From ffbd517d5a8c8e93ddd11046434fb029f3df73aa Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Mon, 20 Apr 2009 15:50:09 -0400
Subject: Btrfs: use WRITE_SYNC for synchronous writes

Part of reducing fsync/O_SYNC/O_DIRECT latencies is using WRITE_SYNC for
writes we plan on waiting on in the near future.  This patch
mirrors recent changes in other filesystems and the generic code to
use WRITE_SYNC when WB_SYNC_ALL is passed and to use WRITE_SYNC for
other latency critical writes.

Btrfs uses async worker threads for checksumming before the write is done,
and then again to actually submit the bios.  The bio submission code just
runs a per-device list of bios that need to be sent down the pipe.

This list is split into low priority and high priority lists so the
WRITE_SYNC IO happens first.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 92caa8035f3..fec18b43c2c 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2095,10 +2095,10 @@ static int write_dev_supers(struct btrfs_device *device,
 				device->barriers = 0;
 				get_bh(bh);
 				lock_buffer(bh);
-				ret = submit_bh(WRITE, bh);
+				ret = submit_bh(WRITE_SYNC, bh);
 			}
 		} else {
-			ret = submit_bh(WRITE, bh);
+			ret = submit_bh(WRITE_SYNC, bh);
 		}
 
 		if (!ret && wait) {
-- 
cgit v1.2.3-70-g09d2


From d313d7a31a752c88f7288692bd98e66d0789779b Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Mon, 20 Apr 2009 15:50:09 -0400
Subject: Btrfs: add a priority queue to the async thread helpers

Btrfs is using WRITE_SYNC_PLUG to send down synchronous IOs with a
higher priority.  But, the checksumming helper threads prevent it
from being fully effective.

There are two problems.  First, a big queue of pending checksumming
will delay the synchronous IO behind other lower priority writes.  Second,
the checksumming uses an ordered async work queue.  The ordering makes sure
that IOs are sent to the block layer in the same order they are sent
to the checksumming threads.  Usually this gives us less seeky IO.

But, when we start mixing IO priorities, the lower priority IO can delay
the higher priority IO.

This patch solves both problems by adding a high priority list to the async
helper threads, and a new btrfs_set_work_high_prio(), which is used
to make put a new async work item onto the higher priority list.

The ordering is still done on high priority IO, but all of the high
priority bios are ordered separately from the low priority bios.  This
ordering is purely an IO optimization, it is not involved in data
or metadata integrity.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/async-thread.c | 60 ++++++++++++++++++++++++++++++++++++++-----------
 fs/btrfs/async-thread.h |  2 ++
 fs/btrfs/disk-io.c      |  5 +++++
 fs/btrfs/extent_io.c    |  2 +-
 fs/btrfs/file.c         |  2 +-
 5 files changed, 56 insertions(+), 15 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c
index 51bfdfc8fcd..502c3d61de6 100644
--- a/fs/btrfs/async-thread.c
+++ b/fs/btrfs/async-thread.c
@@ -25,6 +25,7 @@
 #define WORK_QUEUED_BIT 0
 #define WORK_DONE_BIT 1
 #define WORK_ORDER_DONE_BIT 2
+#define WORK_HIGH_PRIO_BIT 3
 
 /*
  * container for the kthread task pointer and the list of pending work
@@ -36,6 +37,7 @@ struct btrfs_worker_thread {
 
 	/* list of struct btrfs_work that are waiting for service */
 	struct list_head pending;
+	struct list_head prio_pending;
 
 	/* list of worker threads from struct btrfs_workers */
 	struct list_head worker_list;
@@ -103,10 +105,16 @@ static noinline int run_ordered_completions(struct btrfs_workers *workers,
 
 	spin_lock_irqsave(&workers->lock, flags);
 
