X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=fs%2Ffs-writeback.c;h=15944f754e1512a21e614d0e47fe6756584f10ea;hb=5b0830cb9085f4b69f9d57d7f3aaff322ffbec26;hp=f8cd7a97f5b7671c20dcbfc26d992eba831b0134;hpb=cfc4ba5365449cb6b5c9f68d755a142f17da1e47;p=safe%2Fjmp%2Flinux-2.6 diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index f8cd7a9..15944f7 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -41,23 +41,24 @@ struct wb_writeback_args { long nr_pages; struct super_block *sb; enum writeback_sync_modes sync_mode; - int for_kupdate; - int range_cyclic; + int for_kupdate:1; + int range_cyclic:1; + int for_background:1; }; /* * Work items for the bdi_writeback threads */ struct bdi_work { - struct list_head list; - struct rcu_head rcu_head; + struct list_head list; /* pending work list */ + struct rcu_head rcu_head; /* for RCU free/clear of work */ - unsigned long seen; - atomic_t pending; + unsigned long seen; /* threads that have seen this work */ + atomic_t pending; /* number of threads still to do work */ - struct wb_writeback_args args; + struct wb_writeback_args args; /* writeback arguments */ - unsigned long state; + unsigned long state; /* flag bits, see WS_* */ }; enum { @@ -74,14 +75,10 @@ static inline bool bdi_work_on_stack(struct bdi_work *work) } static inline void bdi_work_init(struct bdi_work *work, - struct writeback_control *wbc) + struct wb_writeback_args *args) { INIT_RCU_HEAD(&work->rcu_head); - work->args.sb = wbc->sb; - work->args.nr_pages = wbc->nr_to_write; - work->args.sync_mode = wbc->sync_mode; - work->args.range_cyclic = wbc->range_cyclic; - work->args.for_kupdate = 0; + work->args = *args; work->state = WS_USED; } @@ -101,6 +98,11 @@ static void bdi_work_clear(struct bdi_work *work) { clear_bit(WS_USED_B, &work->state); smp_mb__after_clear_bit(); + /* + * work can have disappeared at this point. bit waitq functions + * should be able to tolerate this, provided bdi_sched_wait does + * not dereference it's pointer argument. + */ wake_up_bit(&work->state, WS_USED_B); } @@ -117,6 +119,7 @@ static void bdi_work_free(struct rcu_head *head) static void wb_work_complete(struct bdi_work *work) { const enum writeback_sync_modes sync_mode = work->args.sync_mode; + int onstack = bdi_work_on_stack(work); /* * For allocated work, we can clear the done/seen bit right here. @@ -124,9 +127,9 @@ static void wb_work_complete(struct bdi_work *work) * to after the RCU grace period, since the stack could be invalidated * as soon as bdi_work_clear() has done the wakeup. */ - if (!bdi_work_on_stack(work)) + if (!onstack) bdi_work_clear(work); - if (sync_mode == WB_SYNC_NONE || bdi_work_on_stack(work)) + if (sync_mode == WB_SYNC_NONE || onstack) call_rcu(&work->rcu_head, bdi_work_free); } @@ -149,21 +152,19 @@ static void wb_clear_pending(struct bdi_writeback *wb, struct bdi_work *work) static void bdi_queue_work(struct backing_dev_info *bdi, struct bdi_work *work) { - if (work) { - work->seen = bdi->wb_mask; - BUG_ON(!work->seen); - atomic_set(&work->pending, bdi->wb_cnt); - BUG_ON(!bdi->wb_cnt); - - /* - * Make sure stores are seen before it appears on the list - */ - smp_mb(); + work->seen = bdi->wb_mask; + BUG_ON(!work->seen); + atomic_set(&work->pending, bdi->wb_cnt); + BUG_ON(!bdi->wb_cnt); - spin_lock(&bdi->wb_lock); - list_add_tail_rcu(&work->list, &bdi->work_list); - spin_unlock(&bdi->wb_lock); - } + /* + * list_add_tail_rcu() contains the necessary barriers to + * make sure the above stores are seen before the item is + * noticed on the list + */ + spin_lock(&bdi->wb_lock); + list_add_tail_rcu(&work->list, &bdi->work_list); + spin_unlock(&bdi->wb_lock); /* * If the default thread isn't there, make sure we add it. When @@ -174,15 +175,7 @@ static void bdi_queue_work(struct backing_dev_info *bdi, struct bdi_work *work) else { struct bdi_writeback *wb = &bdi->wb; - /* - * If we failed allocating the bdi work item, wake up the wb - * thread always. As a safety precaution, it'll