bcache: add CACHE_SET_IO_DISABLE to struct cache_set flags

When too many I/Os failed on cache device, bch_cache_set_error() is called
in the error handling code path to retire whole problematic cache set. If
new I/O requests continue to come and take refcount dc->count, the cache
set won't be retired immediately, this is a problem.

Further more, there are several kernel thread and self-armed kernel work
may still running after bch_cache_set_error() is called. It needs to wait
quite a while for them to stop, or they won't stop at all. They also
prevent the cache set from being retired.

The solution in this patch is, to add per cache set flag to disable I/O
request on this cache and all attached backing devices. Then new coming I/O
requests can be rejected in *_make_request() before taking refcount, kernel
threads and self-armed kernel worker can stop very fast when flags bit
CACHE_SET_IO_DISABLE is set.

Because bcache also do internal I/Os for writeback, garbage collection,
bucket allocation, journaling, this kind of I/O should be disabled after
bch_cache_set_error() is called. So closure_bio_submit() is modified to
check whether CACHE_SET_IO_DISABLE is set on cache_set->flags. If set,
closure_bio_submit() will set bio->bi_status to BLK_STS_IOERR and
return, generic_make_request() won't be called.

A sysfs interface is also added to set or clear CACHE_SET_IO_DISABLE bit
from cache_set->flags, to disable or enable cache set I/O for debugging. It
is helpful to trigger more corner case issues for failed cache device.

Changelog
v4, add wait_for_kthread_stop(), and call it before exits writeback and gc
    kernel threads.
v3, change CACHE_SET_IO_DISABLE from 4 to 3, since it is bit index.
    remove "bcache: " prefix when printing out kernel message.
v2, more changes by previous review,
- Use CACHE_SET_IO_DISABLE of cache_set->flags, suggested by Junhui.
- Check CACHE_SET_IO_DISABLE in bch_btree_gc() to stop a while-loop, this
  is reported and inspired from origal patch of Pavel Vazharov.
v1, initial version.

Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Michael Lyle <mlyle@lyle.org>
Cc: Junhui Tang <tang.junhui@zte.com.cn>
Cc: Michael Lyle <mlyle@lyle.org>
Cc: Pavel Vazharov <freakpv@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

10 files changed, 116 insertions, 30 deletions

diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c
index 458e1d38577d..004cc3cc6123 100644
--- a/drivers/md/bcache/alloc.c
+++ b/drivers/md/bcache/alloc.c
@@ -287,7 +287,8 @@ do {									\
 			break;						\
 									\
 		mutex_unlock(&(ca)->set->bucket_lock);			\
-		if (kthread_should_stop()) {				\
+		if (kthread_should_stop() ||				\
+		    test_bit(CACHE_SET_IO_DISABLE, &ca->set->flags)) {	\
 			set_current_state(TASK_RUNNING);		\
 			return 0;					\
 		}							\
diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h
index b5ddb848cd31..8a0327581d62 100644
--- a/drivers/md/bcache/bcache.h
+++ b/drivers/md/bcache/bcache.h
@@ -188,6 +188,7 @@
 #include <linux/refcount.h>
 #include <linux/types.h>
 #include <linux/workqueue.h>
+#include <linux/kthread.h>
 
 #include "bset.h"
 #include "util.h"
@@ -475,10 +476,15 @@ struct gc_stat {
  *
  * CACHE_SET_RUNNING means all cache devices have been registered and journal
  * replay is complete.
+ *
+ * CACHE_SET_IO_DISABLE is set when bcache is stopping the whold cache set, all
+ * external and internal I/O should be denied when this flag is set.
+ *
  */
 #define CACHE_SET_UNREGISTERING		0
 #define	CACHE_SET_STOPPING		1
 #define	CACHE_SET_RUNNING		2
+#define CACHE_SET_IO_DISABLE		3
 
 struct cache_set {
 	struct closure		cl;
@@ -868,6 +874,33 @@ static inline void wake_up_allocators(struct cache_set *c)
 		wake_up_process(ca->alloc_thread);
 }
 