-	while (!list_empty(&workers->order_list)) {
-		work = list_entry(workers->order_list.next,
-				  struct btrfs_work, order_list);
-
+	while (1) {
+		if (!list_empty(&workers->prio_order_list)) {
+			work = list_entry(workers->prio_order_list.next,
+					  struct btrfs_work, order_list);
+		} else if (!list_empty(&workers->order_list)) {
+			work = list_entry(workers->order_list.next,
+					  struct btrfs_work, order_list);
+		} else {
+			break;
+		}
 		if (!test_bit(WORK_DONE_BIT, &work->flags))
 			break;
 
@@ -143,8 +151,14 @@ static int worker_loop(void *arg)
 	do {
 		spin_lock_irq(&worker->lock);
 again_locked:
-		while (!list_empty(&worker->pending)) {
-			cur = worker->pending.next;
+		while (1) {
+			if (!list_empty(&worker->prio_pending))
+				cur = worker->prio_pending.next;
+			else if (!list_empty(&worker->pending))
+				cur = worker->pending.next;
+			else
+				break;
+
 			work = list_entry(cur, struct btrfs_work, list);
 			list_del(&work->list);
 			clear_bit(WORK_QUEUED_BIT, &work->flags);
@@ -163,7 +177,6 @@ again_locked:
 
 			spin_lock_irq(&worker->lock);
 			check_idle_worker(worker);
-
 		}
 		if (freezing(current)) {
 			worker->working = 0;
@@ -178,7 +191,8 @@ again_locked:
 				 * jump_in?
 				 */
 				smp_mb();
-				if (!list_empty(&worker->pending))
+				if (!list_empty(&worker->pending) ||
+				    !list_empty(&worker->prio_pending))
 					continue;
 
 				/*
@@ -191,7 +205,8 @@ again_locked:
 				 */
 				schedule_timeout(1);
 				smp_mb();
-				if (!list_empty(&worker->pending))
+				if (!list_empty(&worker->pending) ||
+				    !list_empty(&worker->prio_pending))
 					continue;
 
 				if (kthread_should_stop())
@@ -200,7 +215,8 @@ again_locked:
 				/* still no more work?, sleep for real */
 				spin_lock_irq(&worker->lock);
 				set_current_state(TASK_INTERRUPTIBLE);
-				if (!list_empty(&worker->pending))
+				if (!list_empty(&worker->pending) ||
+				    !list_empty(&worker->prio_pending))
 					goto again_locked;
 
 				/*
@@ -248,6 +264,7 @@ void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max)
 	INIT_LIST_HEAD(&workers->worker_list);
 	INIT_LIST_HEAD(&workers->idle_list);
 	INIT_LIST_HEAD(&workers->order_list);
+	INIT_LIST_HEAD(&workers->prio_order_list);
 	spin_lock_init(&workers->lock);
 	workers->max_workers = max;
 	workers->idle_thresh = 32;
@@ -273,6 +290,7 @@ int btrfs_start_workers(struct btrfs_workers *workers, int num_workers)
 		}
 
 		INIT_LIST_HEAD(&worker->pending);
+		INIT_LIST_HEAD(&worker->prio_pending);
 		INIT_LIST_HEAD(&worker->worker_list);
 		spin_lock_init(&worker->lock);
 		atomic_set(&worker->num_pending, 0);
@@ -396,7 +414,10 @@ int btrfs_requeue_work(struct btrfs_work *work)
 		goto out;
 
 	spin_lock_irqsave(&worker->lock, flags);
-	list_add_tail(&work->list, &worker->pending);
+	if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
+		list_add_tail(&work->list, &worker->prio_pending);
+	else
+		list_add_tail(&work->list, &worker->pending);
 	atomic_inc(&worker->num_pending);
 
 	/* by definition we're busy, take ourselves off the idle
@@ -422,6 +443,11 @@ out:
 	return 0;
 }
 
