*/
int nr_pdflush_threads;
+/*
+ * Passed into wb_writeback(), essentially a subset of writeback_control
+ */
+struct wb_writeback_args {
+ long nr_pages;
+ struct super_block *sb;
+ enum writeback_sync_modes sync_mode;
+ int for_kupdate;
+ int range_cyclic;
+};
+
/*
* Work items for the bdi_writeback threads
*/
struct bdi_work {
- struct list_head list;
- struct list_head wait_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 super_block *sb;
- unsigned long nr_pages;
- enum writeback_sync_modes sync_mode;
+ struct wb_writeback_args args; /* writeback arguments */
- unsigned long state;
+ unsigned long state; /* flag bits, see WS_* */
};
enum {
}
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->sb = wbc->sb;
- work->nr_pages = wbc->nr_to_write;
- work->sync_mode = wbc->sync_mode;
+ work->args = *args;
work->state = WS_USED;
}
-static inline void bdi_work_init_on_stack(struct bdi_work *work,
- struct writeback_control *wbc)
-{
- bdi_work_init(work, wbc);
- work->state |= WS_ONSTACK;
-}
-
/**
* writeback_in_progress - determine whether there is writeback in progress
* @bdi: the device's backing_dev_info structure.
{
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);
}
static void wb_work_complete(struct bdi_work *work)
{
- const enum writeback_sync_modes sync_mode = work->sync_mode;
+ 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.
* 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);
}
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
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);
}
}
TASK_UNINTERRUPTIBLE);
}
-static struct bdi_work *bdi_alloc_work(struct writeback_control *wbc)
+static void bdi_alloc_queue_work(struct backing_dev_info *bdi,
+ 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;
- return work;
+ if (wb->task)
+ wake_up_process(wb->task);
+ }
}
-void bdi_start_writeback(struct writeback_control *wbc)
+/**
+ * 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)
{
- const bool must_wait = wbc->sync_mode == WB_SYNC_ALL;
- struct bdi_work work_stack, *work = NULL;
+ struct wb_writeback_args args = {
+ .sb = sb,
+ .sync_mode = WB_SYNC_ALL,
+ .nr_pages = LONG_MAX,
+ .range_cyclic = 0,
+ };
+ struct bdi_work work;
- if (!must_wait)
- work = bdi_alloc_work(wbc);
+ bdi_work_init(&work, &args);
+ work.state |= WS_ONSTACK;
- if (!work) {
- work = &work_stack;
- bdi_work_init_on_stack(work, wbc);
- }
+ bdi_queue_work(bdi, &work);
+ bdi_wait_on_work_clear(&work);
+}
- bdi_queue_work(wbc->bdi, work);
+/**
+ * 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,
+ };
- /*
- * If the sync mode is WB_SYNC_ALL, block waiting for the work to
- * complete. If not, we only need to wait for the work to be started,
- * if we allocated it on-stack. We use the same mechanism, if the
- * wait bit is set in the bdi_work struct, then threads will not
- * clear pending until after they are done.
- *
- * Note that work == &work_stack if must_wait is true, so we don't
- * need to do call_rcu() here ever, since the completion path will
- * have done that for us.
- */
- if (must_wait || work == &work_stack) {
- bdi_wait_on_work_clear(work);
- if (work != &work_stack)
- call_rcu(&work->rcu_head, bdi_work_free);
- }
+ bdi_alloc_queue_work(bdi, &args);
}
/*
* older_than_this takes precedence over nr_to_write. So we'll only write back
* all dirty pages if they are all attached to "old" mappings.