+static inline void closure_bio_submit(struct cache_set *c,
+				      struct bio *bio,
+				      struct closure *cl)
+{
+	closure_get(cl);
+	if (unlikely(test_bit(CACHE_SET_IO_DISABLE, &c->flags))) {
+		bio->bi_status = BLK_STS_IOERR;
+		bio_endio(bio);
+		return;
+	}
+	generic_make_request(bio);
+}
+
+/*
+ * Prevent the kthread exits directly, and make sure when kthread_stop()
+ * is called to stop a kthread, it is still alive. If a kthread might be
+ * stopped by CACHE_SET_IO_DISABLE bit set, wait_for_kthread_stop() is
+ * necessary before the kthread returns.
+ */
+static inline void wait_for_kthread_stop(void)
+{
+	while (!kthread_should_stop()) {
+		set_current_state(TASK_INTERRUPTIBLE);
+		schedule();
+	}
+}
+
 /* Forward declarations */
 
 void bch_count_io_errors(struct cache *, blk_status_t, int, const char *);
diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c
index fad9fe8817eb..39cc8a549091 100644
--- a/drivers/md/bcache/btree.c
+++ b/drivers/md/bcache/btree.c
@@ -1744,6 +1744,7 @@ static void bch_btree_gc(struct cache_set *c)
 
 	btree_gc_start(c);
 
+	/* if CACHE_SET_IO_DISABLE set, gc thread should stop too */
 	do {
 		ret = btree_root(gc_root, c, &op, &writes, &stats);
 		closure_sync(&writes);
@@ -1751,7 +1752,7 @@ static void bch_btree_gc(struct cache_set *c)
 
 		if (ret && ret != -EAGAIN)
 			pr_warn("gc failed!");
-	} while (ret);
+	} while (ret && !test_bit(CACHE_SET_IO_DISABLE, &c->flags));
 
 	bch_btree_gc_finish(c);
 	wake_up_allocators(c);
@@ -1789,15 +1790,19 @@ static int bch_gc_thread(void *arg)
 
 	while (1) {
 		wait_event_interruptible(c->gc_wait,
-			   kthread_should_stop() || gc_should_run(c));
+			   kthread_should_stop() ||
+			   test_bit(CACHE_SET_IO_DISABLE, &c->flags) ||
+			   gc_should_run(c));
 
-		if (kthread_should_stop())
+		if (kthread_should_stop() ||
+		    test_bit(CACHE_SET_IO_DISABLE, &c->flags))
 			break;
 
 		set_gc_sectors(c);
 		bch_btree_gc(c);
 	}
 
+	wait_for_kthread_stop();
 	return 0;
 }
 
diff --git a/drivers/md/bcache/io.c b/drivers/md/bcache/io.c
index a783c5a41ff1..8013ecbcdbda 100644
--- a/drivers/md/bcache/io.c
+++ b/drivers/md/bcache/io.c
@@ -38,7 +38,7 @@ void __bch_submit_bbio(struct bio *bio, struct cache_set *c)
 	bio_set_dev(bio, PTR_CACHE(c, &b->key, 0)->bdev);
 
 	b->submit_time_us = local_clock_us();
-	closure_bio_submit(bio, bio->bi_private);
+	closure_bio_submit(c, bio, bio->bi_private);
 }
 
 void bch_submit_bbio(struct bio *bio, struct cache_set *c,
diff --git a/drivers/md/bcache/journal.c b/drivers/md/bcache/journal.c
index 1b736b860739..c94085f400a4 100644
--- a/drivers/md/bcache/journal.c
+++ b/drivers/md/bcache/journal.c
@@ -62,7 +62,7 @@ reread:		left = ca->sb.bucket_size - offset;
 		bio_set_op_attrs(bio, REQ_OP_READ, 0);
 		bch_bio_map(bio, data);
 
-		closure_bio_submit(bio, &cl);
+		closure_bio_submit(ca->set, bio, &cl);
 		closure_sync(&cl);
 
 		/* This function could be simpler now since we no longer write
@@ -674,7 +674,7 @@ static void journal_write_unlocked(struct closure *cl)
 	spin_unlock(&c->journal.lock);
 
 	while ((bio = bio_list_pop(&list)))
-		closure_bio_submit(bio, cl);
+		closure_bio_submit(c, bio, cl);
 
 	continue_at(cl, journal_write_done, NULL);
 }
diff --git a/drivers/md/bcache/request.c b/drivers/md/bcache/request.c
index 6422846b546e..7aca308bee5b 100644
--- a/drivers/md/bcache/request.c
+++ b/drivers/md/bcache/request.c
@@ -747,7 +747,7 @@ static void cached_dev_read_error(struct closure *cl)
 
