From 08239ca2a053dbc3b082916bdfbd88e5a9ad9267 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Thu, 28 Nov 2013 10:31:35 +0800 Subject: bcache: fix sparse non static symbol warning Fixes the following sparse warning: drivers/md/bcache/btree.c:2220:5: warning: symbol 'btree_insert_fn' was not declared. Should it be static? Signed-off-by: Wei Yongjun Signed-off-by: Kent Overstreet --- drivers/md/bcache/btree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index 5e2765aadce1..dc6e2be265b7 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -2217,7 +2217,7 @@ struct btree_insert_op { struct bkey *replace_key; }; -int btree_insert_fn(struct btree_op *b_op, struct btree *b) +static int btree_insert_fn(struct btree_op *b_op, struct btree *b) { struct btree_insert_op *op = container_of(b_op, struct btree_insert_op, op); -- cgit v1.2.3 From f665c0f852316ff99e9eb7f71f34d43003f8e139 Mon Sep 17 00:00:00 2001 From: Stefan Priebe Date: Sat, 16 Nov 2013 21:26:52 +0100 Subject: bcache: kthread don't set writeback task to INTERUPTIBLE at the beginning (schedule_timout_interuptible) and others do his on their own This prevents wrong load average calculation (load of 1 per thread) Signed-off-by: Kent Overstreet --- drivers/md/bcache/writeback.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index 99053b1251be..484e57d7012c 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -500,8 +500,6 @@ int bch_cached_dev_writeback_init(struct cached_dev *dc) if (IS_ERR(dc->writeback_thread)) return PTR_ERR(dc->writeback_thread); - set_task_state(dc->writeback_thread, TASK_INTERRUPTIBLE); - INIT_DELAYED_WORK(&dc->writeback_rate_update, update_writeback_rate); schedule_delayed_work(&dc->writeback_rate_update, dc->writeback_rate_update_seconds * HZ); -- cgit v1.2.3 From ce2b3f595e1c56639085645e0130426e443008c0 Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Thu, 28 Nov 2013 17:28:37 -0800 Subject: bcache: Use uninterruptible sleep in writeback We're just waiting on kthread_should_stop(), nothing else, so interruptible sleep was wrong here. Signed-off-by: Kent Overstreet --- drivers/md/bcache/writeback.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index 484e57d7012c..3cd931d3f26c 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -241,7 +241,7 @@ static void read_dirty(struct cached_dev *dc) if (KEY_START(&w->key) != dc->last_read || jiffies_to_msecs(delay) > 50) while (!kthread_should_stop() && delay) - delay = schedule_timeout_interruptible(delay); + delay = schedule_timeout_uninterruptible(delay); dc->last_read = KEY_OFFSET(&w->key); @@ -438,7 +438,7 @@ static int bch_writeback_thread(void *arg) while (delay && !kthread_should_stop() && !test_bit(BCACHE_DEV_DETACHING, &dc->disk.flags)) - delay = schedule_timeout_interruptible(delay); + delay = schedule_timeout_uninterruptible(delay); } } -- cgit v1.2.3 From d24a6e1087030b6da286df9433add5fa2f21b83b Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Sun, 10 Nov 2013 21:55:27 -0800 Subject: bcache: Fix dirty_data accounting Dirty data accounting wasn't quite right - firstly, we were adding the key we're inserting after it could have merged with another dirty key already in the btree, and secondly we could sometimes pass the wrong offset to bcache_dev_sectors_dirty_add() for dirty data we were overwriting - which is important when tracking dirty data by stripe. NOTE FOR BACKPORTERS: For 3.10 (and 3.11?) there's other accounting fixes necessary that got squashed in with other patches; the full patch against 3.10 is 408cc2f47eeac93a, available at: