cfq-iosched: Fixed boot warning with BLK_CGROUP=y and CFQ_GROUP_IOSCHED=n
Vivek Goyal [Fri, 18 Jun 2010 14:39:47 +0000 (10:39 -0400)]
Hi Jens,

Few days back Ingo noticed a CFQ boot time warning. This patch fixes it.
The issue here is that with CFQ_GROUP_IOSCHED=n, CFQ should not really
be making blkio stat related calls.

> Hm, it's still not entirely fixed, as of 2.6.35-rc2-00131-g7908a9e. With
> some
> configs i get bad spinlock warnings during bootup:
>
> [   28.968013] initcall net_olddevs_init+0x0/0x82 returned 0 after 93750
> usecs
> [   28.972003] calling  b44_init+0x0/0x55 @ 1
> [   28.976009] bus: 'pci': add driver b44
> [   28.976374]  sda:
> [   28.978157] BUG: spinlock bad magic on CPU#1, async/0/117
> [   28.980000]  lock: 7e1c5bbc, .magic: 00000000, .owner: <none>/-1, +.owner_cpu: 0
> [   28.980000] Pid: 117, comm: async/0 Not tainted +2.6.35-rc2-tip-01092-g010e7ef-dirty #8183
> [   28.980000] Call Trace:
> [   28.980000]  [<41ba6d55>] ? printk+0x20/0x24
> [   28.980000]  [<4134b7b7>] spin_bug+0x7c/0x87
> [   28.980000]  [<4134b853>] do_raw_spin_lock+0x1e/0x123
> [   28.980000]  [<41ba92ca>] ? _raw_spin_lock_irqsave+0x12/0x20
> [   28.980000]  [<41ba92d2>] _raw_spin_lock_irqsave+0x1a/0x20
> [   28.980000]  [<4133476f>] blkiocg_update_io_add_stats+0x25/0xfb
> [   28.980000]  [<41335dae>] ? cfq_prio_tree_add+0xb1/0xc1
> [   28.980000]  [<41337bc7>] cfq_insert_request+0x8c/0x425

Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>

block/cfq-iosched.c
block/cfq.h [new file with mode: 0644]

index 153f627..7982b83 100644 (file)
@@ -14,7 +14,7 @@
 #include <linux/rbtree.h>
 #include <linux/ioprio.h>
 #include <linux/blktrace_api.h>
-#include "blk-cgroup.h"
+#include "cfq.h"
 
 /*
  * tunables
@@ -879,7 +879,7 @@ cfq_group_service_tree_del(struct cfq_data *cfqd, struct cfq_group *cfqg)
        if (!RB_EMPTY_NODE(&cfqg->rb_node))
                cfq_rb_erase(&cfqg->rb_node, st);
        cfqg->saved_workload_slice = 0;
-       blkiocg_update_dequeue_stats(&cfqg->blkg, 1);
+       cfq_blkiocg_update_dequeue_stats(&cfqg->blkg, 1);
 }
 
 static inline unsigned int cfq_cfqq_slice_usage(struct cfq_queue *cfqq)
@@ -939,8 +939,8 @@ static void cfq_group_served(struct cfq_data *cfqd, struct cfq_group *cfqg,
 
        cfq_log_cfqg(cfqd, cfqg, "served: vt=%llu min_vt=%llu", cfqg->vdisktime,
                                        st->min_vdisktime);
-       blkiocg_update_timeslice_used(&cfqg->blkg, used_sl);
-       blkiocg_set_start_empty_time(&cfqg->blkg);
+       cfq_blkiocg_update_timeslice_used(&cfqg->blkg, used_sl);
+       cfq_blkiocg_set_start_empty_time(&cfqg->blkg);
 }
 
 #ifdef CONFIG_CFQ_GROUP_IOSCHED
@@ -995,7 +995,7 @@ cfq_find_alloc_cfqg(struct cfq_data *cfqd, struct cgroup *cgroup, int create)
 
        /* Add group onto cgroup list */
        sscanf(dev_name(bdi->dev), "%u:%u", &major, &minor);
-       blkiocg_add_blkio_group(blkcg, &cfqg->blkg, (void *)cfqd,
+       cfq_blkiocg_add_blkio_group(blkcg, &cfqg->blkg, (void *)cfqd,
                                        MKDEV(major, minor));
        cfqg->weight = blkcg_get_weight(blkcg, cfqg->blkg.dev);
 
