From ab74c7b23f3770935016e3eb3ecdf1e42b73efaa Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 15 Jul 2020 11:48:55 -0400 Subject: ext4: indicate via a block bitmap read is prefetched via a tracepoint Modify the ext4_read_block_bitmap_load tracepoint so that it tells us whether a block bitmap is being prefetched. Signed-off-by: Theodore Ts'o Reviewed-by: Artem Blagodarenko --- include/trace/events/ext4.h | 24 ++++++++++++++++++++---- 1 file changed, 20 insertions(+), 4 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index cc41d692ae8e..cbcd2e1a608d 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -1312,18 +1312,34 @@ DEFINE_EVENT(ext4__bitmap_load, ext4_mb_buddy_bitmap_load, TP_ARGS(sb, group) ); -DEFINE_EVENT(ext4__bitmap_load, ext4_read_block_bitmap_load, +DEFINE_EVENT(ext4__bitmap_load, ext4_load_inode_bitmap, TP_PROTO(struct super_block *sb, unsigned long group), TP_ARGS(sb, group) ); -DEFINE_EVENT(ext4__bitmap_load, ext4_load_inode_bitmap, +TRACE_EVENT(ext4_read_block_bitmap_load, + TP_PROTO(struct super_block *sb, unsigned long group, bool prefetch), - TP_PROTO(struct super_block *sb, unsigned long group), + TP_ARGS(sb, group, prefetch), - TP_ARGS(sb, group) + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + __field( bool, prefetch ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + __entry->prefetch = prefetch; + ), + + TP_printk("dev %d,%d group %u prefetch %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->group, __entry->prefetch) ); TRACE_EVENT(ext4_direct_IO_enter, -- cgit v1.2.3 From 3d392b2676bf3199863a1e5efb2c087ad9d442a4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 17 Jul 2020 00:14:40 -0400 Subject: ext4: add prefetch_block_bitmaps mount option For file systems where we can afford to keep the buddy bitmaps cached, we can speed up initial writes to large file systems by starting to load the block allocation bitmaps as soon as the file system is mounted. This won't work well for _super_ large file systems, or memory constrained systems, so we only enable this when it is requested via a mount option. Addresses-Google-Bug: 159488342 Signed-off-by: Theodore Ts'o Reviewed-by: Andreas Dilger --- fs/ext4/ext4.h | 15 +++++++++++- fs/ext4/mballoc.c | 10 ++++---- fs/ext4/super.c | 57 ++++++++++++++++++++++++++++++++------------- include/trace/events/ext4.h | 44 ++++++++++++++++++++++++++++++++++ 4 files changed, 103 insertions(+), 23 deletions(-) (limited to 'include/trace') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 26ae31a994a2..e5f0c66a6156 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1188,6 +1188,7 @@ struct ext4_inode_info { #define EXT4_MOUNT_JOURNAL_CHECKSUM 0x800000 /* Journal checksums */ #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT 0x1000000 /* Journal Async Commit */ #define EXT4_MOUNT_WARN_ON_ERROR 0x2000000 /* Trigger WARN_ON on error */ +#define EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS 0x4000000 #define EXT4_MOUNT_DELALLOC 0x8000000 /* Delalloc support */ #define EXT4_MOUNT_DATA_ERR_ABORT 0x10000000 /* Abort on file data write */ #define EXT4_MOUNT_BLOCK_VALIDITY 0x20000000 /* Block validity checking */ @@ -2334,9 +2335,15 @@ struct ext4_lazy_init { struct mutex li_list_mtx; }; +enum ext4_li_mode { + EXT4_LI_MODE_PREFETCH_BBITMAP, + EXT4_LI_MODE_ITABLE, +}; + struct ext4_li_request { struct super_block *lr_super; - struct ext4_sb_info *lr_sbi; + enum ext4_li_mode lr_mode; + ext4_group_t lr_first_not_zeroed; ext4_group_t lr_next_group; struct list_head lr_request; unsigned long lr_next_sched; @@ -2676,6 +2683,12 @@ extern int ext4_mb_reserve_blocks(struct super_block *, int); extern void ext4_discard_preallocations(struct inode *); extern int __init ext4_init_mballoc(void); extern void ext4_exit_mballoc(void); +extern ext4_group_t ext4_mb_prefetch(struct super_block *sb, + ext4_group_t group, + unsigned int nr, int *cnt); +extern void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr); + extern void ext4_free_blocks(handle_t *handle, struct inode *inode, struct buffer_head *bh, ext4_fsblk_t block, unsigned long count, int flags); diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 9a07da53ab7b..8ecd49dd3906 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2232,9 +2232,8 @@ out: * Start prefetching @nr block bitmaps starting at @group. * Return the next group which needs to be prefetched. */ -static ext4_group_t -ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, - unsigned int nr, int *cnt) +ext4_group_t ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, + unsigned int nr, int *cnt) { ext4_group_t ngroups = ext4_get_groups_count(sb); struct buffer_head *bh; @@ -2284,9 +2283,8 @@ ext4_mb_prefetch(struct super_block *sb, ext4_group_t group, * waiting for the block allocation bitmap read to finish when * ext4_mb_prefetch_fini is called from ext4_mb_regular_allocator(). */ -static void -ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, - unsigned int nr) +void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group, + unsigned int nr) { while (nr-- > 0) { struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group, diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 8c00f0a09f4d..1cbe3f248faf 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1521,6 +1521,7 @@ 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, }; static const match_table_t tokens = { @@ -1612,6 +1613,7 @@ static const match_table_t tokens = { {Opt_test_dummy_encryption, "test_dummy_encryption"}, {Opt_nombcache, "nombcache"}, {Opt_nombcache, "no_mbcache"}, /* for backward compatibility */ + {Opt_prefetch_block_bitmaps, "prefetch_block_bitmaps"}, {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 */ @@ -1829,6 +1831,8 @@ static const struct mount_opts { {Opt_max_dir_size_kb, 0, MOPT_GTE0}, {Opt_test_dummy_encryption, 0, MOPT_STRING}, {Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET}, + {Opt_prefetch_block_bitmaps, EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS, + MOPT_SET}, {Opt_err, 0, 0} }; @@ -3201,15 +3205,34 @@ static void print_daily_error_info(struct timer_list *t) static int ext4_run_li_request(struct ext4_li_request *elr) { struct ext4_group_desc *gdp = NULL; - ext4_group_t group, ngroups; - struct super_block *sb; + struct super_block *sb = elr->lr_super; + ext4_group_t ngroups = EXT4_SB(sb)->s_groups_count; + ext4_group_t group = elr->lr_next_group; unsigned long timeout = 0; + unsigned int prefetch_ios = 0; int ret = 0; - sb = elr->lr_super; - ngroups = EXT4_SB(sb)->s_groups_count; + if (elr->lr_mode == EXT4_LI_MODE_PREFETCH_BBITMAP) { + elr->lr_next_group = ext4_mb_prefetch(sb, group, + EXT4_SB(sb)->s_mb_prefetch, &prefetch_ios); + if (prefetch_ios) + ext4_mb_prefetch_fini(sb, elr->lr_next_group, + prefetch_ios); + trace_ext4_prefetch_bitmaps(sb, group, elr->lr_next_group, + prefetch_ios); + if (group >= elr->lr_next_group) { + ret = 1; + if (elr->lr_first_not_zeroed != ngroups && + !sb_rdonly(sb) && test_opt(sb, INIT_INODE_TABLE)) { + elr->lr_next_group = elr->lr_first_not_zeroed; + elr->lr_mode = EXT4_LI_MODE_ITABLE; + ret = 0; + } + } + return ret; + } - for (group = elr->lr_next_group; group < ngroups; group++) { + for (; group < ngroups; group++) { gdp = ext4_get_group_desc(sb, group, NULL); if (!gdp) { ret = 1; @@ -3227,9 +3250,10 @@ static int ext4_run_li_request(struct ext4_li_request *elr) timeout = jiffies; ret = ext4_init_inode_table(sb, group, elr->lr_timeout ? 