 		/* XXX: invalidate cache */
 
-		closure_bio_submit(bio, cl);
+		closure_bio_submit(s->iop.c, bio, cl);
 	}
 
 	continue_at(cl, cached_dev_cache_miss_done, NULL);
@@ -872,7 +872,7 @@ static int cached_dev_cache_miss(struct btree *b, struct search *s,
 	s->cache_miss	= miss;
 	s->iop.bio	= cache_bio;
 	bio_get(cache_bio);
-	closure_bio_submit(cache_bio, &s->cl);
+	closure_bio_submit(s->iop.c, cache_bio, &s->cl);
 
 	return ret;
 out_put:
@@ -880,7 +880,7 @@ out_put:
 out_submit:
 	miss->bi_end_io		= request_endio;
 	miss->bi_private	= &s->cl;
-	closure_bio_submit(miss, &s->cl);
+	closure_bio_submit(s->iop.c, miss, &s->cl);
 	return ret;
 }
 
@@ -945,7 +945,7 @@ static void cached_dev_write(struct cached_dev *dc, struct search *s)
 
 		if ((bio_op(bio) != REQ_OP_DISCARD) ||
 		    blk_queue_discard(bdev_get_queue(dc->bdev)))
-			closure_bio_submit(bio, cl);
+			closure_bio_submit(s->iop.c, bio, cl);
 	} else if (s->iop.writeback) {
 		bch_writeback_add(dc);
 		s->iop.bio = bio;
@@ -960,12 +960,12 @@ static void cached_dev_write(struct cached_dev *dc, struct search *s)
 			flush->bi_private = cl;
 			flush->bi_opf = REQ_OP_WRITE | REQ_PREFLUSH;
 
-			closure_bio_submit(flush, cl);
+			closure_bio_submit(s->iop.c, flush, cl);
 		}
 	} else {
 		s->iop.bio = bio_clone_fast(bio, GFP_NOIO, dc->disk.bio_split);
 
-		closure_bio_submit(bio, cl);
+		closure_bio_submit(s->iop.c, bio, cl);
 	}
 
 	closure_call(&s->iop.cl, bch_data_insert, NULL, cl);
@@ -981,7 +981,7 @@ static void cached_dev_nodata(struct closure *cl)
 		bch_journal_meta(s->iop.c, cl);
 
 	/* If it's a flush, we send the flush to the backing device too */
-	closure_bio_submit(bio, cl);
+	closure_bio_submit(s->iop.c, bio, cl);
 
 	continue_at(cl, cached_dev_bio_complete, NULL);
 }
@@ -996,6 +996,12 @@ static blk_qc_t cached_dev_make_request(struct request_queue *q,
 	struct cached_dev *dc = container_of(d, struct cached_dev, disk);
 	int rw = bio_data_dir(bio);
 
+	if (unlikely(d->c && test_bit(CACHE_SET_IO_DISABLE, &d->c->flags))) {
+		bio->bi_status = BLK_STS_IOERR;
+		bio_endio(bio);
+		return BLK_QC_T_NONE;
+	}
+
 	atomic_set(&dc->backing_idle, 0);
 	generic_start_io_acct(q, rw, bio_sectors(bio), &d->disk->part0);
 
@@ -1112,6 +1118,12 @@ static blk_qc_t flash_dev_make_request(struct request_queue *q,
 	struct bcache_device *d = bio->bi_disk->private_data;
 	int rw = bio_data_dir(bio);
 
+	if (unlikely(d->c && test_bit(CACHE_SET_IO_DISABLE, &d->c->flags))) {
+		bio->bi_status = BLK_STS_IOERR;
+		bio_endio(bio);
+		return BLK_QC_T_NONE;
+	}
+
 	generic_start_io_acct(q, rw, bio_sectors(bio), &d->disk->part0);
 
 	s = search_alloc(bio, d);
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index e5be599338c5..dda4ccdd1360 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -521,7 +521,7 @@ static void prio_io(struct cache *ca, uint64_t bucket, int op,
 	bio_set_op_attrs(bio, op, REQ_SYNC|REQ_META|op_flags);
 	bch_bio_map(bio, ca->disk_buckets);
 
-	closure_bio_submit(bio, &ca->prio);
+	closure_bio_submit(ca->set, bio, &ca->prio);
 	closure_sync(cl);
 }
 
@@ -1350,6 +1350,9 @@ bool bch_cache_set_error(struct cache_set *c, const char *fmt, ...)
 	    test_bit(CACHE_SET_STOPPING, &c->flags))
 		return false;
 