@@ -1079,7 +1079,7 @@ static void cfq_release_cfq_groups(struct cfq_data *cfqd)
                 * it from cgroup list, then it will take care of destroying
                 * cfqg also.
                 */
-               if (!blkiocg_del_blkio_group(&cfqg->blkg))
+               if (!cfq_blkiocg_del_blkio_group(&cfqg->blkg))
                        cfq_destroy_cfqg(cfqd, cfqg);
        }
 }
@@ -1421,10 +1421,10 @@ static void cfq_reposition_rq_rb(struct cfq_queue *cfqq, struct request *rq)
 {
        elv_rb_del(&cfqq->sort_list, rq);
        cfqq->queued[rq_is_sync(rq)]--;
-       blkiocg_update_io_remove_stats(&(RQ_CFQG(rq))->blkg, rq_data_dir(rq),
-                                               rq_is_sync(rq));
+       cfq_blkiocg_update_io_remove_stats(&(RQ_CFQG(rq))->blkg,
+                                       rq_data_dir(rq), rq_is_sync(rq));
        cfq_add_rq_rb(rq);
-       blkiocg_update_io_add_stats(&(RQ_CFQG(rq))->blkg,
+       cfq_blkiocg_update_io_add_stats(&(RQ_CFQG(rq))->blkg,
                        &cfqq->cfqd->serving_group->blkg, rq_data_dir(rq),
                        rq_is_sync(rq));
 }
@@ -1482,8 +1482,8 @@ static void cfq_remove_request(struct request *rq)
        cfq_del_rq_rb(rq);
 
        cfqq->cfqd->rq_queued--;
-       blkiocg_update_io_remove_stats(&(RQ_CFQG(rq))->blkg, rq_data_dir(rq),
-                                               rq_is_sync(rq));
+       cfq_blkiocg_update_io_remove_stats(&(RQ_CFQG(rq))->blkg,
+                                       rq_data_dir(rq), rq_is_sync(rq));
        if (rq_is_meta(rq)) {
                WARN_ON(!cfqq->meta_pending);
                cfqq->meta_pending--;
@@ -1518,8 +1518,8 @@ static void cfq_merged_request(struct request_queue *q, struct request *req,
 static void cfq_bio_merged(struct request_queue *q, struct request *req,
                                struct bio *bio)
 {
-       blkiocg_update_io_merged_stats(&(RQ_CFQG(req))->blkg, bio_data_dir(bio),
-                                       cfq_bio_sync(bio));
+       cfq_blkiocg_update_io_merged_stats(&(RQ_CFQG(req))->blkg,
+                                       bio_data_dir(bio), cfq_bio_sync(bio));
 }
 
 static void
@@ -1539,8 +1539,8 @@ cfq_merged_requests(struct request_queue *q, struct request *rq,
        if (cfqq->next_rq == next)
                cfqq->next_rq = rq;
        cfq_remove_request(next);
-       blkiocg_update_io_merged_stats(&(RQ_CFQG(rq))->blkg, rq_data_dir(next),
-                                       rq_is_sync(next));
+       cfq_blkiocg_update_io_merged_stats(&(RQ_CFQG(rq))->blkg,
+                                       rq_data_dir(next), rq_is_sync(next));
 }
 
 static int cfq_allow_merge(struct request_queue *q, struct request *rq,
@@ -1571,7 +1571,7 @@ static int cfq_allow_merge(struct request_queue *q, struct request *rq,
 static inline void cfq_del_timer(struct cfq_data *cfqd, struct cfq_queue *cfqq)
 {
        del_timer(&cfqd->idle_slice_timer);
-       blkiocg_update_idle_time_stats(&cfqq->cfqg->blkg);
+       cfq_blkiocg_update_idle_time_stats(&cfqq->cfqg->blkg);
 }
 
 static void __cfq_set_active_queue(struct cfq_data *cfqd,
@@ -1580,7 +1580,7 @@ static void __cfq_set_active_queue(struct cfq_data *cfqd,
        if (cfqq) {
                cfq_log_cfqq(cfqd, cfqq, "set_active wl_prio:%d wl_type:%d",
                                cfqd->serving_prio, cfqd->serving_type);
-               blkiocg_update_avg_queue_size_stats(&cfqq->cfqg->blkg);
+               cfq_blkiocg_update_avg_queue_size_stats(&cfqq->cfqg->blkg);
                cfqq->slice_start = 0;
                cfqq->dispatch_start = jiffies;
                cfqq->allocated_slice = 0;
@@ -1911,7 +1911,7 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd)
        sl = cfqd->cfq_slice_idle;
 
        mod_timer(&cfqd->idle_slice_timer, jiffies + sl);
-       blkiocg_update_set_idle_time_stats(&cfqq->cfqg->blkg);
+       cfq_blkiocg_update_set_idle_time_stats(&cfqq->cfqg->blkg);
        cfq_log_cfqq(cfqd, cfqq, "arm_idle: %lu", sl);
 }
 
