From b7a2441f9966fe3e1be960a876ab52e6029ea005 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Wed, 21 Jul 2010 22:19:51 -0600 Subject: writeback: remove writeback_control.more_io When wbc.more_io was first introduced, it indicates whether there are at least one superblock whose s_more_io contains more IO work. Now with the per-bdi writeback, it can be replaced with a simple b_more_io test. Acked-by: Jan Kara Acked-by: Mel Gorman Reviewed-by: Minchan Kim Signed-off-by: Wu Fengguang --- include/trace/events/ext4.h | 6 ++---- include/trace/events/writeback.h | 5 +---- 2 files changed, 3 insertions(+), 8 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index e09592d2f916..b225d0d8c87f 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -404,7 +404,6 @@ TRACE_EVENT(ext4_da_writepages_result, __field( int, pages_written ) __field( long, pages_skipped ) __field( int, sync_mode ) - __field( char, more_io ) __field( pgoff_t, writeback_index ) ), @@ -415,16 +414,15 @@ TRACE_EVENT(ext4_da_writepages_result, __entry->pages_written = pages_written; __entry->pages_skipped = wbc->pages_skipped; __entry->sync_mode = wbc->sync_mode; - __entry->more_io = wbc->more_io; __entry->writeback_index = inode->i_mapping->writeback_index; ), TP_printk("dev %d,%d ino %lu ret %d pages_written %d pages_skipped %ld " - " more_io %d sync_mode %d writeback_index %lu", + "sync_mode %d writeback_index %lu", MAJOR(__entry->dev), MINOR(__entry->dev), (unsigned long) __entry->ino, __entry->ret, __entry->pages_written, __entry->pages_skipped, - __entry->more_io, __entry->sync_mode, + __entry->sync_mode, (unsigned long) __entry->writeback_index) ); diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index 4e249b927eaa..b2cfac5f3313 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -101,7 +101,6 @@ DECLARE_EVENT_CLASS(wbc_class, __field(int, for_background) __field(int, for_reclaim) __field(int, range_cyclic) - __field(int, more_io) __field(unsigned long, older_than_this) __field(long, range_start) __field(long, range_end) @@ -116,7 +115,6 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->for_background = wbc->for_background; __entry->for_reclaim = wbc->for_reclaim; __entry->range_cyclic = wbc->range_cyclic; - __entry->more_io = wbc->more_io; __entry->older_than_this = wbc->older_than_this ? *wbc->older_than_this : 0; __entry->range_start = (long)wbc->range_start; @@ -124,7 +122,7 @@ DECLARE_EVENT_CLASS(wbc_class, ), TP_printk("bdi %s: towrt=%ld skip=%ld mode=%d kupd=%d " - "bgrd=%d reclm=%d cyclic=%d more=%d older=0x%lx " + "bgrd=%d reclm=%d cyclic=%d older=0x%lx " "start=0x%lx end=0x%lx", __entry->name, __entry->nr_to_write, @@ -134,7 +132,6 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->for_background, __entry->for_reclaim, __entry->range_cyclic, - __entry->more_io, __entry->older_than_this, __entry->range_start, __entry->range_end) -- cgit v1.3-14-g43fede From 846d5a091b0506b75489577cde27f39b37a192a4 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Thu, 5 May 2011 21:10:38 -0600 Subject: writeback: remove .nonblocking and .encountered_congestion Remove two unused struct writeback_control fields: .encountered_congestion (completely unused) .nonblocking (never set, checked/showed in XFS,NFS/btrfs) The .for_background check in nfs_write_inode() is also removed btw, as .for_background implies WB_SYNC_NONE. Reviewed-by: Jan Kara Proposed-by: Christoph Hellwig Signed-off-by: Wu Fengguang --- fs/nfs/write.c | 3 +-- fs/xfs/linux-2.6/xfs_aops.c | 2 +- include/linux/writeback.h | 2 -- include/trace/events/btrfs.h | 6 ++---- 4 files changed, 4 insertions(+), 9 deletions(-) (limited to 'include/trace') diff --git a/fs/nfs/write.c b/fs/nfs/write.c index e268e3b23497..dd6a6cee39a7 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1564,8 +1564,7 @@ int nfs_write_inode(struct inode *inode, struct writeback_control *wbc) int status; bool sync = true; - if (wbc->sync_mode == WB_SYNC_NONE || wbc->nonblocking || - wbc->for_background) + if (wbc->sync_mode == WB_SYNC_NONE) sync = false; status = pnfs_layoutcommit_inode(inode, sync); diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 79ce38be15a1..7559861481aa 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -970,7 +970,7 @@ xfs_vm_writepage( offset = page_offset(page); type = IO_OVERWRITE; - if (wbc->sync_mode == WB_SYNC_NONE && wbc->nonblocking) + if (wbc->sync_mode == WB_SYNC_NONE) nonblocking = 1; do { diff --git a/include/linux/writeback.h b/include/linux/writeback.h index 8797b20dd22b..2f1b512bd6e0 100644 --- a/include/linux/writeback.h +++ b/include/linux/writeback.h @@ -39,8 +39,6 @@ struct writeback_control { loff_t range_start; loff_t range_end; - unsigned nonblocking:1; /* Don't get stuck on request queues */ - unsigned encountered_congestion:1; /* An output: a queue is full */ unsigned for_kupdate:1; /* A kupdate writeback */ unsigned for_background:1; /* A background writeback */ unsigned tagged_writepages:1; /* tag-and-write to avoid livelock */ diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index 4114129f0794..b31702ac15be 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -284,7 +284,6 @@ DECLARE_EVENT_CLASS(btrfs__writepage, __field( long, pages_skipped ) __field( loff_t, range_start ) __field( loff_t, range_end ) - __field( char, nonblocking ) __field( char, for_kupdate ) __field( char, for_reclaim ) __field( char, range_cyclic ) @@ -299,7 +298,6 @@ DECLARE_EVENT_CLASS(btrfs__writepage, __entry->pages_skipped = wbc->pages_skipped; __entry->range_start = wbc->range_start; __entry->range_end = wbc->range_end; - __entry->nonblocking = wbc->nonblocking; __entry->for_kupdate = wbc->for_kupdate; __entry->for_reclaim = wbc->for_reclaim; __entry->range_cyclic = wbc->range_cyclic; @@ -310,13 +308,13 @@ DECLARE_EVENT_CLASS(btrfs__writepage, TP_printk("root = %llu(%s), ino = %lu, page_index = %lu, " "nr_to_write = %ld, pages_skipped = %ld, range_start = %llu, " - "range_end = %llu, nonblocking = %d, for_kupdate = %d, " + "range_end = %llu, for_kupdate = %d, " "for_reclaim = %d, range_cyclic = %d, writeback_index = %lu", show_root_type(__entry->root_objectid), (unsigned long)__entry->ino, __entry->index, __entry->nr_to_write, __entry->pages_skipped, __entry->range_start, __entry->range_end, - __entry->nonblocking, __entry->for_kupdate, + __entry->for_kupdate, __entry->for_reclaim, __entry->range_cyclic, (unsigned long)__entry->writeback_index) ); -- cgit v1.3-14-g43fede From 251d6a471c831e22880b3c146bb4556ddfb1dc82 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Wed, 1 Dec 2010 17:33:37 -0600 Subject: writeback: trace event writeback_single_inode It is valuable to know how the dirty inodes are iterated and their IO size. "writeback_single_inode: bdi 8:0: ino=134246746 state=I_DIRTY_SYNC|I_SYNC age=414 index=0 to_write=1024 wrote=0" - "state" reflects inode->i_state at the end of writeback_single_inode() - "index" reflects mapping->writeback_index after the ->writepages() call - "to_write" is the wbc->nr_to_write at entrance of writeback_single_inode() - "wrote" is the number of pages actually written v2: add trace event writeback_single_inode_requeue as proposed by Dave. CC: Dave Chinner Signed-off-by: Wu Fengguang --- fs/fs-writeback.c | 4 +++ include/trace/events/writeback.h | 70 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 74 insertions(+) (limited to 'include/trace') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index f43c479feee9..5185fad48b62 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -346,6 +346,7 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, struct writeback_control *wbc) { struct address_space *mapping = inode->i_mapping; + long nr_to_write = wbc->nr_to_write; unsigned dirty; int ret; @@ -368,6 +369,8 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, */ if (wbc->sync_mode != WB_SYNC_ALL) { requeue_io(inode, wb); + trace_writeback_single_inode_requeue(inode, wbc, + nr_to_write); return 0; } @@ -467,6 +470,7 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, } } inode_sync_complete(inode); + trace_writeback_single_inode(inode, wbc, nr_to_write); return ret; } diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index b2cfac5f3313..898277bc89b4 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -8,6 +8,19 @@ #include #include +#define show_inode_state(state) \ + __print_flags(state, "|", \ + {I_DIRTY_SYNC, "I_DIRTY_SYNC"}, \ + {I_DIRTY_DATASYNC, "I_DIRTY_DATASYNC"}, \ + {I_DIRTY_PAGES, "I_DIRTY_PAGES"}, \ + {I_NEW, "I_NEW"}, \ + {I_WILL_FREE, "I_WILL_FREE"}, \ + {I_FREEING, "I_FREEING"}, \ + {I_CLEAR, "I_CLEAR"}, \ + {I_SYNC, "I_SYNC"}, \ + {I_REFERENCED, "I_REFERENCED"} \ + ) + struct wb_writeback_work; DECLARE_EVENT_CLASS(writeback_work_class, @@ -184,6 +197,63 @@ DEFINE_EVENT(writeback_congest_waited_template, writeback_wait_iff_congested, TP_ARGS(usec_timeout, usec_delayed) ); +DECLARE_EVENT_CLASS(writeback_single_inode_template, + + TP_PROTO(struct inode *inode, + struct writeback_control *wbc, + unsigned long nr_to_write + ), + + TP_ARGS(inode, wbc, nr_to_write), + + TP_STRUCT__entry( + __array(char, name, 32) + __field(unsigned long, ino) + __field(unsigned long, state) + __field(unsigned