+void btrfs_set_work_high_prio(struct btrfs_work *work)
+{
+	set_bit(WORK_HIGH_PRIO_BIT, &work->flags);
+}
+
 /*
  * places a struct btrfs_work into the pending queue of one of the kthreads
  */
@@ -438,7 +464,12 @@ int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
 	worker = find_worker(workers);
 	if (workers->ordered) {
 		spin_lock_irqsave(&workers->lock, flags);
-		list_add_tail(&work->order_list, &workers->order_list);
+		if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags)) {
+			list_add_tail(&work->order_list,
+				      &workers->prio_order_list);
+		} else {
+			list_add_tail(&work->order_list, &workers->order_list);
+		}
 		spin_unlock_irqrestore(&workers->lock, flags);
 	} else {
 		INIT_LIST_HEAD(&work->order_list);
@@ -446,7 +477,10 @@ int btrfs_queue_worker(struct btrfs_workers *workers, struct btrfs_work *work)
 
 	spin_lock_irqsave(&worker->lock, flags);
 
-	list_add_tail(&work->list, &worker->pending);
+	if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags))
+		list_add_tail(&work->list, &worker->prio_pending);
+	else
+		list_add_tail(&work->list, &worker->pending);
 	atomic_inc(&worker->num_pending);
 	check_busy_worker(worker);
 
diff --git a/fs/btrfs/async-thread.h b/fs/btrfs/async-thread.h
index 31be4ed8b63..1b511c109db 100644
--- a/fs/btrfs/async-thread.h
+++ b/fs/btrfs/async-thread.h
@@ -85,6 +85,7 @@ struct btrfs_workers {
 	 * of work items waiting for completion
 	 */
 	struct list_head order_list;
+	struct list_head prio_order_list;
 
 	/* lock for finding the next worker thread to queue on */
 	spinlock_t lock;
@@ -98,4 +99,5 @@ int btrfs_start_workers(struct btrfs_workers *workers, int num_workers);
 int btrfs_stop_workers(struct btrfs_workers *workers);
 void btrfs_init_workers(struct btrfs_workers *workers, char *name, int max);
 int btrfs_requeue_work(struct btrfs_work *work);
+void btrfs_set_work_high_prio(struct btrfs_work *work);
 #endif
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index fec18b43c2c..a6b83744b05 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -579,6 +579,10 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 	async->bio_flags = bio_flags;
 
 	atomic_inc(&fs_info->nr_async_submits);
+
+	if (rw & (1 << BIO_RW_SYNCIO))
+		btrfs_set_work_high_prio(&async->work);
+
 	btrfs_queue_worker(&fs_info->workers, &async->work);
 #if 0
 	int limit = btrfs_async_submit_limit(fs_info);
@@ -656,6 +660,7 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 		return btrfs_map_bio(BTRFS_I(inode)->root, rw, bio,
 				     mirror_num, 0);
 	}
+
 	/*
 	 * kthread helpers are used to submit writes so that checksumming
 	 * can happen in parallel across all CPUs
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 483b6727aaa..5d66cb27e42 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -2501,7 +2501,7 @@ int extent_write_full_page(struct extent_io_tree *tree, struct page *page,
 	};
 	struct writeback_control wbc_writepages = {
 		.bdi		= wbc->bdi,
-		.sync_mode	= WB_SYNC_NONE,
+		.sync_mode	= wbc->sync_mode,
 		.older_than_this = NULL,
 		.nr_to_write	= 64,
 		.range_start	= page_offset(page) + PAGE_CACHE_SIZE,
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 9c9fb46ccd0..e21c0060ee7 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1131,7 +1131,7 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf,
 		if (will_write) {
 			btrfs_fdatawrite_range(inode->i_mapping, pos,
 					       pos + write_bytes - 1,
-					       WB_SYNC_NONE);
+					       WB_SYNC_ALL);
 		} else {
 			balance_dirty_pages_ratelimited_nr(inode->i_mapping,
 							   num_pages);
-- 
cgit v1.2.3-70-g09d2