0 : 1); + trace_ext4_lazy_itable_init(sb, group); if (elr->lr_timeout == 0) { timeout = (jiffies - timeout) * - elr->lr_sbi->s_li_wait_mult; + EXT4_SB(elr->lr_super)->s_li_wait_mult; elr->lr_timeout = timeout; } elr->lr_next_sched = jiffies + elr->lr_timeout; @@ -3244,15 +3268,11 @@ static int ext4_run_li_request(struct ext4_li_request *elr) */ static void ext4_remove_li_request(struct ext4_li_request *elr) { - struct ext4_sb_info *sbi; - if (!elr) return; - sbi = elr->lr_sbi; - list_del(&elr->lr_request); - sbi->s_li_request = NULL; + EXT4_SB(elr->lr_super)->s_li_request = NULL; kfree(elr); } @@ -3461,7 +3481,6 @@ static int ext4_li_info_new(void) static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, ext4_group_t start) { - struct ext4_sb_info *sbi = EXT4_SB(sb); struct ext4_li_request *elr; elr = kzalloc(sizeof(*elr), GFP_KERNEL); @@ -3469,8 +3488,13 @@ static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, return NULL; elr->lr_super = sb; - elr->lr_sbi = sbi; - elr->lr_next_group = start; + elr->lr_first_not_zeroed = start; + if (test_opt(sb, PREFETCH_BLOCK_BITMAPS)) + elr->lr_mode = EXT4_LI_MODE_PREFETCH_BBITMAP; + else { + elr->lr_mode = EXT4_LI_MODE_ITABLE; + elr->lr_next_group = start; + } /* * Randomize first schedule time of the request to @@ -3500,8 +3524,9 @@ int ext4_register_li_request(struct super_block *sb, goto out; } - if (first_not_zeroed == ngroups || sb_rdonly(sb) || - !test_opt(sb, INIT_INODE_TABLE)) + if (!test_opt(sb, PREFETCH_BLOCK_BITMAPS) && + (first_not_zeroed == ngroups || sb_rdonly(sb) || + !test_opt(sb, INIT_INODE_TABLE))) goto out; elr = ext4_li_request_new(sb, first_not_zeroed); diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index cbcd2e1a608d..8008d2e116b9 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -2742,6 +2742,50 @@ TRACE_EVENT(ext4_error, __entry->function, __entry->line) ); +TRACE_EVENT(ext4_prefetch_bitmaps, + TP_PROTO(struct super_block *sb, ext4_group_t group, + ext4_group_t next, unsigned int prefetch_ios), + + TP_ARGS(sb, group, next, prefetch_ios), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + __field( __u32, next ) + __field( __u32, ios ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + __entry->next = next; + __entry->ios = prefetch_ios; + ), + + TP_printk("dev %d,%d group %u next %u ios %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->group, __entry->next, __entry->ios) +); + +TRACE_EVENT(ext4_lazy_itable_init, + TP_PROTO(struct super_block *sb, ext4_group_t group), + + TP_ARGS(sb, group), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u32, group ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->group = group; + ), + + TP_printk("dev %d,%d group %u", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->group) +); + #endif /* _TRACE_EXT4_H */ /* This part must be outside protection */ -- cgit v1.2.3 From 27bc446e2def38db3244a6eb4bb1d6312936610a Mon Sep 17 00:00:00 2001 From: brookxu Date: Mon, 17 Aug 2020 15:36:15 +0800 Subject: ext4: limit the length of per-inode prealloc list In the scenario of writing sparse files, the per-inode prealloc list may be very long, resulting in high overhead for ext4_mb_use_preallocated(). To circumvent this problem, we limit the maximum length of per-inode