long, age) + __field(unsigned long, writeback_index) + __field(long, nr_to_write) + __field(unsigned long, wrote) + ), + + TP_fast_assign( + strncpy(__entry->name, + dev_name(inode->i_mapping->backing_dev_info->dev), 32); + __entry->ino = inode->i_ino; + __entry->state = inode->i_state; + __entry->age = (jiffies - inode->dirtied_when) * + 1000 / HZ; + __entry->writeback_index = inode->i_mapping->writeback_index; + __entry->nr_to_write = nr_to_write; + __entry->wrote = nr_to_write - wbc->nr_to_write; + ), + + TP_printk("bdi %s: ino=%lu state=%s age=%lu " + "index=%lu to_write=%ld wrote=%lu", + __entry->name, + __entry->ino, + show_inode_state(__entry->state), + __entry->age, + __entry->writeback_index, + __entry->nr_to_write, + __entry->wrote + ) +); + +DEFINE_EVENT(writeback_single_inode_template, writeback_single_inode_requeue, + TP_PROTO(struct inode *inode, + struct writeback_control *wbc, + unsigned long nr_to_write), + TP_ARGS(inode, wbc, nr_to_write) +); + +DEFINE_EVENT(writeback_single_inode_template, writeback_single_inode, + TP_PROTO(struct inode *inode, + struct writeback_control *wbc, + unsigned long nr_to_write), + TP_ARGS(inode, wbc, nr_to_write) +); + #endif /* _TRACE_WRITEBACK_H */ /* This part must be outside protection */ -- cgit v1.3-14-g43fede From e84d0a4f8e39a73003a6ec9a11b07702745f4c1f Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Sat, 23 Apr 2011 12:27:27 -0600 Subject: writeback: trace event writeback_queue_io Note that it adds a little overheads to account the moved/enqueued inodes from b_dirty to b_io. The "moved" accounting may be later used to limit the number of inodes that can be moved in one shot, in order to keep spinlock hold time under control. Signed-off-by: Wu Fengguang --- fs/fs-writeback.c | 14 ++++++++++---- include/trace/events/writeback.h | 25 +++++++++++++++++++++++++ 2 files changed, 35 insertions(+), 4 deletions(-) (limited to 'include/trace') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 5185fad48b62..6caa98247a5b 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -248,15 +248,16 @@ static bool inode_dirtied_after(struct inode *inode, unsigned long t) /* * Move expired dirty inodes from @delaying_queue to @dispatch_queue. */ -static void move_expired_inodes(struct list_head *delaying_queue, +static int move_expired_inodes(struct list_head *delaying_queue, struct list_head *dispatch_queue, - unsigned long *older_than_this) + unsigned long *older_than_this) { LIST_HEAD(tmp); struct list_head *pos, *node; struct super_block *sb = NULL; struct inode *inode; int do_sb_sort = 0; + int moved = 0; while (!list_empty(delaying_queue)) { inode = wb_inode(delaying_queue->prev); @@ -267,12 +268,13 @@ static void move_expired_inodes(struct list_head *delaying_queue, do_sb_sort = 1; sb = inode->i_sb; list_move(&inode->i_wb_list, &tmp); + moved++; } /* just one sb in list, splice to dispatch_queue and we're done */ if (!do_sb_sort) { list_splice(&tmp, dispatch_queue); - return; + goto out; } /* Move inodes from one superblock together */ @@ -284,6 +286,8 @@ static void move_expired_inodes(struct list_head *delaying_queue, list_move(&inode->i_wb_list, dispatch_queue); } } +out: + return moved; } /* @@ -299,9 +303,11 @@ static void move_expired_inodes(struct list_head *delaying_queue, */ static void queue_io(struct bdi_writeback *wb, unsigned long *older_than_this) { + int moved; assert_spin_locked(&wb->list_lock); list_splice_init(&wb->b_more_io, &wb->b_io); - move_expired_inodes(&wb->b_dirty, &wb->b_io, older_than_this); + moved = move_expired_inodes(&wb->b_dirty, &wb->b_io, older_than_this); + trace_writeback_queue_io(wb, older_than_this, moved); } static int write_inode(struct inode *inode, struct writeback_control *wbc) diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index 898277bc89b4..205d14919ef2 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -162,6 +162,31 @@ DEFINE_WBC_EVENT(wbc_balance_dirty_written); DEFINE_WBC_EVENT(wbc_balance_dirty_wait); DEFINE_WBC_EVENT(wbc_writepage); +TRACE_EVENT(writeback_queue_io, + TP_PROTO(struct bdi_writeback *wb, + unsigned long *older_than_this, + int moved), + TP_ARGS(wb, older_than_this, moved), + TP_STRUCT__entry( + __array(char, name, 32) + __field(unsigned long, older) + __field(long, age) + __field(int, moved) + ), + TP_fast_assign( + strncpy(__entry->name, dev_name(wb->bdi->dev), 32); + __entry->older = older_than_this ? *older_than_this : 0; + __entry->age = older_than_this ? + (jiffies - *older_than_this) * 1000 / HZ : -1; + __entry->moved = moved; + ), + TP_printk("bdi %s: older=%lu age=%ld enqueue=%d", + __entry->name, + __entry->older, /* older_than_this in jiffies */ + __entry->age, /* older_than_this in relative milliseconds */ + __entry->moved) +); + DECLARE_EVENT_CLASS(writeback_congest_waited_template, TP_PROTO(unsigned int usec_timeout, unsigned int usec_delayed), -- cgit v1.3-14-g43fede From 785c4bcc0d88ff006a0b2120815a71e86ecf21ce Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Mon, 23 May 2011 18:33:01 +0200 Subject: ext3: Add fixed tracepoints This commit adds fixed tracepoints to the ext3 code. It is based on ext4 tracepoints, however due to the differences of both file systems, there are some tracepoints missing (those for delaloc and for multi-block allocator) and there are some ext3 specific as well (for reservation windows). Here is a list: ext3_free_inode ext3_request_inode ext3_allocate_inode ext3_evict_inode ext3_drop_inode ext3_mark_inode_dirty ext3_write_begin ext3_ordered_write_end ext3_writeback_write_end ext3_journalled_write_end ext3_ordered_writepage ext3_writeback_writepage ext3_journalled_writepage ext3_readpage ext3_releasepage ext3_invalidatepage ext3_discard_blocks ext3_request_blocks ext3_allocate_blocks ext3_free_blocks ext3_sync_file_enter ext3_sync_file_exit ext3_sync_fs ext3_rsv_window_add ext3_discard_reservation ext3_alloc_new_reservation ext3_reserved ext3_forget ext3_read_block_bitmap ext3_direct_IO_enter ext3_direct_IO_exit ext3_unlink_enter ext3_unlink_exit ext3_truncate_enter ext3_truncate_exit ext3_get_blocks_enter ext3_get_blocks_exit ext3_load_inode Signed-off-by: Lukas Czerner Cc: Jan Kara Signed-off-by: Jan Kara --- fs/ext3/balloc.c | 34 +- fs/ext3/fsync.c | 15 +- fs/ext3/ialloc.c | 4 + fs/ext3/inode.c | 29 ++ fs/ext3/namei.c | 3 + fs/ext3/super.c | 13 + include/trace/events/ext3.h | 864 ++++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 946 insertions(+), 16 deletions(-) create mode 100644 include/trace/events/ext3.h (limited to 'include/trace') diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index fe52297e31ad..f7d111e499ad 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -21,6 +21,7 @@ #include #include #include +#include /* * balloc.c contains the blocks allocation and deallocation routines @@ -161,6 +162,7 @@ read_block_bitmap(struct super_block *sb, unsigned int block_group) desc = ext3_get_group_desc(sb, block_group, NULL); if (!desc) return NULL; + trace_ext3_read_block_bitmap(sb, block_group); bitmap_blk = le32_to_cpu(desc->bg_block_bitmap); bh = sb_getblk(sb, bitmap_blk); if (unlikely(!bh)) { @@ -351,6 +353,7 @@ void ext3_rsv_window_add(struct super_block *sb, struct rb_node * parent = NULL; struct ext3_reserve_window_node *this; + trace_ext3_rsv_window_add(sb, rsv); while (*p) { parent = *p; @@ -476,8 +479,10 @@ void ext3_discard_reservation(struct inode *inode) rsv = &block_i->rsv_window_node; if (!rsv_is_empty(&rsv->rsv_window)) { spin_lock(rsv_lock); - if (!rsv_is_empty(&rsv->rsv_window)) + if (!rsv_is_empty(&rsv->rsv_window)) { + trace_ext3_discard_reservation(inode, rsv); rsv_window_remove(inode->i_sb, rsv); + } spin_unlock(rsv_lock); } } @@ -683,14 +688,10 @@ error_return: void ext3_free_blocks(handle_t *handle, struct inode *inode, ext3_fsblk_t block, unsigned long count) { - struct super_block * sb; + struct super_block *sb = inode->i_sb; unsigned long dquot_freed_blocks; - sb = inode->i_sb; - if (!sb) { - printk ("ext3_free_blocks: nonexistent device"); - return; - } + trace_ext3_free_blocks(inode, block, count); ext3_free_blocks_sb(handle, sb, block, count, &dquot_freed_blocks); if (dquot_freed_blocks) dquot_free_block(inode, dquot_freed_blocks); @@ -1136,6 +1137,7 @@ static int alloc_new_reservation(struct ext3_reserve_window_node *my_rsv, else start_block = grp_goal + group_first_block; + trace_ext3_alloc_new_reservation(sb, start_block); size = my_rsv->rsv_goal_size; if (!rsv_is_empty(&my_rsv->rsv_window)) { @@ -1230,8 +1232,11 @@ retry: * check if the first free block is within the * free space we just reserved */ - if (start_block >= my_rsv->rsv_start && start_block <= my_rsv->rsv_end) + if (start_block >= my_rsv->rsv_start && + start_block <= my_rsv->rsv_end) { + trace_ext3_reserved(sb, start_block, my_rsv); return 0; /* success */ + } /* * if the first free bit we found is out of the reservable space * continue search for next reservable space, @@ -1514,10 +1519,6 @@ ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode, *errp = -ENOSPC; sb = inode->i_sb; - if (!sb) { - printk("ext3_new_block: nonexistent device"); - return 0; - } /* * Check quota for allocation of this block. @@ -1528,8 +1529,10 @@ ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode, return 0; } + trace_ext3_request_blocks(inode, goal, num); + sbi = EXT3_SB(sb); - es = EXT3_SB(sb)->s_es; + es = sbi->s_es; ext3_debug("goal=%lu.