@@ -1931,7 +1931,7 @@ static void cfq_dispatch_insert(struct request_queue *q, struct request *rq)
        elv_dispatch_sort(q, rq);
 
        cfqd->rq_in_flight[cfq_cfqq_sync(cfqq)]++;
-       blkiocg_update_dispatch_stats(&cfqq->cfqg->blkg, blk_rq_bytes(rq),
+       cfq_blkiocg_update_dispatch_stats(&cfqq->cfqg->blkg, blk_rq_bytes(rq),
                                        rq_data_dir(rq), rq_is_sync(rq));
 }
 
@@ -3257,7 +3257,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
                                cfq_clear_cfqq_wait_request(cfqq);
                                __blk_run_queue(cfqd->queue);
                        } else {
-                               blkiocg_update_idle_time_stats(
+                               cfq_blkiocg_update_idle_time_stats(
                                                &cfqq->cfqg->blkg);
                                cfq_mark_cfqq_must_dispatch(cfqq);
                        }
@@ -3285,7 +3285,7 @@ static void cfq_insert_request(struct request_queue *q, struct request *rq)
        rq_set_fifo_time(rq, jiffies + cfqd->cfq_fifo_expire[rq_is_sync(rq)]);
        list_add_tail(&rq->queuelist, &cfqq->fifo);
        cfq_add_rq_rb(rq);
-       blkiocg_update_io_add_stats(&(RQ_CFQG(rq))->blkg,
+       cfq_blkiocg_update_io_add_stats(&(RQ_CFQG(rq))->blkg,
                        &cfqd->serving_group->blkg, rq_data_dir(rq),
                        rq_is_sync(rq));
        cfq_rq_enqueued(cfqd, cfqq, rq);
@@ -3373,9 +3373,9 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq)
        WARN_ON(!cfqq->dispatched);
        cfqd->rq_in_driver--;
        cfqq->dispatched--;
-       blkiocg_update_completion_stats(&cfqq->cfqg->blkg, rq_start_time_ns(rq),
-                       rq_io_start_time_ns(rq), rq_data_dir(rq),
-                       rq_is_sync(rq));
+       cfq_blkiocg_update_completion_stats(&cfqq->cfqg->blkg,
+                       rq_start_time_ns(rq), rq_io_start_time_ns(rq),
+                       rq_data_dir(rq), rq_is_sync(rq));
 
        cfqd->rq_in_flight[cfq_cfqq_sync(cfqq)]--;
 
@@ -3739,7 +3739,7 @@ static void cfq_exit_queue(struct elevator_queue *e)
 
        cfq_put_async_queues(cfqd);
        cfq_release_cfq_groups(cfqd);
-       blkiocg_del_blkio_group(&cfqd->root_group.blkg);
+       cfq_blkiocg_del_blkio_group(&cfqd->root_group.blkg);
 
        spin_unlock_irq(q->queue_lock);
 