git://evilpiepirate.org/~kent/linux-bcache.git bcache-3.10-writeback-fixes Signed-off-by: Kent Overstreet Cc: linux-stable # >= v3.10 diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index 2a46036..4a12b2f 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -1817,7 +1817,8 @@ static bool fix_overlapping_extents(struct btree *b, struct bkey *insert, if (KEY_START(k) > KEY_START(insert) + sectors_found) goto check_failed; - if (KEY_PTRS(replace_key) != KEY_PTRS(k)) + if (KEY_PTRS(k) != KEY_PTRS(replace_key) || + KEY_DIRTY(k) != KEY_DIRTY(replace_key)) goto check_failed; /* skip past gen */ --- drivers/md/bcache/btree.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index dc6e2be265b7..c36745e5986a 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -1817,7 +1817,8 @@ static bool fix_overlapping_extents(struct btree *b, struct bkey *insert, if (KEY_START(k) > KEY_START(insert) + sectors_found) goto check_failed; - if (KEY_PTRS(replace_key) != KEY_PTRS(k)) + if (KEY_PTRS(k) != KEY_PTRS(replace_key) || + KEY_DIRTY(k) != KEY_DIRTY(replace_key)) goto check_failed; /* skip past gen */ -- cgit v1.2.3 From 9eb8ebeb2471b8cbaa078066aeb85fe5d46f5304 Mon Sep 17 00:00:00 2001 From: Nicholas Swenson Date: Tue, 22 Oct 2013 13:19:23 -0700 Subject: bcache: Fix for can_attach_cache() Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c index dec15cd2d797..c57bfa071a57 100644 --- a/drivers/md/bcache/super.c +++ b/drivers/md/bcache/super.c @@ -1676,7 +1676,7 @@ err: static bool can_attach_cache(struct cache *ca, struct cache_set *c) { return ca->sb.block_size == c->sb.block_size && - ca->sb.bucket_size == c->sb.block_size && + ca->sb.bucket_size == c->sb.bucket_size && ca->sb.nr_in_set == c->sb.nr_in_set; } -- cgit v1.2.3 From 97d11a660fd906dbea3dccd2638495d8497c3c81 Mon Sep 17 00:00:00 2001 From: Nicholas Swenson Date: Wed, 23 Oct 2013 17:35:26 -0700 Subject: bcache: Fix heap_peek() macro Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/util.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/util.h b/drivers/md/bcache/util.h index 362c4b3f8b4a..1030c6020e98 100644 --- a/drivers/md/bcache/util.h +++ b/drivers/md/bcache/util.h @@ -110,7 +110,7 @@ do { \ _r; \ }) -#define heap_peek(h) ((h)->size ? (h)->data[0] : NULL) +#define heap_peek(h) ((h)->used ? (h)->data[0] : NULL) #define heap_full(h) ((h)->used == (h)->size) -- cgit v1.2.3 From bee63f40cb5f5e8ab2abfbc85acde99cc0acd4b5 Mon Sep 17 00:00:00 2001 From: Nicholas Swenson Date: Thu, 31 Oct 2013 19:25:18 -0700 Subject: bcache: fix for gc crashing when no sectors are used Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/movinggc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/movinggc.c b/drivers/md/bcache/movinggc.c index 7c1275e66025..46c952379fab 100644 --- a/drivers/md/bcache/movinggc.c +++ b/drivers/md/bcache/movinggc.c @@ -184,7 +184,8 @@ static bool bucket_cmp(struct bucket *l, struct bucket *r) static unsigned bucket_heap_top(struct cache *ca) { - return GC_SECTORS_USED(heap_peek(&ca->heap)); + struct bucket *b; + return (b = heap_peek(&ca->heap)) ? GC_SECTORS_USED(b) : 0; } void bch_moving_gc(struct cache_set *c) -- cgit v1.2.3 From 981aa8c091e164ea51dd1e81b71a1f3852bbcceb Mon Sep 17 00:00:00 2001 From: Nicholas Swenson Date: Thu, 7 Nov 2013 17:53:19 -0800 Subject: bcache: bugfix - moving_gc now moves only correct buckets Removed gc_move_threshold because picking buckets only by threshold could lead moving extra buckets (ei. if there are buckets at the threshold that aren't supposed to be moved do to space considerations). This is replaced by a GC_MOVE bit in the gc_mark bitmask. Now only marked buckets get moved. Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/alloc.c | 2 ++ drivers/md/bcache/bcache.h | 6 +++--- drivers/md/bcache/movinggc.c | 8 +++----- 3 files changed, 8 insertions(+), 8 deletions(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/alloc.c b/drivers/md/bcache/alloc.c index 2b46bf1d7e40..4c9852d92b0a 100644 --- a/drivers/md/bcache/alloc.c +++ b/drivers/md/bcache/alloc.c @@ -421,9 +421,11 @@ out: if (watermark <= WATERMARK_METADATA) { SET_GC_MARK(b, GC_MARK_METADATA); + SET_GC_MOVE(b, 0); b->prio = BTREE_PRIO; } else { SET_GC_MARK(b, GC_MARK_RECLAIMABLE); + SET_GC_MOVE(b, 0); b->prio = INITIAL_PRIO; } diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h index 4beb55a0ff30..a7b1a7631ed2 100644 --- a/drivers/md/bcache/bcache.h +++ b/drivers/md/bcache/bcache.h @@ -197,7 +197,7 @@ struct bucket { uint8_t disk_gen; uint8_t last_gc; /* Most out of date gen in the btree */ uint8_t gc_gen; - uint16_t gc_mark; + uint16_t gc_mark; /* Bitfield used by GC. See below for field */ }; /* @@ -209,7 +209,8 @@ BITMASK(GC_MARK, struct bucket, gc_mark, 0, 2); #define GC_MARK_RECLAIMABLE 0 #define GC_MARK_DIRTY 1 #define GC_MARK_METADATA 2 -BITMASK(GC_SECTORS_USED, struct bucket, gc_mark, 2, 14); +BITMASK(GC_SECTORS_USED, struct bucket, gc_mark, 2, 13); +BITMASK(GC_MOVE, struct bucket, gc_mark, 15, 1); #include "journal.h" #include "stats.h" @@ -445,7 +446,6 @@ struct cache { * call prio_write() to keep gens from wrapping. */ uint8_t need_save_prio; - unsigned gc_move_threshold; /* * If nonzero, we know we aren't going to find any buckets to invalidate diff --git a/drivers/md/bcache/movinggc.c b/drivers/md/bcache/movinggc.c index 46c952379fab..30f347d4e609 100644 --- a/drivers/md/bcache/movinggc.c +++ b/drivers/md/bcache/movinggc.c @@ -25,10 +25,9 @@ static bool moving_pred(struct keybuf *buf, struct bkey *k) unsigned i; for (i = 0; i < KEY_PTRS(k); i++) { - struct cache *ca = PTR_CACHE(c, k, i); struct bucket *g = PTR_BUCKET(c, k, i); - if (GC_SECTORS_USED(g) < ca->gc_move_threshold) + if (GC_MOVE(g)) return true; } @@ -227,9 +226,8 @@ void bch_moving_gc(struct cache_set *c) sectors_to_move -= GC_SECTORS_USED(b); } - ca->gc_move_threshold = bucket_heap_top(ca); - - pr_debug("threshold %u", ca->gc_move_threshold); + while (heap_pop(&ca->heap, b, bucket_cmp)) + SET_GC_MOVE(b, 1); } mutex_unlock(&c->bucket_lock); -- cgit v1.2.3 From bf0a628a95dba7f983b6047cea695fb066fb2512 Mon Sep 17 00:00:00 2001 From: Nicholas Swenson Date: Tue, 26 Nov 2013 19:14:23 -0800 Subject: bcache: fix for gc and writeback race Garbage collector needs to check keys in the writeback keybuf to make sure it's not invalidating buckets to which the writeback keys point to. Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/btree.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c index c36745e5986a..31bb53fcc67a 100644 --- a/drivers/md/bcache/btree.c +++ b/drivers/md/bcache/btree.c @@ -1561,6 +1561,28 @@ size_t bch_btree_gc_finish(struct cache_set *c) SET_GC_MARK(PTR_BUCKET(c, &c->uuid_bucket, i), GC_MARK_METADATA); + /* don't reclaim buckets to which writeback keys point */ + rcu_read_lock(); + for (i = 0; i < c->nr_uuids; i++) { + struct bcache_device *d = c->devices[i]; + struct cached_dev *dc; + struct keybuf_key *w, *n; + unsigned j; + + if (!d || UUID_FLASH_ONLY(&c->uuids[i])) + continue; + dc = container_of(d, struct cached_dev, disk); + + spin_lock(&dc->writeback_keys.lock); + rbtree_postorder_for_each_entry_safe(w, n, + &dc->writeback_keys.keys, node) + for (j = 0; j < KEY_PTRS(&w->key); j++) + SET_GC_MARK(PTR_BUCKET(c, &w->key, j), + GC_MARK_DIRTY); + spin_unlock(&dc->writeback_keys.lock); + } + rcu_read_unlock(); + for_each_cache(ca, c, i) { uint64_t *i; -- cgit v1.2.3 From 6d3d1a9c542b19dff1c7d7c8354d0869e4655287 Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Mon, 16 Dec 2013 14:12:09 -0800 Subject: bcache: bugfix for race between moving_gc and bucket_invalidate There is a possibility for a bucket to be invalidated by the allocator while moving_gc was copying it's contents to another bucket, if the bucket only held cached data. To prevent this moving checks for a stale ptr (to an invalidated bucket), before and after reads. It it finds one, it simply ignores moving that data. This only affects bcache if the moving_gc was turned on, note that it's off by default. Signed-off-by: Nicholas Swenson Signed-off-by: Kent Overstreet --- drivers/md/bcache/movinggc.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/movinggc.c b/drivers/md/bcache/movinggc.c index 30f347d4e609..f2f0998c4a91 100644 --- a/drivers/md/bcache/movinggc.c +++ b/drivers/md/bcache/movinggc.c @@ -64,11 +64,16 @@ static void write_moving_finish(struct closure *cl) static void read_moving_endio(struct bio *bio, int error) { + struct bbio *b = container_of(bio, struct bbio, bio); struct moving_io *io = container_of(bio->bi_private, struct moving_io, cl); if (error) io->op.error = error; + else if (!KEY_DIRTY(&b->key) && + ptr_stale(io->op.c, &b->key, 0)) { + io->op.error = -EINTR; + } bch_bbio_endio(io->op.c, bio, error, "reading data to move"); } @@ -140,6 +145,11 @@ static void read_moving(struct cache_set *c) if (!w) break; + if (ptr_stale(c, &w->key, 0)) { + bch_keybuf_del(&c->moving_gc_keys, w); + continue; + } + io = kzalloc(sizeof(struct moving_io) + sizeof(struct bio_vec) * DIV_ROUND_UP(KEY_SIZE(&w->key), PAGE_SECTORS), GFP_KERNEL); -- cgit v1.2.3 From 16749c23c00c686ed168471963e3ddb0f3fcd855 Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Mon, 11 Nov 2013 13:58:34 -0800 Subject: bcache: New writeback PD controller The old writeback PD controller could get into states where it had throttled all the way down and take way too long to recover - it was too complicated to really understand what it was doing. This rewrites a good chunk of it to hopefully be simpler and make more sense, and it also pays more attention to units which should make the behaviour a bit easier to understand. Signed-off-by: Kent Overstreet --- drivers/md/bcache/bcache.h | 6 +++--- drivers/md/bcache/sysfs.c | 50 +++++++++++++++++++++++++------------------ drivers/md/bcache/util.c | 8 ++++++- drivers/md/bcache/writeback.c | 47 ++++++++++++++++++++-------------------- 4 files changed, 62 insertions(+), 49 deletions(-) (limited to 'drivers/md/bcache') diff --git a/drivers/md/bcache/bcache.h b/drivers/md/bcache/bcache.h index a7b1a7631ed2..754f43177483 100644 --- a/drivers/md/bcache/bcache.h +++ b/drivers/md/bcache/bcache.h @@ -373,14 +373,14 @@ struct cached_dev { unsigned char writeback_percent; unsigned writeback_delay; - int writeback_rate_change; - int64_t writeback_rate_derivative; uint64_t writeback_rate_target; + int64_t writeback_rate_proportional; + int64_t writeback_rate_derivative; + int64_t writeback_rate_change; unsigned writeback_rate_update_seconds; unsigned writeback_rate_d_term; unsigned writeback_rate_p_term_inverse; - unsigned writeback_rate_d_smooth; }; enum alloc_watermarks { diff --git a/drivers/md/bcache/sysfs.c b/drivers/md/bcache/sysfs.c index 80d4c2bee18a..a1f85612f0b3 100644 --- a/drivers/md/bcache/sysfs.c +++ b/drivers/md/bcache/sysfs.c @@ -83,7 +83,6 @@ rw_attribute(writeback_rate); rw_attribute(writeback_rate_update_seconds); rw_attribute(writeback_rate_d_term); rw_attribute(writeback_rate_p_term_inverse); -rw_attribute(writeback_rate_d_smooth); read_attribute(writeback_rate_debug); read_attribute(stripe_size); @@ -129,31 +128,41 @@ SHOW(__bch_cached_dev) var_printf(writeback_running, "%i"); var_print(writeback_delay); var_print(writeback_percent); - sysfs_print(writeback_rate, dc->writeback_rate.rate); + sysfs_hprint(writeback_rate, dc->writeback_rate.rate << 9); var_print(writeback_rate_update_seconds); var_print(writeback_rate_d_term); var_print(writeback_rate_p_term_inverse); - var_print(writeback_rate_d_smooth); if (attr == &sysfs_writeback_rate_debug) { + char rate[20]; char dirty[20]; - char derivative[20]; char target[20]; - bch_hprint(dirty, - bcache_dev_sectors_dirty(&dc->disk) << 9); - bch_hprint(derivative, dc->writeback_rate_derivative << 9); + char proportional[20]; + char derivative[20]; + char change[20]; + s64 next_io; + + bch_hprint(rate, dc->writeback_rate.rate << 9); + bch_hprint(dirty, bcache_dev_sectors_dirty(&dc->disk) << 9); bch_hprint(target, dc->writeback_rate_target << 9); + bch_hprint(proportional,dc->writeback_rate_proportional << 9); + bch_hprint(derivative, dc->writeback_rate_derivative << 9); + bch_hprint(change, dc->writeback_rate_change << 9); + + next_io = div64_s64(dc->writeback_rate.next - local_clock(), + NSEC_PER_MSEC); return sprintf(buf, - "rate:\t\t%u\n" - "change:\t\t%i\n" + "rate:\t\t%s/sec\n" "dirty:\t\t%s\n" + "target:\t\t%s\n" + "proportional:\t%s\n" "derivative:\t%s\n" - "target:\t\t%s\n", - dc->writeback_rate.rate, - dc->writeback_rate_change, - dirty, derivative, target); + "change:\t\t%s/sec\n" + "next io:\t%llims\n", + rate, dirty, target, proportional, + derivative, change, next_io); } sysfs_hprint(dirty_data, @@ -189,6 +198,7 @@ STORE(__cached_dev) struct kobj_uevent_env *env; #define d_strtoul(var) sysfs_strtoul(var, dc->var) +#define d_strtoul_nonzero(var) sysfs_strtoul_clamp(var, dc->var, 1, INT_MAX) #define d_strtoi_h(var) sysfs_hatoi(var, dc->var) sysfs_strtoul(data_csum, dc->disk.data_csum); @@ -197,16 +207,15 @@ STORE(__cached_dev) d_strtoul(writeback_metadata); d_strtoul(writeback_running); d_strtoul(writeback_delay); - sysfs_strtoul_clamp(writeback_rate, - dc->writeback_rate.rate, 1, 1000000); + sysfs_strtoul_clamp(writeback_percent, dc->writeback_percent, 0, 40); - d_strtoul(writeback_rate_update_seconds); + sysfs_strtoul_clamp(writeback_rate, + dc->writeback_rate.rate, 1, INT_MAX); + + d_strtoul_nonzero(writeback_rate_update_seconds); d_strtoul(writeback_rate_d_term); - d_strtoul(writeback_rate_p_term_inverse); - sysfs_strtoul_clamp(writeback_rate_p_term_inverse, - dc->writeback_rate_p_term_inverse, 1, INT_MAX); - d_strtoul(writeback_rate_d_smooth); + d_strtoul_nonzero(writeback_rate_p_term_inverse); d_strtoi_h(sequential_cutoff); d_strtoi_h(readahead); @@ -313,7 +322,6 @@ static struct attribute *bch_cached_dev_files[] = { &sysfs_writeback_rate_update_seconds, &sysfs_writeback_rate_d_term, &sysfs_writeback_rate_p_term_inverse, - &sysfs_writeback_rate_d_smooth, &sysfs_writeback_rate_debug, &sysfs_dirty_data, &sysfs_stripe_size, diff --git a/drivers/md/bcache/util.c b/drivers/md/bcache/util.c index 462214eeacbe..bb37618e7664 100644 --- a/drivers/md/bcache/util.c +++ b/drivers/md/bcache/util.c @@ -209,7 +209,13 @@ uint64_t bch_next_delay(struct bch_ratelimit *d, uint64_t done) { uint64_t now = local_clock(); - d->next += div_u64(done, d->rate); + d->next += div_u64(done * NSEC_PER_SEC, d->rate); + + if (time_before64(now + NSEC_PER_SEC, d->next)) + d->next = now + NSEC_PER_SEC; + + if (time_after64(now - NSEC_PER_SEC * 2, d->next)) + d->next = now - NSEC_PER_SEC * 2; return time_after64(d->next, now) ? div_u64(d->next - now, NSEC_PER_SEC / HZ) diff --git a/drivers/md/bcache/writeback.c b/drivers/md/bcache/writeback.c index 3cd931d3f26c..6c44fe059c27 100644 --- a/drivers/md/bcache/writeback.c +++ b/drivers/md/bcache/writeback.c @@ -30,38 +30,40 @@ static void __update_writeback_rate(struct cached_dev *dc) /* PD controller */ - int change = 0; - int64_t error; int64_t dirty = bcache_dev_sectors_dirty(&dc->disk); int64_t derivative = dirty - dc->disk.sectors_dirty_last; + int64_t proportional = dirty - target; + int64_t change; dc->disk.sectors_dirty_last = dirty; - derivative *= dc->writeback_rate_d_term; - derivative = clamp(derivative, -dirty, dirty); + /* Scale to sectors per second */ - derivative = ewma_add(dc->disk.sectors_dirty_derivative, derivative, - dc->writeback_rate_d_smooth, 0); + proportional *= dc->writeback_rate_update_seconds; + proportional = div_s64(proportional, dc->writeback_rate_p_term_inverse); - /* Avoid divide by zero */ - if (!target) - goto out; + derivative = div_s64(derivative, dc->writeback_rate_update_seconds); - error = div64_s64((dirty + derivative - target) << 8, target); + derivative = ewma_add(dc->disk.sectors_dirty_derivative, derivative, + (dc->writeback_rate_d_term / + dc->writeback_rate_update_seconds) ?: 1, 0); + + derivative *= dc->writeback_rate_d_term; + derivative = div_s64(derivative, dc->writeback_rate_p_term_inverse); - change = div_s64((dc->writeback_rate.rate * error) >> 8, - dc->writeback_rate_p_term_inverse); + change = proportional + derivative; /* Don't increase writeback rate if the device isn't keeping up */ if (change > 0 && time_after64(local_clock(), - dc->writeback_rate.next + 10 * NSEC_PER_MSEC)) + dc->writeback_rate.next + NSEC_PER_MSEC)) change = 0; dc->writeback_rate.rate = - clamp_t(int64_t, dc->writeback_rate.rate + change, + clamp_t(int64_t, (int64_t) dc->writeback_rate.rate + change, 1, NSEC_PER_MSEC); -out: + + dc->writeback_rate_proportional = proportional; dc->writeback_rate_derivative = derivative; dc->writeback_rate_change = change; dc->writeback_rate_target = target; @@ -87,15 +89,11 @@ static void update_writeback_rate(struct work_struct *work) static unsigned writeback_delay(struct cached_dev *dc, unsigned sectors) { - uint64_t ret; - if (test_bit(BCACHE_DEV_DETACHING, &dc->disk.flags) || !dc->writeback_percent) return 0; - ret = bch_next_delay(&dc->writeback_rate, sectors * 10000000ULL); - - return min_t(uint64_t, ret, HZ); + return bch_next_delay(&dc->writeback_rate, sectors); } struct dirty_io { @@ -476,6 +474,8 @@ void bch_sectors_dirty_init(struct cached_dev *dc) bch_btree_map_keys(&op.op, dc->disk.c, &KEY(op.inode, 0, 0), sectors_dirty_init_fn, 0); + + dc->disk.sectors_dirty_last = bcache_dev_sectors_dirty(&dc->disk); } int bch_cached_dev_writeback_init(struct cached_dev *dc) @@ -490,10 +490,9 @@ int bch_cached_dev_writeback_init(struct cached_dev *dc) dc->writeback_delay = 30; dc->writeback_rate.rate = 1024; - dc->writeback_rate_update_seconds = 30; - dc->writeback_rate_d_term = 16; - dc->writeback_rate_p_term_inverse = 64; - dc->writeback_rate_d_smooth = 8; + dc->writeback_rate_update_seconds = 5; + dc->writeback_rate_d_term = 30; + dc->writeback_rate_p_term_inverse = 6000; dc->writeback_thread = kthread_create(bch_writeback_thread, dc, "bcache_writeback"); -- cgit v1.2.3