ext4: improve cr 0 / cr 1 group scanning

Instead of traversing through groups linearly, scan groups in specific
orders at cr 0 and cr 1. At cr 0, we want to find groups that have the
largest free order >= the order of the request. So, with this patch,
we maintain lists for each possible order and insert each group into a
list based on the largest free order in its buddy bitmap. During cr 0
allocation, we traverse these lists in the increasing order of largest
free orders. This allows us to find a group with the best available cr
0 match in constant time. If nothing can be found, we fallback to cr 1
immediately.

At CR1, the story is slightly different. We want to traverse in the
order of increasing average fragment size. For CR1, we maintain a rb
tree of groupinfos which is sorted by average fragment size. Instead
of traversing linearly, at CR1, we traverse in the order of increasing
average fragment size, starting at the most optimal group. This brings
down cr 1 search complexity to log(num groups).

For cr >= 2, we just perform the linear search as before. Also, in
case of lock contention, we intermittently fallback to linear search
even in CR 0 and CR 1 cases. This allows us to proceed during the
allocation path even in case of high contention.

There is an opportunity to do optimization at CR2 too. That's because
at CR2 we only consider groups where bb_free counter (number of free
blocks) is greater than the request extent size. That's left as future
work.

All the changes introduced in this patch are protected under a new
mount option "mb_optimize_scan".

With this patchset, following experiment was performed:

Created a highly fragmented disk of size 65TB. The disk had no
contiguous 2M regions. Following command was run consecutively for 3
times:

time dd if=/dev/urandom of=file bs=2M count=10

Here are the results with and without cr 0/1 optimizations introduced
in this patch:

|---------+------------------------------+---------------------------|
|         | Without CR 0/1 Optimizations | With CR 0/1 Optimizations |
|---------+------------------------------+---------------------------|
| 1st run | 5m1.871s                     | 2m47.642s                 |
| 2nd run | 2m28.390s                    | 0m0.611s                  |
| 3rd run | 2m26.530s                    | 0m1.255s                  |
|---------+------------------------------+---------------------------|
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Link: https://lore.kernel.org/r/20210401172129.189766-6-harshadshirwadkar@gmail.comSigned-off-by: Theodore Ts'o <tytso@mit.edu>

fs/ext4/ext4.h

@@ -162,7 +162,12 @@ enum SHIFT_DIRECTION {
 #define EXT4_MB_USE_RESERVED		0x2000
 /* Do strict check for free blocks while retrying block allocation */
 #define EXT4_MB_STRICT_CHECK		0x4000
-
+/* Large fragment size list lookup succeeded at least once for cr = 0 */
+#define EXT4_MB_CR0_OPTIMIZED		0x8000
+/* Avg fragment size rb tree lookup succeeded at least once for cr = 1 */
+#define EXT4_MB_CR1_OPTIMIZED		0x00010000
+/* Perform linear traversal for one group */
+#define EXT4_MB_SEARCH_NEXT_LINEAR	0x00020000
 struct ext4_allocation_request {
 	/* target inode for block we're allocating */
 	struct inode *inode;
@@ -1247,7 +1252,9 @@ struct ext4_inode_info {
 #define EXT4_MOUNT2_JOURNAL_FAST_COMMIT	0x00000010 /* Journal fast commit */
 #define EXT4_MOUNT2_DAX_NEVER		0x00000020 /* Do not allow Direct Access */
 #define EXT4_MOUNT2_DAX_INODE		0x00000040 /* For printing options only */
-
+#define EXT4_MOUNT2_MB_OPTIMIZE_SCAN	0x00000080 /* Optimize group
+						    * scanning in mballoc
+						    */
 
 #define clear_opt(sb, opt)		EXT4_SB(sb)->s_mount_opt &= \
 						~EXT4_MOUNT_##opt
@@ -1528,9 +1535,14 @@ struct ext4_sb_info {
 	unsigned int s_mb_free_pending;
 	struct list_head s_freed_data_list;	/* List of blocks to be freed
 						   after commit completed */
+	struct rb_root s_mb_avg_fragment_size_root;
+	rwlock_t s_mb_rb_lock;
+	struct list_head *s_mb_largest_free_orders;
+	rwlock_t *s_mb_largest_free_orders_locks;
 
