btrfs: add the beginning of async discard, discard workqueue

When discard is enabled, everytime a pinned extent is released back to
the block_group's free space cache, a discard is issued for the extent.
This is an overeager approach when it comes to discarding and helping
the SSD maintain enough free space to prevent severe garbage collection
situations.

This adds the beginning of async discard. Instead of issuing a discard
prior to returning it to the free space, it is just marked as untrimmed.
The block_group is then added to a LRU which then feeds into a workqueue
to issue discards at a much slower rate. Full discarding of unused block
groups is still done and will be addressed in a future patch of the
series.

For now, we don't persist the discard state of extents and bitmaps.
Therefore, our failure recovery mode will be to consider extents
untrimmed. This lets us handle failure and unmounting as one in the
same.

On a number of Facebook webservers, I collected data every minute
accounting the time we spent in btrfs_finish_extent_commit() (col. 1)
and in btrfs_commit_transaction() (col. 2). btrfs_finish_extent_commit()
is where we discard extents synchronously before returning them to the
free space cache.

discard=sync:
                 p99 total per minute       p99 total per minute
      Drive   |   extent_commit() (ms)  |    commit_trans() (ms)
    ---------------------------------------------------------------
     Drive A  |           434           |          1170
     Drive B  |           880           |          2330
     Drive C  |          2943           |          3920
     Drive D  |          4763           |          5701

discard=async:
                 p99 total per minute       p99 total per minute
      Drive   |   extent_commit() (ms)  |    commit_trans() (ms)
    --------------------------------------------------------------
     Drive A  |           134           |           956
     Drive B  |            64           |          1972
     Drive C  |            59           |          1032
     Drive D  |            62           |          1200

While it's not great that the stats are cumulative over 1m, all of these
servers are running the same workload and and the delta between the two
are substantial. We are spending significantly less time in
btrfs_finish_extent_commit() which is responsible for discarding.

Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Dennis Zhou <dennis@kernel.org>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>

1 files changed, 48 insertions, 6 deletions

diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 063f4db44024..fdc5401f3877 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -21,6 +21,7 @@
 #include "space-info.h"
 #include "delalloc-space.h"
 #include "block-group.h"
+#include "discard.h"
 
 #define BITS_PER_BITMAP		(PAGE_SIZE * 8UL)
 #define MAX_CACHE_BYTES_PER_GIG	SZ_32K
@@ -755,9 +756,11 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
 		/*
 		 * Sync discard ensures that the free space cache is always
 		 * trimmed.  So when reading this in, the state should reflect
-		 * that.
+		 * that.  We also do this for async as a stop gap for lack of
+		 * persistence.
 		 */
-		if (btrfs_test_opt(fs_info, DISCARD_SYNC))
+		if (btrfs_test_opt(fs_info, DISCARD_SYNC) ||
+		    btrfs_test_opt(fs_info, DISCARD_ASYNC))
 			e->trim_state = BTRFS_TRIM_STATE_TRIMMED;
 
 		if (!e->bytes) {
@@ -2382,6 +2385,7 @@ int __btrfs_add_free_space(struct btrfs_fs_info *fs_info,
 			   u64 offset, u64 bytes,
 			   enum btrfs_trim_state trim_state)
 {
+	struct btrfs_block_group *block_group = ctl->private;
 	struct btrfs_free_space *info;
 	int ret = 0;
 
@@ -2431,6 +2435,9 @@ out:
 		ASSERT(ret != -EEXIST);
 	}
 
+	if (trim_state != BTRFS_TRIM_STATE_TRIMMED)
+		btrfs_discard_queue_work(&fs_info->discard_ctl, block_group);
+
 	return ret;
 }
 
@@ -2447,6 +2454,25 @@ int btrfs_add_free_space(struct btrfs_block_group *block_group,
 				      bytenr, size, trim_state);
 }
 
