ blk_mq_queue_reinit(q, &cpuhp_online_new);

list_for_each_entry(q, &all_q_list, all_q_node)
blk_mq_unfreeze_queue(q);

mutex_unlock(&all_q_mutex);
- return NOTIFY_OK;
+ return 0;
+}
+
+static int blk_mq_queue_reinit_dead(unsigned int cpu)
+{
+ cpumask_copy(&cpuhp_online_new, cpu_online_mask);
+ return blk_mq_queue_reinit_work();
+}
+
+static int blk_mq_queue_reinit_prepare(unsigned int cpu)
+{
+ /*
+ * Before hotadded cpu starts handling requests, new mappings must
+ * be established. Otherwise, these requests in hw queue might
+ * never be dispatched.
+ *
+ * For example, there is a single hw queue (hctx) and two CPU queues
+ * (ctx0 for CPU0, and ctx1 for CPU1).
+ *
+ * Now CPU1 is just onlined and a request is inserted into
+ * ctx1->rq_list and set bit0 in pending bitmap as ctx1->index_hw is
+ * still zero.
+ *
+ * And then while running hw queue, flush_busy_ctxs() finds bit0 is
+ * set in pending bitmap and tries to retrieve requests in
+ * hctx->ctxs[0]->rq_list. But htx->ctxs[0] is a pointer to ctx0,
+ * so the request in ctx1->rq_list is ignored.
+ */
+
+ cpumask_copy(&cpuhp_online_new, cpu_online_mask);
+ cpumask_set_cpu(cpu, &cpuhp_online_new);
+ return blk_mq_queue_reinit_work();
}

static int __blk_mq_alloc_rq_maps(struct blk_mq_tag_set *set)
@@ -2447,8 +2446,9 @@ static int __init blk_mq_init(void)
{
blk_mq_cpu_init();

- hotcpu_notifier(blk_mq_queue_reinit_notify, 0);
-
+ cpuhp_setup_state_nocalls(CPUHP_BLK_MQ_PREPARE, "block/mq:prepare",
+ blk_mq_queue_reinit_prepare,
+ blk_mq_queue_reinit_dead);
return 0;
}
subsys_initcall(blk_mq_init);