From 97e728d4353f38c87bf0804cdfd79a9b13fc2c3e Mon Sep 17 00:00:00 2001
From: Josef Bacik <jbacik@redhat.com>
Date: Tue, 21 Apr 2009 17:40:57 -0400
Subject: Btrfs: try to keep a healthy ratio of metadata vs data block groups

This patch makes the chunk allocator keep a good ratio of metadata vs data
block groups.  By default for every 8 data block groups, we'll allocate 1
metadata chunk, or about 12% of the disk will be allocated for metadata.  This
can be changed by specifying the metadata_ratio mount option.

This is simply the number of data block groups that have to be allocated to
force a metadata chunk allocation.  By making sure we allocate metadata chunks
more often, we are less likely to get into situations where the whole disk
has been allocated as data block groups.

Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/ctree.h       |  3 +++
 fs/btrfs/disk-io.c     |  1 +
 fs/btrfs/extent-tree.c | 28 +++++++++++++++++++++++++++-
 fs/btrfs/super.c       | 12 +++++++++++-
 4 files changed, 42 insertions(+), 2 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index ad96495dedc..213535f45da 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -881,6 +881,9 @@ struct btrfs_fs_info {
 	u64 metadata_alloc_profile;
 	u64 system_alloc_profile;
 
+	unsigned data_chunk_allocations;
+	unsigned metadata_ratio;
+
 	void *bdev_holder;
 };
 
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index a6b83744b05..44c94d808e2 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1604,6 +1604,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->btree_inode = new_inode(sb);
 	fs_info->btree_inode->i_ino = 1;
 	fs_info->btree_inode->i_nlink = 1;
+	fs_info->metadata_ratio = 8;
 
 	fs_info->thread_pool_size = min_t(unsigned long,
 					  num_online_cpus() + 2, 8);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 178df4c67de..2895a837323 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1918,15 +1918,29 @@ void btrfs_delalloc_free_space(struct btrfs_root *root, struct inode *inode,
 	spin_unlock(&info->lock);
 }
 
+static void force_metadata_allocation(struct btrfs_fs_info *info)
+{
+	struct list_head *head = &info->space_info;
+	struct btrfs_space_info *found;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(found, head, list) {
+		if (found->flags & BTRFS_BLOCK_GROUP_METADATA)
+			found->force_alloc = 1;
+	}
+	rcu_read_unlock();
+}
+
 static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 			  struct btrfs_root *extent_root, u64 alloc_bytes,
 			  u64 flags, int force)
 {
 	struct btrfs_space_info *space_info;
+	struct btrfs_fs_info *fs_info = extent_root->fs_info;
 	u64 thresh;
 	int ret = 0;
 
-	mutex_lock(&extent_root->fs_info->chunk_mutex);
+	mutex_lock(&fs_info->chunk_mutex);
 
 	flags = btrfs_reduce_alloc_profile(extent_root, flags);
 
@@ -1958,6 +1972,18 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 	}
 	spin_unlock(&space_info->lock);
 
+	/*
+	 * if we're doing a data chunk, go ahead and make sure that
+	 * we keep a reasonable number of metadata chunks allocated in the
+	 * FS as well.
+	 */
+	if (flags & BTRFS_BLOCK_GROUP_DATA) {
+		fs_info->data_chunk_allocations++;
+		if (!(fs_info->data_chunk_allocations %
+		      fs_info->metadata_ratio))
+			force_metadata_allocation(fs_info);
+	}
+
 	ret = btrfs_alloc_chunk(trans, extent_root, flags);
 	if (ret)
 		space_info->full = 1;
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 9744af9d71e..30c9a8ca2a5 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -68,7 +68,7 @@ enum {
 	Opt_degraded, Opt_subvol, Opt_device, Opt_nodatasum, Opt_nodatacow,
 	Opt_max_extent, Opt_max_inline, Opt_alloc_start, Opt_nobarrier,
 	Opt_ssd, Opt_thread_pool, Opt_noacl,  Opt_compress, Opt_notreelog,
-	Opt_flushoncommit, Opt_err,
+	Opt_ratio, Opt_flushoncommit, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -87,6 +87,7 @@ static match_table_t tokens = {
 	{Opt_noacl, "noacl"},
 	{Opt_notreelog, "notreelog"},
 	{Opt_flushoncommit, "flushoncommit"},
+	{Opt_ratio, "metadata_ratio=%d"},
 	{Opt_err, NULL},
 };
 