 	/* tunables */
 	unsigned long s_stripe;
+	unsigned int s_mb_max_linear_groups;
 	unsigned int s_mb_stream_request;
 	unsigned int s_mb_max_to_scan;
 	unsigned int s_mb_min_to_scan;
@@ -1554,6 +1566,8 @@ struct ext4_sb_info {
 	atomic_t s_bal_goals;	/* goal hits */
 	atomic_t s_bal_breaks;	/* too long searches */
 	atomic_t s_bal_2orders;	/* 2^order hits */
+	atomic_t s_bal_cr0_bad_suggestions;
+	atomic_t s_bal_cr1_bad_suggestions;
 	atomic64_t s_bal_cX_groups_considered[4];
 	atomic64_t s_bal_cX_hits[4];
 	atomic64_t s_bal_cX_failed[4];		/* cX loop didn't find blocks */
@@ -3360,11 +3374,14 @@ struct ext4_group_info {
 	ext4_grpblk_t	bb_free;	/* total free blocks */
 	ext4_grpblk_t	bb_fragments;	/* nr of freespace fragments */
 	ext4_grpblk_t	bb_largest_free_order;/* order of largest frag in BG */
+	ext4_group_t	bb_group;	/* Group number */
 	struct          list_head bb_prealloc_list;
 #ifdef DOUBLE_CHECK
 	void            *bb_bitmap;
 #endif
 	struct rw_semaphore alloc_sem;
+	struct rb_node	bb_avg_fragment_size_rb;
+	struct list_head bb_largest_free_order_node;
 	ext4_grpblk_t	bb_counters[];	/* Nr of free power-of-two-block
 					 * regions, index is order.
 					 * bb_counters[3] = 5 means

fs/ext4/mballoc.h

@@ -78,6 +78,18 @@
  */
 #define MB_DEFAULT_MAX_INODE_PREALLOC	512
 
+/*
+ * Number of groups to search linearly before performing group scanning
+ * optimization.
+ */
+#define MB_DEFAULT_LINEAR_LIMIT		4
+
+/*
+ * Minimum number of groups that should be present in the file system to perform
+ * group scanning optimizations.
+ */
+#define MB_DEFAULT_LINEAR_SCAN_THRESHOLD	16
+
 /*
  * Number of valid buddy orders
  */
@@ -166,11 +178,14 @@ struct ext4_allocation_context {
 	/* copy of the best found extent taken before preallocation efforts */
 	struct ext4_free_extent ac_f_ex;
 
+	ext4_group_t ac_last_optimal_group;
+	__u32 ac_groups_considered;
+	__u32 ac_flags;		/* allocation hints */
 	__u16 ac_groups_scanned;
+	__u16 ac_groups_linear_remaining;
 	__u16 ac_found;
 	__u16 ac_tail;
 	__u16 ac_buddy;
-	__u16 ac_flags;		/* allocation hints */
 	__u8 ac_status;
 	__u8 ac_criteria;
 	__u8 ac_2order;		/* if request is to allocate 2^N blocks and

fs/ext4/super.c

@@ -1688,7 +1688,7 @@ enum {
 	Opt_dioread_nolock, Opt_dioread_lock,
 	Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
 	Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache,
-	Opt_prefetch_block_bitmaps,
+	Opt_prefetch_block_bitmaps, Opt_mb_optimize_scan,
 #ifdef CONFIG_EXT4_DEBUG
 	Opt_fc_debug_max_replay, Opt_fc_debug_force
 #endif
@@ -1789,6 +1789,7 @@ static const match_table_t tokens = {
 	{Opt_nombcache, "nombcache"},
 	{Opt_nombcache, "no_mbcache"},	/* for backward compatibility */
 	{Opt_prefetch_block_bitmaps, "prefetch_block_bitmaps"},
+	{Opt_mb_optimize_scan, "mb_optimize_scan=%d"},
 	{Opt_removed, "check=none"},	/* mount option from ext2/3 */
 	{Opt_removed, "nocheck"},	/* mount option from ext2/3 */
 	{Opt_removed, "reservation"},	/* mount option from ext2/3 */
@@ -1821,6 +1822,8 @@ static ext4_fsblk_t get_sb_block(void **data)
 }
 