\n", goal); /* * Allocate a block from reservation only when @@ -1742,6 +1745,10 @@ allocated: brelse(bitmap_bh); dquot_free_block(inode, *count-num); *count = num; + + trace_ext3_allocate_blocks(inode, goal, num, + (unsigned long long)ret_block); + return ret_block; io_error: @@ -1996,6 +2003,7 @@ ext3_grpblk_t ext3_trim_all_free(struct super_block *sb, unsigned int group, if ((next - start) < minblocks) goto free_extent; + trace_ext3_discard_blocks(sb, discard_block, next - start); /* Send the TRIM command down to the device */ err = sb_issue_discard(sb, discard_block, next - start, GFP_NOFS, 0); diff --git a/fs/ext3/fsync.c b/fs/ext3/fsync.c index 09b13bb34c94..06a4394d2bc3 100644 --- a/fs/ext3/fsync.c +++ b/fs/ext3/fsync.c @@ -30,6 +30,7 @@ #include #include #include +#include /* * akpm: A new design for ext3_sync_file(). @@ -51,10 +52,13 @@ int ext3_sync_file(struct file *file, int datasync) int ret, needs_barrier = 0; tid_t commit_tid; + J_ASSERT(ext3_journal_current_handle() == NULL); + + trace_ext3_sync_file_enter(file, datasync); + if (inode->i_sb->s_flags & MS_RDONLY) return 0; - J_ASSERT(ext3_journal_current_handle() == NULL); /* * data=writeback,ordered: @@ -70,8 +74,10 @@ int ext3_sync_file(struct file *file, int datasync) * (they were dirtied by commit). But that's OK - the blocks are * safe in-journal, which is all fsync() needs to ensure. */ - if (ext3_should_journal_data(inode)) - return ext3_force_commit(inode->i_sb); + if (ext3_should_journal_data(inode)) { + ret = ext3_force_commit(inode->i_sb); + goto out; + } if (datasync) commit_tid = atomic_read(&ei->i_datasync_tid); @@ -91,5 +97,8 @@ int ext3_sync_file(struct file *file, int datasync) */ if (needs_barrier) blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); + +out: + trace_ext3_sync_file_exit(inode, ret); return ret; } diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c index bfc2dc43681d..bf09cbf938cc 100644 --- a/fs/ext3/ialloc.c +++ b/fs/ext3/ialloc.c @@ -23,6 +23,7 @@ #include #include #include +#include #include @@ -118,6 +119,7 @@ void ext3_free_inode (handle_t *handle, struct inode * inode) ino = inode->i_ino; ext3_debug ("freeing inode %lu\n", ino); + trace_ext3_free_inode(inode); is_directory = S_ISDIR(inode->i_mode); @@ -426,6 +428,7 @@ struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, return ERR_PTR(-EPERM); sb = dir->i_sb; + trace_ext3_request_inode(dir, mode); inode = new_inode(sb); if (!inode) return ERR_PTR(-ENOMEM); @@ -601,6 +604,7 @@ got: } ext3_debug("allocating inode %lu\n", inode->i_ino); + trace_ext3_allocate_inode(inode, dir, mode); goto really_out; fail: ext3_std_error(sb, err); diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 3451d23c3bae..3aa05eebe0b8 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -38,6 +38,7 @@ #include #include #include +#include #include "xattr.h" #include "acl.h" @@ -70,6 +71,7 @@ int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode, might_sleep(); + trace_ext3_forget(inode, is_metadata, blocknr); BUFFER_TRACE(bh, "enter"); jbd_debug(4, "forgetting bh %p: is_metadata = %d, mode %o, " @@ -198,6 +200,7 @@ void ext3_evict_inode (struct inode *inode) handle_t *handle; int want_delete = 0; + trace_ext3_evict_inode(inode); if (!inode->i_nlink && !is_bad_inode(inode)) { dquot_initialize(inode); want_delete = 1; @@ -842,6 +845,7 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, ext3_fsblk_t first_block = 0; + trace_ext3_get_blocks_enter(inode, iblock, maxblocks, create); J_ASSERT(handle != NULL || create == 0); depth = ext3_block_to_path(inode,iblock,offsets,&blocks_to_boundary); @@ -970,6 +974,9 @@ cleanup: } BUFFER_TRACE(bh_result, "returned"); out: + trace_ext3_get_blocks_exit(inode, iblock, + depth ? le32_to_cpu(chain[depth-1].key) : 0, + count, err); return err; } @@ -1217,6 +1224,8 @@ static int ext3_write_begin(struct file *file, struct address_space *mapping, * we allocate blocks but write fails for some reason */ int needed_blocks = ext3_writepage_trans_blocks(inode) + 1; + trace_ext3_write_begin(inode, pos, len, flags); + index = pos >> PAGE_CACHE_SHIFT; from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; @@ -1332,6 +1341,7 @@ static int ext3_ordered_write_end(struct file *file, unsigned from, to; int ret = 0, ret2; + trace_ext3_ordered_write_end(inode, pos, len, copied); copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); from = pos & (PAGE_CACHE_SIZE - 1); @@ -1367,6 +1377,7 @@ static int ext3_writeback_write_end(struct file *file, struct inode *inode = file->f_mapping->host; int ret; + trace_ext3_writeback_write_end(inode, pos, len, copied); copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); update_file_sizes(inode, pos, copied); /* @@ -1395,6 +1406,7 @@ static int ext3_journalled_write_end(struct file *file, int partial = 0; unsigned from, to; + trace_ext3_journalled_write_end(inode, pos, len, copied); from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; @@ -1577,6 +1589,7 @@ static int ext3_ordered_writepage(struct page *page, if (ext3_journal_current_handle()) goto out_fail; + trace_ext3_ordered_writepage(page); if (!page_has_buffers(page)) { create_empty_buffers(page, inode->i_sb->s_blocksize, (1 << BH_Dirty)|(1 << BH_Uptodate)); @@ -1647,6 +1660,7 @@ static int ext3_writeback_writepage(struct page *page, if (ext3_journal_current_handle()) goto out_fail; + trace_ext3_writeback_writepage(page); if (page_has_buffers(page)) { if (!walk_page_buffers(NULL, page_buffers(page), 0, PAGE_CACHE_SIZE, NULL, buffer_unmapped)) { @@ -1689,6 +1703,7 @@ static int ext3_journalled_writepage(struct page *page, if (ext3_journal_current_handle()) goto no_write; + trace_ext3_journalled_writepage(page); handle = ext3_journal_start(inode, ext3_writepage_trans_blocks(inode)); if (IS_ERR(handle)) { ret = PTR_ERR(handle); @@ -1739,6 +1754,7 @@ out_unlock: static int ext3_readpage(struct file *file, struct page *page) { + trace_ext3_readpage(page); return mpage_readpage(page, ext3_get_block); } @@ -1753,6 +1769,8 @@ static void ext3_invalidatepage(struct page *page, unsigned long offset) { journal_t *journal = EXT3_JOURNAL(page->mapping->host); + trace_ext3_invalidatepage(page, offset); + /* * If it's a full truncate we just forget about the pending dirtying */ @@ -1766,6 +1784,7 @@ static int ext3_releasepage(struct page *page, gfp_t wait) { journal_t *journal = EXT3_JOURNAL(page->mapping->host); + trace_ext3_releasepage(page); WARN_ON(PageChecked(page)); if (!page_has_buffers(page)) return 0; @@ -1794,6 +1813,8 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_length(iov, nr_segs); int retries = 0; + trace_ext3_direct_IO_enter(inode, offset, iov_length(iov, nr_segs), rw); + if (rw == WRITE) { loff_t final_size = offset + count; @@ -1868,6 +1889,8 @@ retry: ret = err; } out: + trace_ext3_direct_IO_exit(inode, offset, + iov_length(iov, nr_segs), rw, ret); return ret; } @@ -2446,6 +2469,8 @@ void ext3_truncate(struct inode *inode) unsigned blocksize = inode->i_sb->s_blocksize; struct page *page; + trace_ext3_truncate_enter(inode); + if (!ext3_can_truncate(inode)) goto out_notrans; @@ -2597,6 +2622,7 @@ out_stop: ext3_orphan_del(handle, inode); ext3_journal_stop(handle); + trace_ext3_truncate_exit(inode); return; out_notrans: /* @@ -2605,6 +2631,7 @@ out_notrans: */ if (inode->i_nlink) ext3_orphan_del(NULL, inode); + trace_ext3_truncate_exit(inode); } static ext3_fsblk_t ext3_get_inode_block(struct super_block *sb, @@ -2746,6 +2773,7 @@ make_io: * has in-inode xattrs, or we don't have this inode in memory. * Read the block from disk. */ + trace_ext3_load_inode(inode); get_bh(bh); bh->b_end_io = end_buffer_read_sync; submit_bh(READ_META, bh); @@ -3372,6 +3400,7 @@ int ext3_mark_inode_dirty(handle_t *handle, struct inode *inode) int err; might_sleep(); + trace_ext3_mark_inode_dirty(inode, _RET_IP_); err = ext3_reserve_inode_write(handle, inode, &iloc); if (!err) err = ext3_mark_iloc_dirty(handle, inode, &iloc); diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 34b6d9bfc48a..51736a4ff0cd 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -36,6 +36,7 @@ #include #include #include +#include #include "namei.h" #include "xattr.h" @@ -2144,6 +2145,7 @@ static int ext3_unlink(struct inode * dir, struct dentry *dentry) struct ext3_dir_entry_2 * de; handle_t *handle; + trace_ext3_unlink_enter(dir, dentry); /* Initialize quotas before so that eventual writes go * in separate transaction */ dquot_initialize(dir); @@ -2189,6 +2191,7 @@ static int ext3_unlink(struct inode * dir, struct dentry *dentry) end_unlink: ext3_journal_stop(handle); brelse (bh); + trace_ext3_unlink_exit(dentry, retval); return retval; } diff --git a/fs/ext3/super.c b/fs/ext3/super.c index aad153ef6b78..662290fb6fff 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -44,6 +44,9 @@ #include "acl.h" #include "namei.h" +#define CREATE_TRACE_POINTS +#include + #ifdef CONFIG_EXT3_DEFAULTS_TO_ORDERED #define EXT3_MOUNT_DEFAULT_DATA_MODE EXT3_MOUNT_ORDERED_DATA #else @@ -497,6 +500,14 @@ static