*/
-static long wb_writeback(struct bdi_writeback *wb, long nr_pages,
- struct super_block *sb,
- enum writeback_sync_modes sync_mode, int for_kupdate)
+static long wb_writeback(struct bdi_writeback *wb,
+ struct wb_writeback_args *args)
{
struct writeback_control wbc = {
.bdi = wb->bdi,
- .sb = sb,
- .sync_mode = sync_mode,
+ .sb = args->sb,
+ .sync_mode = args->sync_mode,
.older_than_this = NULL,
- .for_kupdate = for_kupdate,
- .range_cyclic = 1,
+ .for_kupdate = args->for_kupdate,
+ .range_cyclic = args->range_cyclic,
};
unsigned long oldest_jif;
long wrote = 0;
oldest_jif = jiffies -
msecs_to_jiffies(dirty_expire_interval * 10);
}
+ if (!wbc.range_cyclic) {
+ wbc.range_start = 0;
+ wbc.range_end = LLONG_MAX;
+ }
for (;;) {
/*
* Don't flush anything for non-integrity writeback where
* no nr_pages was given
*/
- if (!for_kupdate && nr_pages <= 0 && sync_mode == WB_SYNC_NONE)
+ if (!args->for_kupdate && args->nr_pages <= 0 &&
+ args->sync_mode == WB_SYNC_NONE)
break;
/*
* periodic background writeout and we are below the
* background dirty threshold, don't do anything
*/
- if (for_kupdate && nr_pages <= 0 && !over_bground_thresh())
+ if (args->for_kupdate && args->nr_pages <= 0 &&
+ !over_bground_thresh())
break;
wbc.more_io = 0;
wbc.nr_to_write = MAX_WRITEBACK_PAGES;
wbc.pages_skipped = 0;
writeback_inodes_wb(wb, &wbc);
- nr_pages -= MAX_WRITEBACK_PAGES - wbc.nr_to_write;
+ args->nr_pages -= MAX_WRITEBACK_PAGES - wbc.nr_to_write;
wrote += MAX_WRITEBACK_PAGES - wbc.nr_to_write;
/*
/*
* 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)
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;
global_page_state(NR_UNSTABLE_NFS) +
(inodes_stat.nr_inodes - inodes_stat.nr_unused);
- if (nr_pages)
- return wb_writeback(wb, nr_pages, NULL, WB_SYNC_NONE, 1);
+ if (nr_pages) {
+ struct wb_writeback_args args = {
+ .nr_pages = nr_pages,
+ .sync_mode = WB_SYNC_NONE,
+ .for_kupdate = 1,
+ .range_cyclic = 1,
+ };
+
+ return wb_writeback(wb, &args);
+ }
return 0;
}
{
struct backing_dev_info *bdi = wb->bdi;
struct bdi_work *work;
- long nr_pages, wrote = 0;
+ long wrote = 0;
while ((work = get_next_work_item(bdi, wb)) != NULL) {
- enum writeback_sync_modes sync_mode;
-
- nr_pages = work->nr_pages;
+ struct wb_writeback_args args = work->args;
/*
* Override sync mode, in case we must wait for completion
*/
if (force_wait)
- work->sync_mode = sync_mode = WB_SYNC_ALL;
- else
- sync_mode = work->sync_mode;
+ work->args.sync_mode = args.sync_mode = WB_SYNC_ALL;
/*
* If this isn't a data integrity operation, just notify
* that we have seen this work and we are now starting it.
*/
- if (sync_mode == WB_SYNC_NONE)
+ if (args.sync_mode == WB_SYNC_NONE)
wb_clear_pending(wb, work);
- wrote += wb_writeback(wb, nr_pages, work->sb, sync_mode, 0);
+ wrote += wb_writeback(wb, &args);
/*
* This is a data integrity writeback, so only do the
* notification when we have completed the work.
*/
- if (sync_mode == WB_SYNC_ALL)
+ if (args.sync_mode == WB_SYNC_ALL)
wb_clear_pending(wb, work);
}
}
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();
}
}
/*
- * Schedule writeback for all backing devices. Expensive! If this is a data
- * integrity operation, writeback will be complete when this returns. If
- * we are simply called for WB_SYNC_NONE, then writeback will merely be
- * scheduled to run.