flush out - * everything - */ - if (!wb_has_dirty_io(wb)) { - if (work) - wb_clear_pending(wb, work); - } else if (wb->task) + if (wb->task) wake_up_process(wb->task); } } @@ -198,35 +191,83 @@ static void bdi_wait_on_work_clear(struct bdi_work *work) } static void bdi_alloc_queue_work(struct backing_dev_info *bdi, - struct writeback_control *wbc) + struct wb_writeback_args *args) { struct bdi_work *work; + /* + * This is WB_SYNC_NONE writeback, so if allocation fails just + * wakeup the thread for old dirty data writeback + */ work = kmalloc(sizeof(*work), GFP_ATOMIC); - if (work) - bdi_work_init(work, wbc); + if (work) { + bdi_work_init(work, args); + bdi_queue_work(bdi, work); + } else { + struct bdi_writeback *wb = &bdi->wb; + + if (wb->task) + wake_up_process(wb->task); + } +} - bdi_queue_work(bdi, work); +/** + * bdi_sync_writeback - start and wait for writeback + * @bdi: the backing device to write from + * @sb: write inodes from this super_block + * + * Description: + * This does WB_SYNC_ALL data integrity writeback and waits for the + * IO to complete. Callers must hold the sb s_umount semaphore for + * reading, to avoid having the super disappear before we are done. + */ +static void bdi_sync_writeback(struct backing_dev_info *bdi, + struct super_block *sb) +{ + struct wb_writeback_args args = { + .sb = sb, + .sync_mode = WB_SYNC_ALL, + .nr_pages = LONG_MAX, + .range_cyclic = 0, + }; + struct bdi_work work; + + bdi_work_init(&work, &args); + work.state |= WS_ONSTACK; + + bdi_queue_work(bdi, &work); + bdi_wait_on_work_clear(&work); } -void bdi_start_writeback(struct writeback_control *wbc) +/** + * bdi_start_writeback - start writeback + * @bdi: the backing device to write from + * @nr_pages: the number of pages to write + * + * Description: + * This does WB_SYNC_NONE opportunistic writeback. The IO is only + * started when this function returns, we make no guarentees on + * completion. Caller need not hold sb s_umount semaphore. + * + */ +void bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages) { + struct wb_writeback_args args = { + .sync_mode = WB_SYNC_NONE, + .nr_pages = nr_pages, + .range_cyclic = 1, + }; + /* - * WB_SYNC_NONE is opportunistic writeback. If this allocation fails, - * bdi_queue_work() will wake up the thread and flush old data. This - * should ensure some amount of progress in freeing memory. + * We treat @nr_pages=0 as the special case to do background writeback, + * ie. to sync pages until the background dirty threshold is reached. */ - if (wbc->sync_mode != WB_SYNC_ALL) - bdi_alloc_queue_work(wbc->bdi, wbc); - else { - struct bdi_work work; - - bdi_work_init(&work, wbc); - work.state |= WS_ONSTACK; - - bdi_queue_work(wbc->bdi, &work); - bdi_wait_on_work_clear(&work); + if (!nr_pages) { + args.nr_pages = LONG_MAX; + args.for_background = 1; } + + bdi_alloc_queue_work(bdi, &args); } /* @@ -279,7 +320,7 @@ static bool inode_dirtied_after(struct inode *inode, unsigned long t) * For inodes being constantly redirtied, dirtied_when can get stuck. * It _appears_ to be in the future, but is actually in distant past. * This test is necessary to prevent such wrapped-around relative times - * from permanently stopping the whole pdflush writeback. + * from permanently stopping the whole bdi writeback. */ ret = ret && time_before_eq(inode->dirtied_when, jiffies); #endif @@ -408,8 +449,13 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc) spin_lock(&inode_lock); inode->i_state &= ~I_SYNC; if (!(inode->i_state & (I_FREEING | I_CLEAR))) { - if (!