+	if (test_and_set_bit(CACHE_SET_IO_DISABLE, &c->flags))
+		pr_warn("CACHE_SET_IO_DISABLE already set");
+
 	/* XXX: we can be called from atomic context
 	acquire_console_sem();
 	*/
@@ -1585,6 +1588,7 @@ struct cache_set *bch_cache_set_alloc(struct cache_sb *sb)
 	c->congested_read_threshold_us	= 2000;
 	c->congested_write_threshold_us	= 20000;
 	c->error_limit	= DEFAULT_IO_ERROR_LIMIT;
+	WARN_ON(test_and_clear_bit(CACHE_SET_IO_DISABLE, &c->flags));
 
 	return c;
 err:
diff --git a/drivers/md/bcache/sysfs.c b/drivers/md/bcache/sysfs.c
index 55673508628f..a3a45de5626d 100644
--- a/drivers/md/bcache/sysfs.c
+++ b/drivers/md/bcache/sysfs.c
@@ -95,6 +95,7 @@ read_attribute(partial_stripes_expensive);
 
 rw_attribute(synchronous);
 rw_attribute(journal_delay_ms);
+rw_attribute(io_disable);
 rw_attribute(discard);
 rw_attribute(running);
 rw_attribute(label);
@@ -591,6 +592,8 @@ SHOW(__bch_cache_set)
 	sysfs_printf(gc_always_rewrite,		"%i", c->gc_always_rewrite);
 	sysfs_printf(btree_shrinker_disabled,	"%i", c->shrinker_disabled);
 	sysfs_printf(copy_gc_enabled,		"%i", c->copy_gc_enabled);
+	sysfs_printf(io_disable,		"%i",
+		     test_bit(CACHE_SET_IO_DISABLE, &c->flags));
 
 	if (attr == &sysfs_bset_tree_stats)
 		return bch_bset_print_stats(c, buf);
@@ -680,6 +683,20 @@ STORE(__bch_cache_set)
 	if (attr == &sysfs_io_error_halflife)
 		c->error_decay = strtoul_or_return(buf) / 88;
 
+	if (attr == &sysfs_io_disable) {
+		int v = strtoul_or_return(buf);
+
+		if (v) {
+			if (test_and_set_bit(CACHE_SET_IO_DISABLE,
+					     &c->flags))
+				pr_warn("CACHE_SET_IO_DISABLE already set");
+		} else {
+			if (!test_and_clear_bit(CACHE_SET_IO_DISABLE,
+						&c->flags))
+				pr_warn("CACHE_SET_IO_DISABLE already cleared");
+		}
+	}
+
 	sysfs_strtoul(journal_delay_ms,		c->journal_delay_ms);
 	sysfs_strtoul(verify,			c->verify);
 	sysfs_strtoul(key_merging_disabled,	c->key_merging_disabled);
@@ -765,6 +782,7 @@ static struct attribute *bch_cache_set_internal_files[] = {
 	&sysfs_gc_always_rewrite,
 	&sysfs_btree_shrinker_disabled,
 	&sysfs_copy_gc_enabled,
+	&sysfs_io_disable,
 	NULL
 };
 KTYPE(bch_cache_set_internal);
diff --git a/drivers/md/bcache/util.h b/drivers/md/bcache/util.h
index a6763db7f061..268024529edd 100644
--- a/drivers/md/bcache/util.h
+++ b/drivers/md/bcache/util.h
@@ -567,12 +567,6 @@ static inline sector_t bdev_sectors(struct block_device *bdev)
 	return bdev->bd_inode->i_size >> 9;
 }
 
-#define closure_bio_submit(bio, cl)					\
-do {									\
-	closure_get(cl);						\
-	generic_make_request(bio);					\
-} while (0)
-
 uint64_t bch_crc64_update(uint64_t, const void *, size_t);
 uint64_t bch_crc64(const void *, size_t);
 
diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c
index 8f98ef1038d3..70092ada68e6 100644
--- a/drivers/md/bcache/writeback.c
+++ b/drivers/md/bcache/writeback.c
@@ -114,6 +114,7 @@ static void update_writeback_rate(struct work_struct *work)
 	struct cached_dev *dc = container_of(to_delayed_work(work),
 					     struct cached_dev,
 					     writeback_rate_update);
+	struct cache_set *c = dc->disk.c;
 
 	/*
 	 * should check BCACHE_DEV_RATE_DW_RUNNING before calling
@@ -123,7 +124,12 @@ static void update_writeback_rate(struct work_struct *work)
 	/* paired with where BCACHE_DEV_RATE_DW_RUNNING is tested */
 	smp_mb();
 