+ * 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)
{
- const bool must_wait = wbc->sync_mode == WB_SYNC_ALL;
+ struct wb_writeback_args args = {
+ .sb = sb,
+ .nr_pages = nr_pages,
+ .sync_mode = WB_SYNC_NONE,
+ };
struct backing_dev_info *bdi;
- struct bdi_work *work;
- LIST_HEAD(list);
-
-restart:
- spin_lock(&bdi_lock);
- list_for_each_entry(bdi, &bdi_list, bdi_list) {
- struct bdi_work *work;
+ rcu_read_lock();
+ list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) {
if (!bdi_has_dirty_io(bdi))
continue;
- /*
- * If work allocation fails, do the writes inline. We drop
- * the lock and restart the list writeout. This should be OK,
- * since this happens rarely and because the writeout should
- * eventually make more free memory available.
- */
- work = bdi_alloc_work(wbc);
- if (!work) {
- struct writeback_control __wbc;
-
- /*
- * Not a data integrity writeout, just continue
- */
- if (!must_wait)
- continue;
-
- spin_unlock(&bdi_lock);
- __wbc = *wbc;
- __wbc.bdi = bdi;
- writeback_inodes_wbc(&__wbc);
- goto restart;
- }
- if (must_wait)
- list_add_tail(&work->wait_list, &list);
-
- bdi_queue_work(bdi, work);
+ bdi_alloc_queue_work(bdi, &args);
}
- spin_unlock(&bdi_lock);
-
- /*
- * If this is for WB_SYNC_ALL, wait for pending work to complete
- * before returning.
- */
- while (!list_empty(&list)) {
- work = list_entry(list.next, struct bdi_work, wait_list);
- list_del(&work->wait_list);
- bdi_wait_on_work_clear(work);
- call_rcu(&work->rcu_head, bdi_work_free);
- }
+ rcu_read_unlock();
}
/*
*/
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)
* 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;
* 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);
* 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))
* 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;
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);
* 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,
- .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_writeback_all(&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);
EXPORT_SYMBOL_GPL(default_backing_dev_info);
static struct class *bdi_class;
+
+/*
+ * bdi_lock protects updates to bdi_list and bdi_pending_list, as well as
+ * reader side protection for bdi_pending_list. bdi_list has RCU reader side
+ * locking.
+ */
DEFINE_SPINLOCK(bdi_lock);
LIST_HEAD(bdi_list);
LIST_HEAD(bdi_pending_list);
/*
* Add us to the active bdi_list
*/
- spin_lock(&bdi_lock);
- list_add(&bdi->bdi_list, &bdi_list);
- spin_unlock(&bdi_lock);
+ spin_lock_bh(&bdi_lock);
+ list_add_rcu(&bdi->bdi_list, &bdi_list);
+ spin_unlock_bh(&bdi_lock);
bdi_task_init(bdi, wb);
if (wb_has_dirty_io(me) || !list_empty(&me->bdi->work_list))
wb_do_writeback(me, 0);
- spin_lock(&bdi_lock);
+ spin_lock_bh(&bdi_lock);
/*
* Check if any existing bdi's have dirty data without
if (list_empty(&bdi_pending_list)) {
unsigned long wait;
- spin_unlock(&bdi_lock);
+ spin_unlock_bh(&bdi_lock);
wait = msecs_to_jiffies(dirty_writeback_interval * 10);
schedule_timeout(wait);
try_to_freeze();
bdi = list_entry(bdi_pending_list.next, struct backing_dev_info,
bdi_list);
list_del_init(&bdi->bdi_list);
- spin_unlock(&bdi_lock);
+ spin_unlock_bh(&bdi_lock);
wb = &bdi->wb;
wb->task = kthread_run(bdi_start_fn, wb, "flush-%s",
* a chance to flush other bdi's to free
* memory.