@@ -3807,8 +3807,8 @@ static void *cfq_init_queue(struct request_queue *q)
         */
        atomic_set(&cfqg->ref, 1);
        rcu_read_lock();
-       blkiocg_add_blkio_group(&blkio_root_cgroup, &cfqg->blkg, (void *)cfqd,
-                                       0);
+       cfq_blkiocg_add_blkio_group(&blkio_root_cgroup, &cfqg->blkg,
+                                       (void *)cfqd, 0);
        rcu_read_unlock();
 #endif
        /*
diff --git a/block/cfq.h b/block/cfq.h
new file mode 100644 (file)
index 0000000..8783318
--- /dev/null
@@ -0,0 +1,115 @@
+#ifndef _CFQ_H
+#define _CFQ_H
+#include "blk-cgroup.h"
+
+#ifdef CONFIG_CFQ_GROUP_IOSCHED
+static inline void cfq_blkiocg_update_io_add_stats(struct blkio_group *blkg,
+       struct blkio_group *curr_blkg, bool direction, bool sync)
+{
+       blkiocg_update_io_add_stats(blkg, curr_blkg, direction, sync);
+}
+
+static inline void cfq_blkiocg_update_dequeue_stats(struct blkio_group *blkg,
+                       unsigned long dequeue)
+{
+       blkiocg_update_dequeue_stats(blkg, dequeue);
+}
+
+static inline void cfq_blkiocg_update_timeslice_used(struct blkio_group *blkg,
+                       unsigned long time)
+{
+       blkiocg_update_timeslice_used(blkg, time);
+}
+
+static inline void cfq_blkiocg_set_start_empty_time(struct blkio_group *blkg)
+{
+       blkiocg_set_start_empty_time(blkg);
+}
+
+static inline void cfq_blkiocg_update_io_remove_stats(struct blkio_group *blkg,
+                               bool direction, bool sync)
+{
+       blkiocg_update_io_remove_stats(blkg, direction, sync);
+}
+
+static inline void cfq_blkiocg_update_io_merged_stats(struct blkio_group *blkg,
+               bool direction, bool sync)
+{
+       blkiocg_update_io_merged_stats(blkg, direction, sync);
+}
+
+static inline void cfq_blkiocg_update_idle_time_stats(struct blkio_group *blkg)
+{
+       blkiocg_update_idle_time_stats(blkg);
+}
+
+static inline void
+cfq_blkiocg_update_avg_queue_size_stats(struct blkio_group *blkg)
+{
+       blkiocg_update_avg_queue_size_stats(blkg);
+}
+
+static inline void
+cfq_blkiocg_update_set_idle_time_stats(struct blkio_group *blkg)
+{
+       blkiocg_update_set_idle_time_stats(blkg);
+}
+
+static inline void cfq_blkiocg_update_dispatch_stats(struct blkio_group *blkg,
+                               uint64_t bytes, bool direction, bool sync)
+{
+       blkiocg_update_dispatch_stats(blkg, bytes, direction, sync);
+}
+
+static inline void cfq_blkiocg_update_completion_stats(struct blkio_group *blkg, uint64_t start_time, uint64_t io_start_time, bool direction, bool sync)
+{
+       cfq_blkiocg_update_completion_stats(blkg, start_time, io_start_time,
+                               direction, sync);
+}
+
+static inline void cfq_blkiocg_add_blkio_group(struct blkio_cgroup *blkcg,
+                       struct blkio_group *blkg, void *key, dev_t dev) {
+       blkiocg_add_blkio_group(blkcg, blkg, key, dev);
+}
+
+static inline int cfq_blkiocg_del_blkio_group(struct blkio_group *blkg)
+{
+       return blkiocg_del_blkio_group(blkg);
+}
+
+#else /* CFQ_GROUP_IOSCHED */
+static inline void cfq_blkiocg_update_io_add_stats(struct blkio_group *blkg,
+       struct blkio_group *curr_blkg, bool direction, bool sync) {}
+
+static inline void cfq_blkiocg_update_dequeue_stats(struct blkio_group *blkg,
+                       unsigned long dequeue) {}
+
+static inline void cfq_blkiocg_update_timeslice_used(struct blkio_group *blkg,
+                       unsigned long time) {}
+static inline void cfq_blkiocg_set_start_empty_time(struct blkio_group *blkg) {}
+static inline void cfq_blkiocg_update_io_remove_stats(struct blkio_group *blkg,
+                               bool direction, bool sync) {}
+static inline void cfq_blkiocg_update_io_merged_stats(struct blkio_group *blkg,
+               bool direction, bool sync) {}
+static inline void cfq_blkiocg_update_idle_time_stats(struct blkio_group *blkg)
+{
+}
+static inline void
+cfq_blkiocg_update_avg_queue_size_stats(struct blkio_group *blkg) {}
+
+static inline void
+cfq_blkiocg_update_set_idle_time_stats(struct blkio_group *blkg) {}
+
+static inline void cfq_blkiocg_update_dispatch_stats(struct blkio_group *blkg,
+                               uint64_t bytes, bool direction, bool sync) {}
+static inline void cfq_blkiocg_update_completion_stats(struct blkio_group *blkg, uint64_t start_time, uint64_t io_start_time, bool direction, bool sync) {}
+
+static inline void cfq_blkiocg_add_blkio_group(struct blkio_cgroup *blkcg,
+                       struct blkio_group *blkg, void *key, dev_t dev) {}
+static inline int cfq_blkiocg_del_blkio_group(struct blkio_group *blkg)
+{
+       return 0;
+}
+
+#endif /* CFQ_GROUP_IOSCHED */
+#endif