-	if (!test_bit(BCACHE_DEV_WB_RUNNING, &dc->disk.flags)) {
+	/*
+	 * CACHE_SET_IO_DISABLE might be set via sysfs interface,
+	 * check it here too.
+	 */
+	if (!test_bit(BCACHE_DEV_WB_RUNNING, &dc->disk.flags) ||
+	    test_bit(CACHE_SET_IO_DISABLE, &c->flags)) {
 		clear_bit(BCACHE_DEV_RATE_DW_RUNNING, &dc->disk.flags);
 		/* paired with where BCACHE_DEV_RATE_DW_RUNNING is tested */
 		smp_mb();
@@ -138,7 +144,12 @@ static void update_writeback_rate(struct work_struct *work)
 
 	up_read(&dc->writeback_lock);
 
-	if (test_bit(BCACHE_DEV_WB_RUNNING, &dc->disk.flags)) {
+	/*
+	 * CACHE_SET_IO_DISABLE might be set via sysfs interface,
+	 * check it here too.
+	 */
+	if (test_bit(BCACHE_DEV_WB_RUNNING, &dc->disk.flags) &&
+	    !test_bit(CACHE_SET_IO_DISABLE, &c->flags)) {
 		schedule_delayed_work(&dc->writeback_rate_update,
 			      dc->writeback_rate_update_seconds * HZ);
 	}
@@ -278,7 +289,7 @@ static void write_dirty(struct closure *cl)
 		bio_set_dev(&io->bio, io->dc->bdev);
 		io->bio.bi_end_io	= dirty_endio;
 
-		closure_bio_submit(&io->bio, cl);
+		closure_bio_submit(io->dc->disk.c, &io->bio, cl);
 	}
 
 	atomic_set(&dc->writeback_sequence_next, next_sequence);
@@ -304,7 +315,7 @@ static void read_dirty_submit(struct closure *cl)
 {
 	struct dirty_io *io = container_of(cl, struct dirty_io, cl);
 
-	closure_bio_submit(&io->bio, cl);
+	closure_bio_submit(io->dc->disk.c, &io->bio, cl);
 
 	continue_at(cl, write_dirty, io->dc->writeback_write_wq);
 }
@@ -330,7 +341,9 @@ static void read_dirty(struct cached_dev *dc)
 
 	next = bch_keybuf_next(&dc->writeback_keys);
 
-	while (!kthread_should_stop() && next) {
+	while (!kthread_should_stop() &&
+	       !test_bit(CACHE_SET_IO_DISABLE, &dc->disk.c->flags) &&
+	       next) {
 		size = 0;
 		nk = 0;
 
@@ -427,7 +440,9 @@ static void read_dirty(struct cached_dev *dc)
 			}
 		}
 
-		while (!kthread_should_stop() && delay) {
+		while (!kthread_should_stop() &&
+		       !test_bit(CACHE_SET_IO_DISABLE, &dc->disk.c->flags) &&
+		       delay) {
 			schedule_timeout_interruptible(delay);
 			delay = writeback_delay(dc, 0);
 		}
@@ -583,11 +598,13 @@ static bool refill_dirty(struct cached_dev *dc)
 static int bch_writeback_thread(void *arg)
 {
 	struct cached_dev *dc = arg;
+	struct cache_set *c = dc->disk.c;
 	bool searched_full_index;
 
 	bch_ratelimit_reset(&dc->writeback_rate);
 
-	while (!kthread_should_stop()) {
+	while (!kthread_should_stop() &&
+	       !test_bit(CACHE_SET_IO_DISABLE, &c->flags)) {
 		down_write(&dc->writeback_lock);
 		set_current_state(TASK_INTERRUPTIBLE);
 		/*
@@ -601,7 +618,8 @@ static int bch_writeback_thread(void *arg)
 		    (!atomic_read(&dc->has_dirty) || !dc->writeback_running)) {
 			up_write(&dc->writeback_lock);
 
-			if (kthread_should_stop()) {
+			if (kthread_should_stop() ||
+			    test_bit(CACHE_SET_IO_DISABLE, &c->flags)) {
 				set_current_state(TASK_RUNNING);
 				break;
 			}
@@ -637,6 +655,7 @@ static int bch_writeback_thread(void *arg)
 
 			while (delay &&
 			       !kthread_should_stop() &&
+			       !test_bit(CACHE_SET_IO_DISABLE, &c->flags) &&
 			       !test_bit(BCACHE_DEV_DETACHING, &dc->disk.flags))
 				delay = schedule_timeout_interruptible(delay);
 
@@ -644,8 +663,8 @@ static int bch_writeback_thread(void *arg)
 		}
 	}
 
-	dc->writeback_thread = NULL;
 	cached_dev_put(dc);
+	wait_for_kthread_stop();
 
 	return 0;
 }