*/
- spin_lock(&bdi_lock);
+ spin_lock_bh(&bdi_lock);
list_add_tail(&bdi->bdi_list, &bdi_pending_list);
- spin_unlock(&bdi_lock);
+ spin_unlock_bh(&bdi_lock);
bdi_flush_io(bdi);
}
return 0;
}
+static void bdi_add_to_pending(struct rcu_head *head)
+{
+ struct backing_dev_info *bdi;
+
+ bdi = container_of(head, struct backing_dev_info, rcu_head);
+ INIT_LIST_HEAD(&bdi->bdi_list);
+
+ spin_lock(&bdi_lock);
+ list_add_tail(&bdi->bdi_list, &bdi_pending_list);
+ spin_unlock(&bdi_lock);
+
+ /*
+ * We are now on the pending list, wake up bdi_forker_task()
+ * to finish the job and add us back to the active bdi_list
+ */
+ wake_up_process(default_backing_dev_info.wb.task);
+}
+
/*
* Add the default flusher task that gets created for any bdi
* that has dirty data pending writeout
* waiting for previous additions to finish.
*/
if (!test_and_set_bit(BDI_pending, &bdi->state)) {
- list_move_tail(&bdi->bdi_list, &bdi_pending_list);
+ list_del_rcu(&bdi->bdi_list);
/*
- * We are now on the pending list, wake up bdi_forker_task()
- * to finish the job and add us back to the active bdi_list
+ * We must wait for the current RCU period to end before
+ * moving to the pending list. So schedule that operation
+ * from an RCU callback.
*/
- wake_up_process(default_backing_dev_info.wb.task);
+ call_rcu(&bdi->rcu_head, bdi_add_to_pending);
}
}
+/*
+ * Remove bdi from bdi_list, and ensure that it is no longer visible
+ */
+static void bdi_remove_from_list(struct backing_dev_info *bdi)
+{
+ spin_lock_bh(&bdi_lock);
+ list_del_rcu(&bdi->bdi_list);
+ spin_unlock_bh(&bdi_lock);
+
+ synchronize_rcu();
+}
+
int bdi_register(struct backing_dev_info *bdi, struct device *parent,
const char *fmt, ...)
{
goto exit;
}
- spin_lock(&bdi_lock);
- list_add_tail(&bdi->bdi_list, &bdi_list);
- spin_unlock(&bdi_lock);
+ spin_lock_bh(&bdi_lock);
+ list_add_tail_rcu(&bdi->bdi_list, &bdi_list);
+ spin_unlock_bh(&bdi_lock);
bdi->dev = dev;
wb->task = NULL;
ret = -ENOMEM;
- spin_lock(&bdi_lock);
- list_del(&bdi->bdi_list);
- spin_unlock(&bdi_lock);
+ bdi_remove_from_list(bdi);
goto exit;
}
}
/*
* Make sure nobody finds us on the bdi_list anymore
*/
- spin_lock(&bdi_lock);
- list_del(&bdi->bdi_list);
- spin_unlock(&bdi_lock);
+ bdi_remove_from_list(bdi);
/*
* Finally, kill the kernel threads. We don't need to be RCU
bdi->max_ratio = 100;
bdi->max_prop_frac = PROP_FRAC_BASE;
spin_lock_init(&bdi->wb_lock);
+ INIT_RCU_HEAD(&bdi->rcu_head);
INIT_LIST_HEAD(&bdi->bdi_list);
INIT_LIST_HEAD(&bdi->wb_list);
INIT_LIST_HEAD(&bdi->work_list);
{
int i;
- WARN_ON(bdi_has_dirty_io(bdi));
+ /*
+ * Splice our entries to the default_backing_dev_info, if this
+ * bdi disappears
+ */
+ if (bdi_has_dirty_io(bdi)) {
+ struct bdi_writeback *dst = &default_backing_dev_info.wb;
+
+ spin_lock(&inode_lock);
+ list_splice(&bdi->wb.b_dirty, &dst->b_dirty);
+ list_splice(&bdi->wb.b_io, &dst->b_io);
+ list_splice(&bdi->wb.b_more_io, &dst->b_more_io);
+ spin_unlock(&inode_lock);
+ }
bdi_unregister(bdi);
git.openpandora.org / pandora-kernel.git / commitdiff