 #define DEFAULT_JOURNAL_IOPRIO (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 3))
+#define DEFAULT_MB_OPTIMIZE_SCAN	(-1)
+
 static const char deprecated_msg[] =
 	"Mount option \"%s\" will be removed by %s\n"
 	"Contact linux-ext4@vger.kernel.org if you think we should keep it.\n";
@@ -2009,6 +2012,7 @@ static const struct mount_opts {
 	{Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET},
 	{Opt_prefetch_block_bitmaps, EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS,
 	 MOPT_SET},
+	{Opt_mb_optimize_scan, EXT4_MOUNT2_MB_OPTIMIZE_SCAN, MOPT_GTE0},
 #ifdef CONFIG_EXT4_DEBUG
 	{Opt_fc_debug_force, EXT4_MOUNT2_JOURNAL_FAST_COMMIT,
 	 MOPT_SET | MOPT_2 | MOPT_EXT4_ONLY},
@@ -2093,6 +2097,7 @@ static int ext4_set_test_dummy_encryption(struct super_block *sb,
 struct ext4_parsed_options {
 	unsigned long journal_devnum;
 	unsigned int journal_ioprio;
+	int mb_optimize_scan;
 };
 
 static int handle_mount_opt(struct super_block *sb, char *opt, int token,
@@ -2389,6 +2394,13 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
 		sbi->s_mount_opt |= m->mount_opt;
 	} else if (token == Opt_data_err_ignore) {
 		sbi->s_mount_opt &= ~m->mount_opt;
+	} else if (token == Opt_mb_optimize_scan) {
+		if (arg != 0 && arg != 1) {
+			ext4_msg(sb, KERN_WARNING,
+				 "mb_optimize_scan should be set to 0 or 1.");
+			return -1;
+		}
+		parsed_opts->mb_optimize_scan = arg;
 	} else {
 		if (!args->from)
 			arg = 1;
@@ -4035,6 +4047,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	/* Set defaults for the variables that will be set during parsing */
 	parsed_opts.journal_ioprio = DEFAULT_JOURNAL_IOPRIO;
 	parsed_opts.journal_devnum = 0;
+	parsed_opts.mb_optimize_scan = DEFAULT_MB_OPTIMIZE_SCAN;
 
 	if ((data && !orig_data) || !sbi)
 		goto out_free_base;
@@ -4979,6 +4992,19 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
 	ext4_fc_replay_cleanup(sb);
 
 	ext4_ext_init(sb);
+
+	/*
+	 * Enable optimize_scan if number of groups is > threshold. This can be
+	 * turned off by passing "mb_optimize_scan=0". This can also be
+	 * turned on forcefully by passing "mb_optimize_scan=1".
+	 */
+	if (parsed_opts.mb_optimize_scan == 1)
+		set_opt2(sb, MB_OPTIMIZE_SCAN);
+	else if (parsed_opts.mb_optimize_scan == 0)
+		clear_opt2(sb, MB_OPTIMIZE_SCAN);
+	else if (sbi->s_groups_count >= MB_DEFAULT_LINEAR_SCAN_THRESHOLD)
+		set_opt2(sb, MB_OPTIMIZE_SCAN);
+
 	err = ext4_mb_init(sb);
 	if (err) {
 		ext4_msg(sb, KERN_ERR, "failed to initialize mballoc (%d)",

fs/ext4/sysfs.c

@@ -215,6 +215,7 @@ EXT4_RW_ATTR_SBI_UI(mb_order2_req, s_mb_order2_reqs);
 EXT4_RW_ATTR_SBI_UI(mb_stream_req, s_mb_stream_request);
 EXT4_RW_ATTR_SBI_UI(mb_group_prealloc, s_mb_group_prealloc);
 EXT4_RW_ATTR_SBI_UI(mb_max_inode_prealloc, s_mb_max_inode_prealloc);
+EXT4_RW_ATTR_SBI_UI(mb_max_linear_groups, s_mb_max_linear_groups);
 EXT4_RW_ATTR_SBI_UI(extent_max_zeroout_kb, s_extent_max_zeroout_kb);
 EXT4_ATTR(trigger_fs_error, 0200, trigger_test_error);
 EXT4_RW_ATTR_SBI_UI(err_ratelimit_interval_ms, s_err_ratelimit_state.interval);
@@ -263,6 +264,7 @@ static struct attribute *ext4_attrs[] = {
 	ATTR_LIST(mb_stream_req),
 	ATTR_LIST(mb_group_prealloc),
 	ATTR_LIST(mb_max_inode_prealloc),
+	ATTR_LIST(mb_max_linear_groups),
 	ATTR_LIST(max_writeback_mb_bump),
 	ATTR_LIST(extent_max_zeroout_kb),
 	ATTR_LIST(trigger_fs_error),