struct inode *ext3_alloc_inode(struct super_block *sb) return &ei->vfs_inode; } +static int ext3_drop_inode(struct inode *inode) +{ + int drop = generic_drop_inode(inode); + + trace_ext3_drop_inode(inode, drop); + return drop; +} + static void ext3_i_callback(struct rcu_head *head) { struct inode *inode = container_of(head, struct inode, i_rcu); @@ -788,6 +799,7 @@ static const struct super_operations ext3_sops = { .destroy_inode = ext3_destroy_inode, .write_inode = ext3_write_inode, .dirty_inode = ext3_dirty_inode, + .drop_inode = ext3_drop_inode, .evict_inode = ext3_evict_inode, .put_super = ext3_put_super, .sync_fs = ext3_sync_fs, @@ -2507,6 +2519,7 @@ static int ext3_sync_fs(struct super_block *sb, int wait) { tid_t target; + trace_ext3_sync_fs(sb, wait); if (journal_start_commit(EXT3_SB(sb)->s_journal, &target)) { if (wait) log_wait_commit(EXT3_SB(sb)->s_journal, target); diff --git a/include/trace/events/ext3.h b/include/trace/events/ext3.h new file mode 100644 index 000000000000..7b53c0573dc9 --- /dev/null +++ b/include/trace/events/ext3.h @@ -0,0 +1,864 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM ext3 + +#if !defined(_TRACE_EXT3_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EXT3_H + +#include + +TRACE_EVENT(ext3_free_inode, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( umode_t, mode ) + __field( uid_t, uid ) + __field( gid_t, gid ) + __field( blkcnt_t, blocks ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->mode = inode->i_mode; + __entry->uid = inode->i_uid; + __entry->gid = inode->i_gid; + __entry->blocks = inode->i_blocks; + ), + + TP_printk("dev %d,%d ino %lu mode 0%o uid %u gid %u blocks %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->mode, __entry->uid, __entry->gid, + (unsigned long) __entry->blocks) +); + +TRACE_EVENT(ext3_request_inode, + TP_PROTO(struct inode *dir, int mode), + + TP_ARGS(dir, mode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, dir ) + __field( umode_t, mode ) + ), + + TP_fast_assign( + __entry->dev = dir->i_sb->s_dev; + __entry->dir = dir->i_ino; + __entry->mode = mode; + ), + + TP_printk("dev %d,%d dir %lu mode 0%o", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->dir, __entry->mode) +); + +TRACE_EVENT(ext3_allocate_inode, + TP_PROTO(struct inode *inode, struct inode *dir, int mode), + + TP_ARGS(inode, dir, mode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( ino_t, dir ) + __field( umode_t, mode ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->dir = dir->i_ino; + __entry->mode = mode; + ), + + TP_printk("dev %d,%d ino %lu dir %lu mode 0%o", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long) __entry->dir, __entry->mode) +); + +TRACE_EVENT(ext3_evict_inode, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( int, nlink ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->nlink = inode->i_nlink; + ), + + TP_printk("dev %d,%d ino %lu nlink %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, __entry->nlink) +); + +TRACE_EVENT(ext3_drop_inode, + TP_PROTO(struct inode *inode, int drop), + + TP_ARGS(inode, drop), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( int, drop ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->drop = drop; + ), + + TP_printk("dev %d,%d ino %lu drop %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, __entry->drop) +); + +TRACE_EVENT(ext3_mark_inode_dirty, + TP_PROTO(struct inode *inode, unsigned long IP), + + TP_ARGS(inode, IP), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field(unsigned long, ip ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->ip = IP; + ), + + TP_printk("dev %d,%d ino %lu caller %pF", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, (void *)__entry->ip) +); + +TRACE_EVENT(ext3_write_begin, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int flags), + + TP_ARGS(inode, pos, len, flags), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, flags ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->flags = flags; + ), + + TP_printk("dev %d,%d ino %lu pos %llu len %u flags %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long long) __entry->pos, __entry->len, + __entry->flags) +); + +DECLARE_EVENT_CLASS(ext3__write_end, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, copied ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->copied = copied; + ), + + TP_printk("dev %d,%d ino %lu pos %llu len %u copied %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long long) __entry->pos, __entry->len, + __entry->copied) +); + +DEFINE_EVENT(ext3__write_end, ext3_ordered_write_end, + + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied) +); + +DEFINE_EVENT(ext3__write_end, ext3_writeback_write_end, + + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied) +); + +DEFINE_EVENT(ext3__write_end, ext3_journalled_write_end, + + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied) +); + +DECLARE_EVENT_CLASS(ext3__page_op, + TP_PROTO(struct page *page), + + TP_ARGS(page), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( pgoff_t, index ) + + ), + + TP_fast_assign( + __entry->index = page->index; + __entry->ino = page->mapping->host->i_ino; + __entry->dev = page->mapping->host->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu page_index %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, __entry->index) +); + +DEFINE_EVENT(ext3__page_op, ext3_ordered_writepage, + + TP_PROTO(struct page *page), + + TP_ARGS(page) +); + +DEFINE_EVENT(ext3__page_op, ext3_writeback_writepage, + + TP_PROTO(struct page *page), + + TP_ARGS(page) +); + +DEFINE_EVENT(ext3__page_op, ext3_journalled_writepage, + + TP_PROTO(struct page *page), + + TP_ARGS(page) +); + +DEFINE_EVENT(ext3__page_op, ext3_readpage, + + TP_PROTO(struct page *page), + + TP_ARGS(page) +); + +DEFINE_EVENT(ext3__page_op, ext3_releasepage, + + TP_PROTO(struct page *page), + + TP_ARGS(page) +); + +TRACE_EVENT(ext3_invalidatepage, + TP_PROTO(struct page *page, unsigned long offset), + + TP_ARGS(page, offset), + + TP_STRUCT__entry( + __field( pgoff_t, index ) + __field( unsigned long, offset ) + __field( ino_t, ino ) + __field( dev_t, dev ) + + ), + + TP_fast_assign( + __entry->index = page->index; + __entry->offset = offset; + __entry->ino = page->mapping->host->i_ino; + __entry->dev = page->mapping->host->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu page_index %lu offset %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->index, __entry->offset) +); + +TRACE_EVENT(ext3_discard_blocks, + TP_PROTO(struct super_block *sb, unsigned long blk, + unsigned long count), + + TP_ARGS(sb, blk, count), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( unsigned long, blk ) + __field( unsigned long, count ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->blk = blk; + __entry->count = count; + ), + + TP_printk("dev %d,%d blk %lu count %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->blk, __entry->count) +); + +TRACE_EVENT(ext3_request_blocks, + TP_PROTO(struct inode *inode, unsigned long goal, + unsigned long count), + + TP_ARGS(inode, goal, count), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( unsigned long, count ) + __field( unsigned long, goal ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->count = count; + __entry->goal = goal; + ), + + TP_printk("dev %d,%d ino %lu count %lu goal %lu ", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->count, __entry->goal) +); + +TRACE_EVENT(ext3_allocate_blocks, + TP_PROTO(struct inode *inode, unsigned long goal, + unsigned long count, unsigned long block), + + TP_ARGS(inode, goal, count, block), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( unsigned long, block ) + __field( unsigned long, count ) + __field( unsigned long, goal ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->block = block; + __entry->count = count; + __entry->goal = goal; + ), + + TP_printk("dev %d,%d ino %lu count %lu block %lu goal %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->count, __entry->block, + __entry->goal) +); + +TRACE_EVENT(ext3_free_blocks, + TP_PROTO(struct inode *inode, unsigned long block, + unsigned long count), + + TP_ARGS(inode, block, count), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( umode_t, mode ) + __field( unsigned long, block ) + __field( unsigned long, count ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->mode = inode->i_mode; + __entry->block = block; + __entry->count = count; + ), + + TP_printk("dev %d,%d ino %lu mode 0%o block %lu count %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->mode, __entry->block, __entry->count) +); + +TRACE_EVENT(ext3_sync_file_enter, + TP_PROTO(struct file *file, int datasync), + + TP_ARGS(file, datasync), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( ino_t, parent ) + __field( int, datasync ) + ), + + TP_fast_assign( + struct dentry *dentry = file->f_path.dentry; + + __entry->dev = dentry->d_inode->i_sb->s_dev; + __entry->ino = dentry->d_inode->i_ino; + __entry->datasync = datasync; + __entry->parent = dentry->d_parent->d_inode->i_ino; + ), + + TP_printk("dev %d,%d ino %lu parent %ld datasync %d ", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long) __entry->parent, __entry->datasync) +); + +TRACE_EVENT(ext3_sync_file_exit, + TP_PROTO(struct