prealloc list to 512 and allow users to modify it. After patching, we observed that the sys ratio of cpu has dropped, and the system throughput has increased significantly. We created a process to write the sparse file, and the running time of the process on the fixed kernel was significantly reduced, as follows: Running time on unfixed kernel： [root@TENCENT64 ~]# time taskset 0x01 ./sparse /data1/sparce.dat real 0m2.051s user 0m0.008s sys 0m2.026s Running time on fixed kernel： [root@TENCENT64 ~]# time taskset 0x01 ./sparse /data1/sparce.dat real 0m0.471s user 0m0.004s sys 0m0.395s Signed-off-by: Chunguang Xu Link: https://lore.kernel.org/r/d7a98178-056b-6db5-6bce-4ead23f4a257@gmail.com Signed-off-by: Theodore Ts'o --- Documentation/admin-guide/ext4.rst | 3 ++ fs/ext4/ext4.h | 4 ++- fs/ext4/extents.c | 10 +++--- fs/ext4/file.c | 2 +- fs/ext4/indirect.c | 2 +- fs/ext4/inode.c | 6 ++-- fs/ext4/ioctl.c | 2 +- fs/ext4/mballoc.c | 74 +++++++++++++++++++++++++++++++++----- fs/ext4/mballoc.h | 4 +++ fs/ext4/move_extent.c | 4 +-- fs/ext4/super.c | 3 +- fs/ext4/sysfs.c | 2 ++ include/trace/events/ext4.h | 17 +++++---- 13 files changed, 104 insertions(+), 29 deletions(-) (limited to 'include/trace') diff --git a/Documentation/admin-guide/ext4.rst b/Documentation/admin-guide/ext4.rst index 7fc6a72920c9..59bcc4a92602 100644 --- a/Documentation/admin-guide/ext4.rst +++ b/Documentation/admin-guide/ext4.rst @@ -482,6 +482,9 @@ Files in /sys/fs/ext4/: multiple of this tuning parameter if the stripe size is not set in the ext4 superblock + mb_max_inode_prealloc + The maximum length of per-inode ext4_prealloc_space list. + mb_max_to_scan The maximum number of extents the multiblock allocator will search to find the best extent. diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 71b4370a3f91..523e00d7b392 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1070,6 +1070,7 @@ struct ext4_inode_info { struct timespec64 i_crtime; /* mballoc */ + atomic_t i_prealloc_active; struct list_head i_prealloc_list; spinlock_t i_prealloc_lock; @@ -1518,6 +1519,7 @@ struct ext4_sb_info { unsigned int s_mb_stats; unsigned int s_mb_order2_reqs; unsigned int s_mb_group_prealloc; + unsigned int s_mb_max_inode_prealloc; unsigned int s_max_dir_size_kb; /* where last allocation was done - for stream allocation */ unsigned long s_mb_last_group; @@ -2682,7 +2684,7 @@ extern int ext4_mb_release(struct super_block *); extern ext4_fsblk_t ext4_mb_new_blocks(handle_t *, struct ext4_allocation_request *, int *); extern int ext4_mb_reserve_blocks(struct super_block *, int); -extern void ext4_discard_preallocations(struct inode *); +extern void ext4_discard_preallocations(struct inode *, unsigned int); extern int __init ext4_init_mballoc(void); extern void ext4_exit_mballoc(void); extern ext4_group_t ext4_mb_prefetch(struct super_block *sb, diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 0eea09aa0f26..a0481582187a 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -100,7 +100,7 @@ static int ext4_ext_trunc_restart_fn(struct inode *inode, int *dropped) * i_mutex. So we can safely drop the i_data_sem here. */ BUG_ON(EXT4_JOURNAL(inode) == NULL); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); *dropped = 1; return 0; @@ -4266,7 +4266,7 @@ got_allocated_blocks: * not a good idea to call discard here directly, * but otherwise we'd need to call it every free(). */ - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) fb_flags = EXT4_FREE_BLOCKS_NO_QUOT_UPDATE; ext4_free_blocks(handle, inode, NULL, newblock, @@ -5293,7 +5293,7 @@ static int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) } down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_es_remove_extent(inode, punch_start, EXT_MAX_BLOCKS - punch_start); @@ -5307,7 +5307,7 @@ static int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) up_write(&EXT4_I(inode)->i_data_sem); goto out_stop; } - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_ext_shift_extents(inode, handle, punch_stop, punch_stop - punch_start, SHIFT_LEFT); @@ -5439,7 +5439,7 @@ static int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len) goto out_stop; down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); path = ext4_find_extent(inode, offset_lblk, NULL, 0); if (IS_ERR(path)) { diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 7a2720517bbb..e608ce3fb535 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -147,7 +147,7 @@ static int ext4_release_file(struct inode *inode, struct file *filp) (atomic_read(&inode->i_writecount) == 1) && !EXT4_I(inode)->i_reserved_data_blocks) { down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); } if (is_dx(inode) && filp->private_data) diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 433ca8415c5a..80c9f33800be 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -696,7 +696,7 @@ static int ext4_ind_trunc_restart_fn(handle_t *handle, struct inode *inode, * i_mutex. So we can safely drop the i_data_sem here. */ BUG_ON(EXT4_JOURNAL(inode) == NULL); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); up_write(&EXT4_I(inode)->i_data_sem); *dropped = 1; return 0; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 0b07576af3bf..77543f988258 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -383,7 +383,7 @@ void ext4_da_update_reserve_space(struct inode *inode, */ if ((ei->i_reserved_data_blocks == 0) && !inode_is_open_for_write(inode)) - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); } static int __check_block_validity(struct inode *inode, const char *func, @@ -4055,7 +4055,7 @@ int ext4_punch_hole(struct inode *inode, loff_t offset, loff_t length) if (stop_block > first_block) { down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ret = ext4_es_remove_extent(inode, first_block, stop_block - first_block); @@ -4210,7 +4210,7 @@ int ext4_truncate(struct inode *inode) down_write(&EXT4_I(inode)->i_data_sem); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) err = ext4_ext_truncate(handle, inode); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 6e70a63dcca7..36eca3bc036a 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -202,7 +202,7 @@ static long swap_inode_boot_loader(struct super_block *sb, reset_inode_seed(inode); reset_inode_seed(inode_bl); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); err = ext4_mark_inode_dirty(handle, inode); if (err < 0) { diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 45ac6088b4ac..132c118d12e1 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2878,6 +2878,7 @@ int ext4_mb_init(struct super_block *sb) sbi->s_mb_stats = MB_DEFAULT_STATS; sbi->s_mb_stream_request = MB_DEFAULT_STREAM_THRESHOLD; sbi->s_mb_order2_reqs = MB_DEFAULT_ORDER2_REQS; + sbi->s_mb_max_inode_prealloc = MB_DEFAULT_MAX_INODE_PREALLOC; /* * The default group preallocation is 512, which for 4k block * sizes translates to 2 megabytes. However for bigalloc file @@ -3816,6 +3817,26 @@ void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, mb_debug(sb, "preallocated %d for group %u\n", preallocated, group); } +static void ext4_mb_mark_pa_deleted(struct super_block *sb, + struct ext4_prealloc_space *pa) +{ + struct ext4_inode_info *ei; + + if (pa->pa_deleted) { + ext4_warning(sb, "deleted pa, type:%d, pblk:%llu, lblk:%u, len:%d\n", + pa->pa_type, pa->pa_pstart, pa->pa_lstart, + pa->pa_len); + return; + } + + pa->pa_deleted = 1; + + if (pa->pa_type == MB_INODE_PA) { + ei = EXT4_I(pa->pa_inode); + atomic_dec(&ei->i_prealloc_active); + } +} + static void ext4_mb_pa_callback(struct rcu_head *head) { struct ext4_prealloc_space *pa; @@ -3848,7 +3869,7 @@ static void ext4_mb_put_pa(struct ext4_allocation_context *ac, return; } - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); grp_blk = pa->pa_pstart; @@ -3972,6 +3993,7 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) spin_lock(pa->pa_obj_lock); list_add_rcu(&pa->pa_inode_list, &ei->i_prealloc_list); spin_unlock(pa->pa_obj_lock); + atomic_inc(&ei->i_prealloc_active); } /* @@ -4182,7 +4204,7 @@ repeat: } /* seems this one can be freed ... */ - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); /* we can trust pa_free ... */ free += pa->pa_free; @@ -4245,7 +4267,7 @@ out_dbg: * * FIXME!! Make sure it is valid at all the call sites */ -void ext4_discard_preallocations(struct inode *inode) +void ext4_discard_preallocations(struct inode *inode, unsigned int needed) { struct ext4_inode_info *ei = EXT4_I(inode); struct super_block *sb = inode->i_sb; @@ -4263,15 +4285,19 @@ void ext4_discard_preallocations(struct inode *inode) mb_debug(sb, "discard preallocation for inode %lu\n", inode->i_ino); - trace_ext4_discard_preallocations(inode); + trace_ext4_discard_preallocations(inode, + atomic_read(&ei->i_prealloc_active), needed); INIT_LIST_HEAD(&list); + if (needed == 0) + needed = UINT_MAX; + repeat: /* first, collect all pa's in the inode */ spin_lock(&ei->i_prealloc_lock); - while (!list_empty(&ei->i_prealloc_list)) { - pa = list_entry(ei->i_prealloc_list.next, + while (!list_empty(&ei->i_prealloc_list) && needed) { + pa = list_entry(ei->i_prealloc_list.prev, struct ext4_prealloc_space, pa_inode_list); BUG_ON(pa->pa_obj_lock != &ei->i_prealloc_lock); spin_lock(&pa->pa_lock); @@ -4288,10 +4314,11 @@ repeat: } if (pa->pa_deleted == 0) { - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); list_del_rcu(&pa->pa_inode_list); list_add(&pa->u.pa_tmp_list, &list); + needed--; continue; } @@ -4592,7 +4619,7 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb, BUG_ON(pa->pa_type != MB_GROUP_PA); /* seems this one can be freed ... */ - pa->pa_deleted = 1; + ext4_mb_mark_pa_deleted(sb, pa); spin_unlock(&pa->pa_lock); list_del_rcu(&pa->pa_inode_list); @@ -4690,11 +4717,30 @@ static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) return ; } +/* + * if per-inode prealloc list is too long, trim some PA + */ +static void ext4_mb_trim_inode_pa(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + int count, delta; + + count = atomic_read(&ei->i_prealloc_active); + delta = (sbi->s_mb_max_inode_prealloc >> 2) + 1; + if (count > sbi->s_mb_max_inode_prealloc + delta) { + count -= sbi->s_mb_max_inode_prealloc; + ext4_discard_preallocations(inode, count); + } +} + /* * release all resource we used in allocation */ static int ext4_mb_release_context(struct ext4_allocation_context *ac) { + struct inode *inode = ac->ac_inode; + struct ext4_inode_info *ei = EXT4_I(inode); struct ext4_sb_info *sbi = EXT4_SB(ac->ac_sb); struct ext4_prealloc_space *pa = ac->ac_pa; if (pa) { @@ -4720,6 +4766,17 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) ext4_mb_add_n_trim(ac); } } + + if (pa->pa_type == MB_INODE_PA) { + /* + * treat per-inode prealloc list as