+/*
+ * This is a subtle distinction because when adding free space back in general,
+ * we want it to be added as untrimmed for async. But in the case where we add
+ * it on loading of a block group, we want to consider it trimmed.
+ */
+int btrfs_add_free_space_async_trimmed(struct btrfs_block_group *block_group,
+				       u64 bytenr, u64 size)
+{
+	enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_UNTRIMMED;
+
+	if (btrfs_test_opt(block_group->fs_info, DISCARD_SYNC) ||
+	    btrfs_test_opt(block_group->fs_info, DISCARD_ASYNC))
+		trim_state = BTRFS_TRIM_STATE_TRIMMED;
+
+	return __btrfs_add_free_space(block_group->fs_info,
+				      block_group->free_space_ctl,
+				      bytenr, size, trim_state);
+}
+
 int btrfs_remove_free_space(struct btrfs_block_group *block_group,
 			    u64 offset, u64 bytes)
 {
@@ -3208,6 +3234,7 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster)
 static int do_trimming(struct btrfs_block_group *block_group,
 		       u64 *total_trimmed, u64 start, u64 bytes,
 		       u64 reserved_start, u64 reserved_bytes,
+		       enum btrfs_trim_state reserved_trim_state,
 		       struct btrfs_trim_range *trim_entry)
 {
 	struct btrfs_space_info *space_info = block_group->space_info;
@@ -3215,6 +3242,9 @@ static int do_trimming(struct btrfs_block_group *block_group,
 	struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl;
 	int ret;
 	int update = 0;
+	const u64 end = start + bytes;
+	const u64 reserved_end = reserved_start + reserved_bytes;
+	enum btrfs_trim_state trim_state = BTRFS_TRIM_STATE_UNTRIMMED;
 	u64 trimmed = 0;
 
 	spin_lock(&space_info->lock);
@@ -3228,11 +3258,20 @@ static int do_trimming(struct btrfs_block_group *block_group,
 	spin_unlock(&space_info->lock);
 
 	ret = btrfs_discard_extent(fs_info, start, bytes, &trimmed);
-	if (!ret)
+	if (!ret) {
 		*total_trimmed += trimmed;
+		trim_state = BTRFS_TRIM_STATE_TRIMMED;
+	}
 
 	mutex_lock(&ctl->cache_writeout_mutex);
-	btrfs_add_free_space(block_group, reserved_start, reserved_bytes);
+	if (reserved_start < start)
+		__btrfs_add_free_space(fs_info, ctl, reserved_start,
+				       start - reserved_start,
+				       reserved_trim_state);
+	if (start + bytes < reserved_start + reserved_bytes)
+		__btrfs_add_free_space(fs_info, ctl, end, reserved_end - end,
+				       reserved_trim_state);
+	__btrfs_add_free_space(fs_info, ctl, start, bytes, trim_state);
 	list_del(&trim_entry->list);
 	mutex_unlock(&ctl->cache_writeout_mutex);
 
@@ -3259,6 +3298,7 @@ static int trim_no_bitmap(struct btrfs_block_group *block_group,
 	int ret = 0;
 	u64 extent_start;
 	u64 extent_bytes;
+	enum btrfs_trim_state extent_trim_state;
 	u64 bytes;
 
 	while (start < end) {
@@ -3300,6 +3340,7 @@ static int trim_no_bitmap(struct btrfs_block_group *block_group,
 
 		extent_start = entry->offset;
 		extent_bytes = entry->bytes;
+		extent_trim_state = entry->trim_state;
 		start = max(start, extent_start);
 		bytes = min(extent_start + extent_bytes, end) - start;
 		if (bytes < minlen) {
@@ -3318,7 +3359,8 @@ static int trim_no_bitmap(struct btrfs_block_group *block_group,
 		mutex_unlock(&ctl->cache_writeout_mutex);
 
 		ret = do_trimming(block_group, total_trimmed, start, bytes,
-				  extent_start, extent_bytes, &trim_entry);
+				  extent_start, extent_bytes, extent_trim_state,
+				  &trim_entry);
 		if (ret)
 			break;
 next:
@@ -3442,7 +3484,7 @@ static int trim_bitmaps(struct btrfs_block_group *block_group,
 		mutex_unlock(&ctl->cache_writeout_mutex);
 
 		ret = do_trimming(block_group, total_trimmed, start, bytes,
-				  start, bytes, &trim_entry);
+				  start, bytes, 0, &trim_entry);
 		if (ret) {
 			reset_trimming_bitmap(ctl, offset);
 			break;