inode *inode, int ret), + + TP_ARGS(inode, ret), + + TP_STRUCT__entry( + __field( int, ret ) + __field( ino_t, ino ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->ret = ret; + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu ret %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->ret) +); + +TRACE_EVENT(ext3_sync_fs, + TP_PROTO(struct super_block *sb, int wait), + + TP_ARGS(sb, wait), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, wait ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->wait = wait; + ), + + TP_printk("dev %d,%d wait %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->wait) +); + +TRACE_EVENT(ext3_rsv_window_add, + TP_PROTO(struct super_block *sb, + struct ext3_reserve_window_node *rsv_node), + + TP_ARGS(sb, rsv_node), + + TP_STRUCT__entry( + __field( unsigned long, start ) + __field( unsigned long, end ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->start = rsv_node->rsv_window._rsv_start; + __entry->end = rsv_node->rsv_window._rsv_end; + ), + + TP_printk("dev %d,%d start %lu end %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->start, __entry->end) +); + +TRACE_EVENT(ext3_discard_reservation, + TP_PROTO(struct inode *inode, + struct ext3_reserve_window_node *rsv_node), + + TP_ARGS(inode, rsv_node), + + TP_STRUCT__entry( + __field( unsigned long, start ) + __field( unsigned long, end ) + __field( ino_t, ino ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->start = rsv_node->rsv_window._rsv_start; + __entry->end = rsv_node->rsv_window._rsv_end; + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu start %lu end %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long)__entry->ino, __entry->start, + __entry->end) +); + +TRACE_EVENT(ext3_alloc_new_reservation, + TP_PROTO(struct super_block *sb, unsigned long goal), + + TP_ARGS(sb, goal), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( unsigned long, goal ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->goal = goal; + ), + + TP_printk("dev %d,%d goal %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->goal) +); + +TRACE_EVENT(ext3_reserved, + TP_PROTO(struct super_block *sb, unsigned long block, + struct ext3_reserve_window_node *rsv_node), + + TP_ARGS(sb, block, rsv_node), + + TP_STRUCT__entry( + __field( unsigned long, block ) + __field( unsigned long, start ) + __field( unsigned long, end ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->block = block; + __entry->start = rsv_node->rsv_window._rsv_start; + __entry->end = rsv_node->rsv_window._rsv_end; + __entry->dev = sb->s_dev; + ), + + TP_printk("dev %d,%d block %lu, start %lu end %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->block, __entry->start, __entry->end) +); + +TRACE_EVENT(ext3_forget, + TP_PROTO(struct inode *inode, int is_metadata, unsigned long block), + + TP_ARGS(inode, is_metadata, block), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( umode_t, mode ) + __field( int, is_metadata ) + __field( unsigned long, block ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->mode = inode->i_mode; + __entry->is_metadata = is_metadata; + __entry->block = block; + ), + + TP_printk("dev %d,%d ino %lu mode 0%o is_metadata %d block %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->mode, __entry->is_metadata, __entry->block) +); + +TRACE_EVENT(ext3_read_block_bitmap, + TP_PROTO(struct super_block *sb, unsigned int 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) +); + +TRACE_EVENT(ext3_direct_IO_enter, + TP_PROTO(struct inode *inode, loff_t offset, unsigned long len, int rw), + + TP_ARGS(inode, offset, len, rw), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( loff_t, pos ) + __field( unsigned long, len ) + __field( int, rw ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + __entry->pos = offset; + __entry->len = len; + __entry->rw = rw; + ), + + TP_printk("dev %d,%d ino %lu pos %llu len %lu rw %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long long) __entry->pos, __entry->len, + __entry->rw) +); + +TRACE_EVENT(ext3_direct_IO_exit, + TP_PROTO(struct inode *inode, loff_t offset, unsigned long len, + int rw, int ret), + + TP_ARGS(inode, offset, len, rw, ret), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( loff_t, pos ) + __field( unsigned long, len ) + __field( int, rw ) + __field( int, ret ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + __entry->pos = offset; + __entry->len = len; + __entry->rw = rw; + __entry->ret = ret; + ), + + TP_printk("dev %d,%d ino %lu pos %llu len %lu rw %d ret %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long long) __entry->pos, __entry->len, + __entry->rw, __entry->ret) +); + +TRACE_EVENT(ext3_unlink_enter, + TP_PROTO(struct inode *parent, struct dentry *dentry), + + TP_ARGS(parent, dentry), + + TP_STRUCT__entry( + __field( ino_t, parent ) + __field( ino_t, ino ) + __field( loff_t, size ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->parent = parent->i_ino; + __entry->ino = dentry->d_inode->i_ino; + __entry->size = dentry->d_inode->i_size; + __entry->dev = dentry->d_inode->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu size %lld parent %ld", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + (unsigned long long)__entry->size, + (unsigned long) __entry->parent) +); + +TRACE_EVENT(ext3_unlink_exit, + TP_PROTO(struct dentry *dentry, int ret), + + TP_ARGS(dentry, ret), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( int, ret ) + ), + + TP_fast_assign( + __entry->ino = dentry->d_inode->i_ino; + __entry->dev = dentry->d_inode->i_sb->s_dev; + __entry->ret = ret; + ), + + TP_printk("dev %d,%d ino %lu ret %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->ret) +); + +DECLARE_EVENT_CLASS(ext3__truncate, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( blkcnt_t, blocks ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + __entry->blocks = inode->i_blocks; + ), + + TP_printk("dev %d,%d ino %lu blocks %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, (unsigned long) __entry->blocks) +); + +DEFINE_EVENT(ext3__truncate, ext3_truncate_enter, + + TP_PROTO(struct inode *inode), + + TP_ARGS(inode) +); + +DEFINE_EVENT(ext3__truncate, ext3_truncate_exit, + + TP_PROTO(struct inode *inode), + + TP_ARGS(inode) +); + +TRACE_EVENT(ext3_get_blocks_enter, + TP_PROTO(struct inode *inode, unsigned long lblk, + unsigned long len, int create), + + TP_ARGS(inode, lblk, len, create), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( unsigned long, lblk ) + __field( unsigned long, len ) + __field( int, create ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + __entry->lblk = lblk; + __entry->len = len; + __entry->create = create; + ), + + TP_printk("dev %d,%d ino %lu lblk %lu len %lu create %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->lblk, __entry->len, __entry->create) +); + +TRACE_EVENT(ext3_get_blocks_exit, + TP_PROTO(struct inode *inode, unsigned long lblk, + unsigned long pblk, unsigned long len, int ret), + + TP_ARGS(inode, lblk, pblk, len, ret), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + __field( unsigned long, lblk ) + __field( unsigned long, pblk ) + __field( unsigned long, len ) + __field( int, ret ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + __entry->lblk = lblk; + __entry->pblk = pblk; + __entry->len = len; + __entry->ret = ret; + ), + + TP_printk("dev %d,%d ino %lu lblk %lu pblk %lu len %lu ret %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino, + __entry->lblk, __entry->pblk, + __entry->len, __entry->ret) +); + +TRACE_EVENT(ext3_load_inode, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( ino_t, ino ) + __field( dev_t, dev ) + ), + + TP_fast_assign( + __entry->ino = inode->i_ino; + __entry->dev = inode->i_sb->s_dev; + ), + + TP_printk("dev %d,%d ino %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + (unsigned long) __entry->ino) +); + +#endif /* _TRACE_EXT3_H */ + +/* This part must be outside protection */ +#include -- cgit v1.3-14-g43fede From 99cb1a318c37bf462c53d43f4dacb7b4896ce0c9 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Mon, 23 May 2011 18:33:02 +0200 Subject: jbd: Add fixed tracepoints This commit adds fixed tracepoint for jbd. It has been based on fixed tracepoints for jbd2, however there are missing those for collecting statistics, since I think that it will require more intrusive patch so I should have its own commit, if someone decide that it is needed. Also there are new tracepoints in __journal_drop_transaction() and journal_update_superblock(). The list of jbd tracepoints: jbd_checkpoint jbd_start_commit jbd_commit_locking jbd_commit_flushing jbd_commit_logging jbd_drop_transaction jbd_end_commit jbd_do_submit_data jbd_cleanup_journal_tail jbd_update_superblock_end Signed-off-by: Lukas Czerner Cc: Jan Kara Signed-off-by: Jan Kara --- fs/jbd/checkpoint.c | 4 + fs/jbd/commit.c | 11 +++ fs/jbd/journal.c | 4 + include/trace/events/jbd.h | 203 +++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 222 insertions(+) create mode 100644 include/trace/events/jbd.h (limited to 'include/trace') diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c index e4b87bc1fa56..dea7503b47e8 100644 --- a/fs/jbd/checkpoint.c +++ b/fs/jbd/checkpoint.c @@ -22,6 +22,7 @@ #include #include #include +#include /* * Unlink a buffer from a transaction checkpoint list. @@ -358,6 +359,7 @@ int log_do_checkpoint(journal_t *journal) * journal straight away. */ result = cleanup_journal_tail(journal); + trace_jbd_checkpoint(journal, result); jbd_debug(1, "cleanup_journal_tail returned %d\n", result); if (result <= 0) return result; @@ -503,6 +505,7 @@ int cleanup_journal_tail(journal_t *journal) if (blocknr < journal->j_tail) freed = freed + journal->j_last - journal->j_first; + trace_jbd_cleanup_journal_tail(journal, first_tid, blocknr, freed); jbd_debug(1, "Cleaning journal tail from %d to %d (offset %u), " "freeing %u\n", @@ -752,6 +755,7 @@ void __journal_drop_transaction(journal_t *journal, transaction_t *transaction) J_ASSERT(journal->j_committing_transaction != transaction); J_ASSERT(journal->j_running_transaction != transaction); + trace_jbd_drop_transaction(journal, transaction); jbd_debug(1, "Dropping transaction %d, all done\n", transaction->t_tid); kfree(transaction); } diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index 72ffa974b0b8..eedd201374a8 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -21,6 +21,7 @@ #include #include #include +#include /* * Default IO end handler for temporary BJ_IO buffer_heads. @@ -204,6 +205,8 @@ write_out_data: if (!trylock_buffer(bh)) { BUFFER_TRACE(bh, "needs blocking lock"); spin_unlock(&journal->j_list_lock); + trace_jbd_do_submit_data(journal, + commit_transaction); /* Write out all data to prevent deadlocks */ journal_do_submit_data(wbuf, bufs, write_op); bufs = 0; @@ -236,6 +239,8 @@ write_out_data: jbd_unlock_bh_state(bh); if (bufs == journal->j_wbufsize) { spin_unlock(&journal->j_list_lock); + trace_jbd_do_submit_data(journal, + commit_transaction); journal_do_submit_data(wbuf, bufs, write_op); bufs = 0; goto write_out_data; @@ -266,6 +271,7 @@ write_out_data: } } spin_unlock(&journal->j_list_lock); + trace_jbd_do_submit_data(journal, commit_transaction); journal_do_submit_data(wbuf, bufs, write_op); return err; @@ -316,12 +322,14 @@ void journal_commit_transaction(journal_t *journal) commit_transaction = journal->j_running_transaction; J_ASSERT(commit_transaction->t_state == T_RUNNING); + trace_jbd_start_commit(journal, commit_transaction); jbd_debug(1, "JBD: starting commit of transaction %d\n", commit_transaction->t_tid); spin_lock(&journal->j_state_lock); commit_transaction->t_state = T_LOCKED; + trace_jbd_commit_locking(journal, commit_transaction); spin_lock(&commit_transaction->t_handle_lock); while (commit_transaction->t_updates) { DEFINE_WAIT(wait); @@ -392,6 +400,7 @@ void journal_commit_transaction(journal_t *journal) */ journal_switch_revoke_table(journal); + trace_jbd_commit_flushing(journal, commit_transaction); commit_transaction->t_state = T_FLUSH; journal->j_committing_transaction = commit_transaction; journal->j_running_transaction = NULL; @@ -493,6 +502,7 @@ void journal_commit_transaction(journal_t *journal) commit_transaction->t_state = T_COMMIT; spin_unlock(&journal->j_state_lock); + trace_jbd_commit_logging(journal, commit_transaction); J_ASSERT(commit_transaction->t_nr_buffers <= commit_transaction->t_outstanding_credits); @@ -946,6 +956,7 @@ restart_loop: } spin_unlock(&journal->j_list_lock); + trace_jbd_end_commit(journal, commit_transaction); jbd_debug(1, "JBD: commit %d complete, head %d\n", journal->j_commit_sequence, journal->j_tail_sequence); diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index e2d4285fbe90..ab019ee77888 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -38,6 +38,9 @@ #include #include +#define CREATE_TRACE_POINTS +#include + #include #include @@ -1065,6 +1068,7 @@ void journal_update_superblock(journal_t *journal, int wait) } else write_dirty_buffer(bh, WRITE); + trace_jbd_update_superblock_end(journal, wait); out: /* If we have just flushed the log (by marking s_start==0), then * any future commit will have to be careful to update the diff --git a/include/trace/events/jbd.h b/include/trace/events/jbd.h new file mode 100644 index 000000000000..aff64d82d713 --- /dev/null +++ b/include/trace/events/jbd.h @@ -0,0 +1,203 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM jbd + +#if !defined(_TRACE_JBD_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_JBD_H + +#include +#include + +TRACE_EVENT(jbd_checkpoint, + + TP_PROTO(journal_t *journal, int result), + + TP_ARGS(journal, result), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, result ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->result = result; + ), + + TP_printk("dev %d,%d result %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->result) +); + +DECLARE_EVENT_CLASS(jbd_commit, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %d,%d transaction %d sync %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit) +); + +DEFINE_EVENT(jbd_commit, jbd_start_commit, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction) +); + +DEFINE_EVENT(jbd_commit, jbd_commit_locking, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction) +); + +DEFINE_EVENT(jbd_commit, jbd_commit_flushing, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction) +); + +DEFINE_EVENT(jbd_commit, jbd_commit_logging, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction) +); + +TRACE_EVENT(jbd_drop_transaction, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %d,%d transaction %d sync %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit) +); + +TRACE_EVENT(jbd_end_commit, + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + __field( int, head ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + __entry->head = journal->j_tail_sequence; + ), + + TP_printk("dev %d,%d transaction %d sync %d head %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit, __entry->head) +); + +TRACE_EVENT(jbd_do_submit_data, + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %d,%d transaction %d sync %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->transaction, __entry->sync_commit) +); + +TRACE_EVENT(jbd_cleanup_journal_tail, + + TP_PROTO(journal_t *journal, tid_t first_tid, + unsigned long block_nr, unsigned long freed), + + TP_ARGS(journal, first_tid, block_nr, freed), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( tid_t, tail_sequence ) + __field( tid_t, first_tid ) + __field(unsigned long, block_nr ) + __field(unsigned long, freed ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->tail_sequence = journal->j_tail_sequence; + __entry->first_tid = first_tid; + __entry->block_nr = block_nr; + __entry->freed = freed; + ), + + TP_printk("dev %d,%d from %u to %u offset %lu freed %lu", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->tail_sequence, __entry->first_tid, + __entry->block_nr, __entry->freed) +); + +TRACE_EVENT(jbd_update_superblock_end, + TP_PROTO(journal_t *journal, int wait), + + TP_ARGS(journal, wait), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, wait ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->wait = wait; + ), + + TP_printk("dev %d,%d wait %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->wait) +); + +#endif /* _TRACE_JBD_H */ + +/* This part must be outside protection */ +#include -- cgit v1.3-14-g43fede From d46db3d58233be4be980eb1e42eebe7808bcabab Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Wed, 4 May 2011 19:54:37 -0600 Subject: writeback: make writeback_control.nr_to_write straight Pass struct wb_writeback_work all the way down to writeback_sb_inodes(), and initialize the struct writeback_control there. struct writeback_control is basically designed to control writeback of a single file, but we keep abuse it for writing multiple files in writeback_sb_inodes() and its callers. It immediately clean things up, e.g. suddenly wbc.nr_to_write vs work->nr_pages starts to make sense, and instead of saving and restoring pages_skipped in writeback_sb_inodes it can always start with a clean zero value. It also makes a neat IO pattern change: large dirty files are now written in the full 4MB writeback chunk size, rather than whatever remained quota in wbc->nr_to_write. Acked-by: Jan Kara Proposed-by: Christoph Hellwig Signed-off-by: Wu Fengguang --- fs/btrfs/extent_io.c | 2 - fs/fs-writeback.c | 196 ++++++++++++++++++++++----------------- include/linux/writeback.h | 6 +- include/trace/events/writeback.h | 39 +++++--- mm/backing-dev.c | 17 +--- mm/page-writeback.c | 17 +--- 6 files changed, 148 insertions(+), 129 deletions(-) (limited to 'include/trace') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 7055d11c1efd..561262d35689 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -2551,7 +2551,6 @@ int extent_write_full_page(struct extent_io_tree *tree, struct page *page, }; struct writeback_control wbc_writepages = { .sync_mode = wbc->sync_mode, - .older_than_this = NULL, .nr_to_write = 64, .range_start = page_offset(page) + PAGE_CACHE_SIZE, .range_end = (loff_t)-1, @@ -2584,7 +2583,6 @@ int extent_write_locked_range(struct extent_io_tree *tree, struct inode *inode, }; struct writeback_control wbc_writepages = { .sync_mode = mode, - .older_than_this = NULL, .nr_to_write = nr_pages * 2, .range_start = start, .range_end = end + 1, diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 6caa98247a5b..2c947da39f6e 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -29,12 +29,22 @@ #include #include "internal.h" +/* + * The maximum number of pages to writeout in a single bdi flush/kupdate + * operation. We do this so we don't hold I_SYNC against an inode for + * enormous amounts of time, which would block a userspace task which has + * been forced to throttle against that inode. Also, the code reevaluates + * the dirty each time it has written this many pages. + */ +#define MAX_WRITEBACK_PAGES 1024L + /* * Passed into wb_writeback(), essentially a subset of writeback_control */ struct wb_writeback_work { long nr_pages; struct super_block *sb; + unsigned long *older_than_this; enum writeback_sync_modes sync_mode; unsigned int tagged_writepages:1; unsigned int for_kupdate:1; @@ -472,7 +482,6 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, * No need to add it back to the LRU. */ list_del_init(&inode->i_wb_list); - wbc->inodes_written++; } } inode_sync_complete(inode); @@ -506,6 +515,31 @@ static bool pin_sb_for_writeback(struct super_block *sb) return false; } +static long writeback_chunk_size(struct wb_writeback_work *work) +{ + long pages; + + /* + * WB_SYNC_ALL mode does livelock avoidance by syncing dirty + * inodes/pages in one big loop. Setting wbc.nr_to_write=LONG_MAX + * here avoids calling into writeback_inodes_wb() more than once. + * + * The intended call sequence for WB_SYNC_ALL writeback is: + * + * wb_writeback() + * writeback_sb_inodes() <== called only once + * write_cache_pages() <== called once for each inode + * (quickly) tag currently dirty pages + * (maybe slowly) sync all tagged pages + */ + if (work->sync_mode == WB_SYNC_ALL || work->tagged_writepages) + pages = LONG_MAX; + else + pages = min(MAX_WRITEBACK_PAGES, work->nr_pages); + + return pages; +} + /* * Write a portion of b_io inodes which belong to @sb. * @@ -513,18 +547,30 @@ static bool pin_sb_for_writeback(struct super_block *sb) * inodes. Otherwise write only ones which go sequentially * in reverse order. * - * Return 1, if the caller writeback routine should be - * interrupted. Otherwise return 0. + * Return the number of pages and/or inodes written. */ -static int writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, - struct writeback_control *wbc, bool only_this_sb) +static long writeback_sb_inodes(struct super_block *sb, + struct bdi_writeback *wb, + struct wb_writeback_work *work) { + struct writeback_control wbc = { + .sync_mode = work->sync_mode, + .tagged_writepages = work->tagged_writepages, + .for_kupdate = work->for_kupdate, + .for_background = work->for_background, + .range_cyclic = work->range_cyclic, + .range_start = 0, + .range_end = LLONG_MAX, + }; + unsigned long start_time = jiffies; + long write_chunk; + long wrote = 0; /* count both pages and inodes */ + while (!list_empty(&wb->b_io)) { - long pages_skipped; struct inode *inode = wb_inode(wb->b_io.prev); if (inode->i_sb != sb) { - if (only_this_sb) { + if (work->sb) { /* * We only want to write back data for this * superblock, move all inodes not belonging @@ -539,7 +585,7 @@ static int writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, * Bounce back to the caller to unpin this and * pin the next superblock. */ - return 0; + break; } /* @@ -553,12 +599,18 @@ static int writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, requeue_io(inode, wb); continue; } - __iget(inode); + write_chunk = writeback_chunk_size(work); + wbc.nr_to_write = write_chunk; + wbc.pages_skipped = 0; + + writeback_single_inode(inode, wb, &wbc); - pages_skipped = wbc->pages_skipped; - writeback_single_inode(inode, wb, wbc); - if (wbc->pages_skipped != pages_skipped) { + work->nr_pages -= write_chunk - wbc.nr_to_write; + wrote += write_chunk - wbc.nr_to_write; + if (!(inode->i_state & I_DIRTY)) + wrote++; + if (wbc.pages_skipped) { /* * writeback is not making progress due to locked * buffers. Skip this inode for now. @@ -570,17 +622,25 @@ static int writeback_sb_inodes(struct super_block *sb, struct bdi_writeback *wb, iput(inode); cond_resched(); spin_lock(&wb->list_lock); - if (wbc->nr_to_write <= 0) - return 1; + /* + * bail out to wb_writeback() often enough to check + * background threshold and other termination conditions. + */ + if (wrote) { + if (time_is_before_jiffies(start_time + HZ / 10UL)) + break; + if (work->nr_pages <= 0) + break; + } } - /* b_io is empty */ - return 1; + return wrote; } -static void __writeback_inodes_wb(struct bdi_writeback *wb, - struct writeback_control *wbc) +static long __writeback_inodes_wb(struct bdi_writeback *wb, + struct wb_writeback_work *work) { - int ret = 0; + unsigned long start_time = jiffies; + long wrote = 0; while (!list_empty(&wb->b_io)) { struct inode *inode = wb_inode(wb->b_io.prev); @@ -590,33 +650,37 @@ static void __writeback_inodes_wb(struct bdi_writeback *wb, requeue_io(inode, wb); continue; } - ret = writeback_sb_inodes(sb, wb, wbc, false); + wrote += writeback_sb_inodes(sb, wb, work); drop_super(sb); - if (ret) - break; + /* refer to the same tests at the end of writeback_sb_inodes */ + if (wrote) { + if (time_is_before_jiffies(start_time + HZ / 10UL)) + break; + if (work->nr_pages <= 0) + break; + } } /* Leave any unwritten inodes on b_io */ + return wrote; } -void writeback_inodes_wb(struct bdi_writeback *wb, - struct writeback_control *wbc) +long writeback_inodes_wb(struct bdi_writeback *wb, long nr_pages) { + struct wb_writeback_work work = { + .nr_pages = nr_pages, + .sync_mode = WB_SYNC_NONE, + .range_cyclic = 1, + }; + spin_lock(&wb->list_lock); if (list_empty(&wb->b_io)) - queue_io(wb, wbc->older_than_this); - __writeback_inodes_wb(wb, wbc); + queue_io(wb, NULL); + __writeback_inodes_wb(wb, &work); spin_unlock(&wb->list_lock); -} -/* - * The maximum number of pages to writeout in a single bdi flush/kupdate - * operation. We do this so we don't hold I_SYNC against an inode for - * enormous amounts of time, which would block a userspace task which has - * been forced to throttle against that inode. Also, the code reevaluates - * the dirty each time it has written this many pages. - */ -#define MAX_WRITEBACK_PAGES 1024 + return nr_pages - work.nr_pages; +} static inline bool over_bground_thresh(void) { @@ -646,42 +710,13 @@ static inline bool over_bground_thresh(void) static long wb_writeback(struct bdi_writeback *wb, struct wb_writeback_work *work) { - struct writeback_control wbc = { - .sync_mode = work->sync_mode, - .tagged_writepages = work->tagged_writepages, - .older_than_this = NULL, - .for_kupdate = work->for_kupdate, - .for_background = work->for_background, - .range_cyclic = work->range_cyclic, - }; + long nr_pages = work->nr_pages; unsigned long oldest_jif; - long wrote = 0; - long write_chunk = MAX_WRITEBACK_PAGES; struct inode *inode; - - if (!wbc.range_cyclic) { - wbc.range_start = 0; - wbc.range_end = LLONG_MAX; - } - - /* - * WB_SYNC_ALL mode does livelock avoidance by syncing dirty - * inodes/pages in one big loop. Setting wbc.nr_to_write=LONG_MAX - * here avoids calling into writeback_inodes_wb() more than once. - * - * The intended call sequence for WB_SYNC_ALL writeback is: - * - * wb_writeback() - * writeback_sb_inodes() <== called only once - * write_cache_pages() <== called once for each inode - * (quickly) tag currently dirty pages - * (maybe slowly) sync all tagged pages - */ - if (wbc.sync_mode == WB_SYNC_ALL || wbc.tagged_writepages) - write_chunk = LONG_MAX; + long progress; oldest_jif = jiffies; - wbc.older_than_this = &oldest_jif; + work->older_than_this = &oldest_jif; spin_lock(&wb->list_lock); for (;;) { @@ -711,24 +746,17 @@ static long wb_writeback(struct bdi_writeback *wb, if (work->for_kupdate) { oldest_jif = jiffies - msecs_to_jiffies(dirty_expire_interval * 10); - wbc.older_than_this = &oldest_jif; + work->older_than_this = &oldest_jif; } - wbc.nr_to_write = write_chunk; - wbc.pages_skipped = 0; - wbc.inodes_written = 0; - - trace_wbc_writeback_start(&wbc, wb->bdi); + trace_writeback_start(wb->bdi, work); if (list_empty(&wb->b_io)) - queue_io(wb, wbc.older_than_this); + queue_io(wb, work->older_than_this); if (work->sb) - writeback_sb_inodes(work->sb, wb, &wbc, true); + progress = writeback_sb_inodes(work->sb, wb, work); else - __writeback_inodes_wb(wb, &wbc); - trace_wbc_writeback_written(&wbc, wb->bdi); - - work->nr_pages -= write_chunk - wbc.nr_to_write; - wrote += write_chunk - wbc.nr_to_write; + progress = __writeback_inodes_wb(wb, work); + trace_writeback_written(wb->bdi, work); /* * Did we write something? Try for more @@ -738,9 +766,7 @@ static long wb_writeback(struct bdi_writeback *wb, * mean the overall work is done. So we keep looping as long * as made some progress on cleaning pages or inodes. */ - if (wbc.nr_to_write < write_chunk) - continue; - if (wbc.inodes_written) + if (progress) continue; /* * No more inodes for IO, bail @@ -753,8 +779,8 @@ static long wb_writeback(struct bdi_writeback *wb, * we'll just busyloop. */ if (!list_empty(&wb->b_more_io)) { + trace_writeback_wait(wb->bdi, work); inode = wb_inode(wb->b_more_io.prev); - trace_wbc_writeback_wait(&wbc, wb->bdi); spin_lock(&inode->i_lock); inode_wait_for_writeback(inode, wb); spin_unlock(&inode->i_lock); @@ -762,7 +788,7 @@ static long wb_writeback(struct bdi_writeback *wb, } spin_unlock(&wb->list_lock); - return wrote; + return nr_pages - work->nr_pages; } /* diff --git a/include/linux/writeback.h b/include/linux/writeback.h index 2f1b512bd6e0..df1b7f18f100 100644 --- a/include/linux/writeback.h +++ b/include/linux/writeback.h @@ -24,12 +24,9 @@ enum writeback_sync_modes { */ struct writeback_control { enum writeback_sync_modes sync_mode; - unsigned long *older_than_this; /* If !NULL, only write back inodes - older than this */ long nr_to_write; /* Write this many pages, and decrement this for each page written */ long pages_skipped; /* Pages which were not written */ - long inodes_written; /* # of inodes written (at least) */ /* * For a_ops->writepages(): is start or end are non-zero then this is @@ -56,8 +53,7 @@ void writeback_inodes_sb_nr(struct super_block *, unsigned long nr); int writeback_inodes_sb_if_idle(struct super_block *); int writeback_inodes_sb_nr_if_idle(struct super_block *, unsigned long nr); void sync_inodes_sb(struct super_block *); -void writeback_inodes_wb(struct bdi_writeback *wb, - struct writeback_control *wbc); +long writeback_inodes_wb(struct bdi_writeback *wb, long nr_pages); long wb_do_writeback(struct bdi_writeback *wb, int force_wait); void wakeup_flusher_threads(long nr_pages); diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index 205d14919ef2..3e7662a0cfa3 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -62,6 +62,9 @@ DEFINE_EVENT(writeback_work_class, name, \ DEFINE_WRITEBACK_WORK_EVENT(writeback_nothread); DEFINE_WRITEBACK_WORK_EVENT(writeback_queue); DEFINE_WRITEBACK_WORK_EVENT(writeback_exec); +DEFINE_WRITEBACK_WORK_EVENT(writeback_start); +DEFINE_WRITEBACK_WORK_EVENT(writeback_written); +DEFINE_WRITEBACK_WORK_EVENT(writeback_wait); TRACE_EVENT(writeback_pages_written, TP_PROTO(long pages_written), @@ -101,6 +104,30 @@ DEFINE_WRITEBACK_EVENT(writeback_bdi_register); DEFINE_WRITEBACK_EVENT(writeback_bdi_unregister); DEFINE_WRITEBACK_EVENT(writeback_thread_start); DEFINE_WRITEBACK_EVENT(writeback_thread_stop); +DEFINE_WRITEBACK_EVENT(balance_dirty_start); +DEFINE_WRITEBACK_EVENT(balance_dirty_wait); + +TRACE_EVENT(balance_dirty_written, + + TP_PROTO(struct backing_dev_info *bdi, int written), + + TP_ARGS(bdi, written), + + TP_STRUCT__entry( + __array(char, name, 32) + __field(int, written) + ), + + TP_fast_assign( + strncpy(__entry->name, dev_name(bdi->dev), 32); + __entry->written = written; + ), + + TP_printk("bdi %s written %d", + __entry->name, + __entry->written + ) +); DECLARE_EVENT_CLASS(wbc_class, TP_PROTO(struct writeback_control *wbc, struct backing_dev_info *bdi), @@ -114,7 +141,6 @@ DECLARE_EVENT_CLASS(wbc_class, __field(int, for_background) __field(int, for_reclaim) __field(int, range_cyclic) - __field(unsigned long, older_than_this) __field(long, range_start) __field(long, range_end) ), @@ -128,14 +154,12 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->for_background = wbc->for_background; __entry->for_reclaim = wbc->for_reclaim; __entry->range_cyclic = wbc->range_cyclic; - __entry->older_than_this = wbc->older_than_this ? - *wbc->older_than_this : 0; __entry->range_start = (long)wbc->range_start; __entry->range_end = (long)wbc->range_end; ), TP_printk("bdi %s: towrt=%ld skip=%ld mode=%d kupd=%d " - "bgrd=%d reclm=%d cyclic=%d older=0x%lx " + "bgrd=%d reclm=%d cyclic=%d " "start=0x%lx end=0x%lx", __entry->name, __entry->nr_to_write, @@ -145,7 +169,6 @@ DECLARE_EVENT_CLASS(wbc_class, __entry->for_background, __entry->for_reclaim, __entry->range_cyclic, - __entry->older_than_this, __entry->range_start, __entry->range_end) ) @@ -154,12 +177,6 @@ DECLARE_EVENT_CLASS(wbc_class, DEFINE_EVENT(wbc_class, name, \ TP_PROTO(struct writeback_control *wbc, struct backing_dev_info *bdi), \ TP_ARGS(wbc, bdi)) -DEFINE_WBC_EVENT(wbc_writeback_start); -DEFINE_WBC_EVENT(wbc_writeback_written); -DEFINE_WBC_EVENT(wbc_writeback_wait); -DEFINE_WBC_EVENT(wbc_balance_dirty_start); -DEFINE_WBC_EVENT(wbc_balance_dirty_written); -DEFINE_WBC_EVENT(wbc_balance_dirty_wait); DEFINE_WBC_EVENT(wbc_writepage); TRACE_EVENT(writeback_queue_io, diff --git a/mm/backing-dev.c b/mm/backing-dev.c index 5f6553ef1ba7..7ba303be5e03 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -260,18 +260,6 @@ int bdi_has_dirty_io(struct backing_dev_info *bdi) return wb_has_dirty_io(&bdi->wb); } -static void bdi_flush_io(struct backing_dev_info *bdi) -{ - struct writeback_control wbc = { - .sync_mode = WB_SYNC_NONE, - .older_than_this = NULL, - .range_cyclic = 1, - .nr_to_write = 1024, - }; - - writeback_inodes_wb(&bdi->wb, &wbc); -} - /* * kupdated() used to do this. We cannot do it from the bdi_forker_thread() * or we risk deadlocking on ->s_umount. The longer term solution would be @@ -457,9 +445,10 @@ static int bdi_forker_thread(void *ptr) if (IS_ERR(task)) { /* * If thread creation fails, force writeout of - * the bdi from the thread. + * the bdi from the thread. Hopefully 1024 is + * large enough for efficient IO. */ - bdi_flush_io(bdi); + writeback_inodes_wb(&bdi->wb, 1024); } else { /* * The spinlock makes sure we do not lose diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 1965d05a29cc..9d6ac2b6d942 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -491,13 +491,6 @@ static void balance_dirty_pages(struct address_space *mapping, struct backing_dev_info *bdi = mapping->backing_dev_info; for (;;) { - struct writeback_control wbc = { - .sync_mode = WB_SYNC_NONE, - .older_than_this = NULL, - .nr_to_write = write_chunk, - .range_cyclic = 1, - }; - nr_reclaimable = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); nr_writeback = global_page_state(NR_WRITEBACK); @@ -559,17 +552,17 @@ static void balance_dirty_pages(struct address_space *mapping, * threshold otherwise wait until the disk writes catch * up. */ - trace_wbc_balance_dirty_start(&wbc, bdi); + trace_balance_dirty_start(bdi); if (bdi_nr_reclaimable > bdi_thresh) { - writeback_inodes_wb(&bdi->wb, &wbc); - pages_written += write_chunk - wbc.nr_to_write; - trace_wbc_balance_dirty_written(&wbc, bdi); + pages_written += writeback_inodes_wb(&bdi->wb, + write_chunk); + trace_balance_dirty_written(bdi, pages_written); if (pages_written >= write_chunk) break; /* We've done our duty */ } - trace_wbc_balance_dirty_wait(&wbc, bdi); __set_current_state(TASK_UNINTERRUPTIBLE); io_schedule_timeout(pause); + trace_balance_dirty_wait(bdi); /* * Increase the delay for each loop, up to our previous -- cgit v1.3-14-g43fede From e1cbe236013c82bcf9a156e98d7b47efb89d2674 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Mon, 6 Dec 2010 22:34:29 -0600 Subject: writeback: trace global_dirty_state Add trace event balance_dirty_state for showing the global dirty page counts and thresholds at each global_dirty_limits() invocation. This will cover the callers throttle_vm_writeout(), over_bground_thresh() and each balance_dirty_pages() loop. Signed-off-by: Wu Fengguang --- include/trace/events/writeback.h | 46 ++++++++++++++++++++++++++++++++++++++++ mm/page-writeback.c | 1 + 2 files changed, 47 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h index 3e7662a0cfa3..6bca4cc0063c 100644 --- a/include/trace/events/writeback.h +++ b/include/trace/events/writeback.h @@ -204,6 +204,52 @@ TRACE_EVENT(writeback_queue_io, __entry->moved) ); +TRACE_EVENT(global_dirty_state, + + TP_PROTO(unsigned long background_thresh, + unsigned long dirty_thresh + ), + + TP_ARGS(background_thresh, + dirty_thresh + ), + + TP_STRUCT__entry( + __field(unsigned long, nr_dirty) + __field(unsigned long, nr_writeback) + __field(unsigned long, nr_unstable) + __field(unsigned long, background_thresh) + __field(unsigned long, dirty_thresh) + __field(unsigned long, dirty_limit) + __field(unsigned long, nr_dirtied) + __field(unsigned long, nr_written) + ), + + TP_fast_assign( + __entry->nr_dirty = global_page_state(NR_FILE_DIRTY); + __entry->nr_writeback = global_page_state(NR_WRITEBACK); + __entry->nr_unstable = global_page_state(NR_UNSTABLE_NFS); + __entry->nr_dirtied = global_page_state(NR_DIRTIED); + __entry->nr_written = global_page_state(NR_WRITTEN); + __entry->background_thresh = background_thresh; + __entry->dirty_thresh = dirty_thresh; + __entry->dirty_limit = global_dirty_limit; + ), + + TP_printk("dirty=%lu writeback=%lu unstable=%lu " + "bg_thresh=%lu thresh=%lu limit=%lu " + "dirtied=%lu written=%lu", + __entry->nr_dirty, + __entry->nr_writeback, + __entry->nr_unstable, + __entry->background_thresh, + __entry->dirty_thresh, + __entry->dirty_limit, + __entry->nr_dirtied, + __entry->nr_written + ) +); + DECLARE_EVENT_CLASS(writeback_congest_waited_template, TP_PROTO(unsigned int usec_timeout, unsigned int usec_delayed), diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 798842a22474..f9d9f5476d58 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -447,6 +447,7 @@ void global_dirty_limits(unsigned long *pbackground, unsigned long *pdirty) } *pbackground = background; *pdirty = dirty; + trace_global_dirty_state(background, dirty); } /** -- cgit v1.3-14-g43fede