a lru list, then try + * to trim the least recently used PA. + */ + spin_lock(pa->pa_obj_lock); + list_move(&pa->pa_inode_list, &ei->i_prealloc_list); + spin_unlock(pa->pa_obj_lock); + } + ext4_mb_put_pa(ac, ac->ac_sb, pa); } if (ac->ac_bitmap_page) @@ -4729,6 +4786,7 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) if (ac->ac_flags & EXT4_MB_HINT_GROUP_ALLOC) mutex_unlock(&ac->ac_lg->lg_mutex); ext4_mb_collect_stats(ac); + ext4_mb_trim_inode_pa(inode); return 0; } diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 6b4d17c2935d..e75b4749aa1c 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -73,6 +73,10 @@ */ #define MB_DEFAULT_GROUP_PREALLOC 512 +/* + * maximum length of inode prealloc list + */ +#define MB_DEFAULT_MAX_INODE_PREALLOC 512 struct ext4_free_data { /* this links the free block information from sb_info */ diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 1ed86fb6c302..0d601b822875 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -686,8 +686,8 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_blk, out: if (*moved_len) { - ext4_discard_preallocations(orig_inode); - ext4_discard_preallocations(donor_inode); + ext4_discard_preallocations(orig_inode, 0); + ext4_discard_preallocations(donor_inode, 0); } ext4_ext_drop_refs(path); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index daa94c7f7271..13bdddc081e0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1127,6 +1127,7 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) inode_set_iversion(&ei->vfs_inode, 1); spin_lock_init(&ei->i_raw_lock); INIT_LIST_HEAD(&ei->i_prealloc_list); + atomic_set(&ei->i_prealloc_active, 0); spin_lock_init(&ei->i_prealloc_lock); ext4_es_init_tree(&ei->i_es_tree); rwlock_init(&ei->i_es_lock); @@ -1220,7 +1221,7 @@ void ext4_clear_inode(struct inode *inode) { invalidate_inode_buffers(inode); clear_inode(inode); - ext4_discard_preallocations(inode); + ext4_discard_preallocations(inode, 0); ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); dquot_drop(inode); if (EXT4_I(inode)->jinode) { diff --git a/fs/ext4/sysfs.c b/fs/ext4/sysfs.c index 7fee11cc30e7..bfabb799fa45 100644 --- a/fs/ext4/sysfs.c +++ b/fs/ext4/sysfs.c @@ -218,6 +218,7 @@ EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan); 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(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); @@ -264,6 +265,7 @@ static struct attribute *ext4_attrs[] = { ATTR_LIST(mb_order2_req), ATTR_LIST(mb_stream_req), ATTR_LIST(mb_group_prealloc), + ATTR_LIST(mb_max_inode_prealloc), ATTR_LIST(max_writeback_mb_bump), ATTR_LIST(extent_max_zeroout_kb), ATTR_LIST(trigger_fs_error), diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 8008d2e116b9..4c8b99ec8606 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -746,24 +746,29 @@ TRACE_EVENT(ext4_mb_release_group_pa, ); TRACE_EVENT(ext4_discard_preallocations, - TP_PROTO(struct inode *inode), + TP_PROTO(struct inode *inode, unsigned int len, unsigned int needed), - TP_ARGS(inode), + TP_ARGS(inode, len, needed), TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( unsigned int, len ) + __field( unsigned int, needed ) ), TP_fast_assign( __entry->dev = inode->i_sb->s_dev; __entry->ino = inode->i_ino; + __entry->len = len; + __entry->needed = needed; ), - TP_printk("dev %d,%d ino %lu", + TP_printk("dev %d,%d ino %lu len: %u needed %u", MAJOR(__entry->dev), MINOR(__entry->dev), - (unsigned long) __entry->ino) + (unsigned long) __entry->ino, __entry->len, + __entry->needed) ); TRACE_EVENT(ext4_mb_discard_preallocations, -- cgit v1.2.3