(inode->i_state & I_DIRTY) && - mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { + if (inode->i_state & I_DIRTY) { + /* + * Someone redirtied the inode while were writing back + * the pages. + */ + redirty_tail(inode); + } else if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { /* * We didn't write back all the pages. nfs_writepages() * sometimes bales out without doing anything. Redirty @@ -453,12 +499,6 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc) inode->i_state |= I_DIRTY_PAGES; redirty_tail(inode); } - } else if (inode->i_state & I_DIRTY) { - /* - * Someone redirtied the inode while were writing back - * the pages. - */ - redirty_tail(inode); } else if (atomic_read(&inode->i_count)) { /* * The inode is clean, inuse @@ -675,6 +715,7 @@ static long wb_writeback(struct bdi_writeback *wb, }; unsigned long oldest_jif; long wrote = 0; + struct inode *inode; if (wbc.for_kupdate) { wbc.older_than_this = &oldest_jif; @@ -688,20 +729,16 @@ static long wb_writeback(struct bdi_writeback *wb, for (;;) { /* - * Don't flush anything for non-integrity writeback where - * no nr_pages was given + * Stop writeback when nr_pages has been consumed */ - if (!args->for_kupdate && args->nr_pages <= 0 && - args->sync_mode == WB_SYNC_NONE) + if (args->nr_pages <= 0) break; /* - * If no specific pages were given and this is just a - * periodic background writeout and we are below the - * background dirty threshold, don't do anything + * For background writeout, stop when we are below the + * background dirty threshold */ - if (args->for_kupdate && args->nr_pages <= 0 && - !over_bground_thresh()) + if (args->for_background && !over_bground_thresh()) break; wbc.more_io = 0; @@ -713,13 +750,32 @@ static long wb_writeback(struct bdi_writeback *wb, wrote += MAX_WRITEBACK_PAGES - wbc.nr_to_write; /* - * If we ran out of stuff to write, bail unless more_io got set + * If we consumed everything, see if we have more */ - if (wbc.nr_to_write > 0 || wbc.pages_skipped > 0) { - if (wbc.more_io && !wbc.for_kupdate) - continue; + if (wbc.nr_to_write <= 0) + continue; + /* + * Didn't write everything and we don't have more IO, bail + */ + if (!wbc.more_io) break; + /* + * Did we write something? Try for more + */ + if (wbc.nr_to_write < MAX_WRITEBACK_PAGES) + continue; + /* + * Nothing written. Wait for some inode to + * become available for writeback. Otherwise + * we'll just busyloop. + */ + spin_lock(&inode_lock); + if (!list_empty(&wb->b_more_io)) { + inode = list_entry(wb->b_more_io.prev, + struct inode, i_list); + inode_wait_for_writeback(inode); } + spin_unlock(&inode_lock); } return wrote; @@ -727,7 +783,11 @@ static long wb_writeback(struct bdi_writeback *wb, /* * Return the next bdi_work struct that hasn't been processed by this - * wb thread yet + * wb thread yet. ->seen is initially set for each thread that exists + * for this device, when a thread first notices a piece of work it + * clears its bit. Depending on writeback type, the thread will notify + * completion on either receiving the work (WB_SYNC_NONE) or after + * it is done (WB_SYNC_ALL). */ static struct bdi_work *get_next_work_item(struct backing_dev_info *bdi, struct bdi_writeback *wb) @@ -737,8 +797,9 @@ static struct bdi_work *get_next_work_item(struct backing_dev_info *bdi, rcu_read_lock(); list_for_each_entry_rcu(work, &bdi->work_list, list) { - if (!test_and_clear_bit(wb->nr, &work->seen)) + if (!test_bit(wb->nr, &work->seen)) continue; + clear_bit(wb->nr, &work->seen); ret = work; break; @@ -849,8 +910,7 @@ int bdi_writeback_task(struct bdi_writeback *wb) } wait_jiffies = msecs_to_jiffies(dirty_writeback_interval * 10); - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(wait_jiffies); + schedule_timeout_interruptible(wait_jiffies); try_to_freeze(); } @@ -858,23 +918,25 @@ int bdi_writeback_task(struct bdi_writeback *wb) } /* - * Schedule writeback for all backing devices. Can only be used for - * WB_SYNC_NONE writeback, WB_SYNC_ALL should use bdi_start_writeback() - * and pass in the superblock. + * Schedule writeback for all backing devices. This does WB_SYNC_NONE + * writeback, for integrity writeback see bdi_sync_writeback(). */ -static void bdi_writeback_all(struct writeback_control *wbc) +static void bdi_writeback_all(struct super_block *sb, long nr_pages) { + struct wb_writeback_args args = { + .sb = sb, + .nr_pages = nr_pages, + .sync_mode = WB_SYNC_NONE, + }; struct backing_dev_info *bdi; - WARN_ON(wbc->sync_mode == WB_SYNC_ALL); - rcu_read_lock(); list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { if (!bdi_has_dirty_io(bdi)) continue; - bdi_alloc_queue_work(bdi, wbc); + bdi_alloc_queue_work(bdi, &args); } rcu_read_unlock(); @@ -886,17 +948,10 @@ static void bdi_writeback_all(struct writeback_control *wbc) */ void wakeup_flusher_threads(long nr_pages) { - struct writeback_control wbc = { - .sync_mode = WB_SYNC_NONE, - .older_than_this = NULL, - .range_cyclic = 1, - }; - if (nr_pages == 0) nr_pages = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); - wbc.nr_to_write = nr_pages; - bdi_writeback_all(&wbc); + bdi_writeback_all(NULL, nr_pages); } static noinline void block_dump___mark_inode_dirty(struct inode *inode) @@ -1030,9 +1085,6 @@ EXPORT_SYMBOL(__mark_inode_dirty); * If older_than_this is non-NULL, then only write out inodes which * had their first dirtying at a time earlier than *older_than_this. * - * If we're a pdlfush thread, then implement pdflush collision avoidance - * against the entire list. - * * If `bdi' is non-zero then we're being asked to writeback a specific queue. * This function assumes that the blockdev superblock's inodes are backed by * a variety of queues, so all inodes are searched. For other superblocks, @@ -1043,7 +1095,7 @@ EXPORT_SYMBOL(__mark_inode_dirty); * on the writer throttling path, and we get decent balancing between many * throttled threads: we don't want them all piling up on inode_sync_wait. */ -static void wait_sb_inodes(struct writeback_control *wbc) +static void wait_sb_inodes(struct super_block *sb) { struct inode *inode, *old_inode = NULL; @@ -1051,7 +1103,7 @@ static void wait_sb_inodes(struct writeback_control *wbc) * We need to be protected against the filesystem going from * r/o to r/w or vice versa. */ - WARN_ON(!rwsem_is_locked(&wbc->sb->s_umount)); + WARN_ON(!rwsem_is_locked(&sb->s_umount)); spin_lock(&inode_lock); @@ -1062,7 +1114,7 @@ static void wait_sb_inodes(struct writeback_control *wbc) * In which case, the inode may not be on the dirty list, but * we still have to wait for that writeout. */ - list_for_each_entry(inode, &wbc->sb->s_inodes, i_sb_list) { + list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { struct address_space *mapping; if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE|I_NEW)) @@ -1102,14 +1154,8 @@ static void wait_sb_inodes(struct writeback_control *wbc) * for IO completion of submitted IO. The number of pages submitted is * returned. */ -long writeback_inodes_sb(struct super_block *sb) +void writeback_inodes_sb(struct super_block *sb) { - struct writeback_control wbc = { - .sb = sb, - .sync_mode = WB_SYNC_NONE, - .range_start = 0, - .range_end = LLONG_MAX, - }; unsigned long nr_dirty = global_page_state(NR_FILE_DIRTY); unsigned long nr_unstable = global_page_state(NR_UNSTABLE_NFS); long nr_to_write; @@ -1117,9 +1163,7 @@ long writeback_inodes_sb(struct super_block *sb) nr_to_write = nr_dirty + nr_unstable + (inodes_stat.nr_inodes - inodes_stat.nr_unused); - wbc.nr_to_write = nr_to_write; - bdi_writeback_all(&wbc); - return nr_to_write - wbc.nr_to_write; + bdi_writeback_all(sb, nr_to_write); } EXPORT_SYMBOL(writeback_inodes_sb); @@ -1130,21 +1174,10 @@ EXPORT_SYMBOL(writeback_inodes_sb); * This function writes and waits on any dirty inode belonging to this * super_block. The number of pages synced is returned. */ -long sync_inodes_sb(struct super_block *sb) +void sync_inodes_sb(struct super_block *sb) { - struct writeback_control wbc = { - .sb = sb, - .bdi = sb->s_bdi, - .sync_mode = WB_SYNC_ALL, - .range_start = 0, - .range_end = LLONG_MAX, - }; - long nr_to_write = LONG_MAX; /* doesn't actually matter */ - - wbc.nr_to_write = nr_to_write; - bdi_start_writeback(&wbc); - wait_sb_inodes(&wbc); - return nr_to_write - wbc.nr_to_write; + bdi_sync_writeback(sb->s_bdi, sb); + wait_sb_inodes(sb); } EXPORT_SYMBOL(sync_inodes_sb);