@@ -234,6 +235,15 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 			printk(KERN_INFO "btrfs: turning on flush-on-commit\n");
 			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
 			break;
+		case Opt_ratio:
+			intarg = 0;
+			match_int(&args[0], &intarg);
+			if (intarg) {
+				info->metadata_ratio = intarg;
+				printk(KERN_INFO "btrfs: metadata ratio %d\n",
+				       info->metadata_ratio);
+			}
+			break;
 		default:
 			break;
 		}
-- 
cgit v1.2.3-70-g09d2


From b7967db75a38df4891b22efe1b0969b9357eb946 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Mon, 27 Apr 2009 07:29:04 -0400
Subject: Btrfs: remove #if 0 code

Btrfs had some old code sitting around under #if 0, this drops it.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c   | 48 +-------------------------------
 fs/btrfs/extent_io.c | 63 ------------------------------------------
 fs/btrfs/file.c      | 78 ----------------------------------------------------
 3 files changed, 1 insertion(+), 188 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 44c94d808e2..77f9a3b824b 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -584,18 +584,7 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 		btrfs_set_work_high_prio(&async->work);
 
 	btrfs_queue_worker(&fs_info->workers, &async->work);
-#if 0
-	int limit = btrfs_async_submit_limit(fs_info);
-	if (atomic_read(&fs_info->nr_async_submits) > limit) {
-		wait_event_timeout(fs_info->async_submit_wait,
-			   (atomic_read(&fs_info->nr_async_submits) < limit),
-			   HZ/10);
 
-		wait_event_timeout(fs_info->async_submit_wait,
-			   (atomic_read(&fs_info->nr_async_bios) < limit),
-			   HZ/10);
-	}
-#endif
 	while (atomic_read(&fs_info->async_submit_draining) &&
 	      atomic_read(&fs_info->nr_async_submits)) {
 		wait_event(fs_info->async_submit_wait,
@@ -770,27 +759,6 @@ static void btree_invalidatepage(struct page *page, unsigned long offset)
 	}
 }
 
-#if 0
-static int btree_writepage(struct page *page, struct writeback_control *wbc)
-{
-	struct buffer_head *bh;
-	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
-	struct buffer_head *head;
-	if (!page_has_buffers(page)) {
-		create_empty_buffers(page, root->fs_info->sb->s_blocksize,
-					(1 << BH_Dirty)|(1 << BH_Uptodate));
-	}
-	head = page_buffers(page);
-	bh = head;
-	do {
-		if (buffer_dirty(bh))
-			csum_tree_block(root, bh, 0);
-		bh = bh->b_this_page;
-	} while (bh != head);
-	return block_write_full_page(page, btree_get_block, wbc);
-}
-#endif
-
 static struct address_space_operations btree_aops = {
 	.readpage	= btree_readpage,
 	.writepage	= btree_writepage,
@@ -1278,11 +1246,7 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 	int ret = 0;
 	struct btrfs_device *device;
 	struct backing_dev_info *bdi;
-#if 0
-	if ((bdi_bits & (1 << BDI_write_congested)) &&
-	    btrfs_congested_async(info, 0))
-		return 1;
-#endif
+
 	list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
 		if (!device->bdev)
 			continue;
@@ -2334,16 +2298,6 @@ int close_ctree(struct btrfs_root *root)
 	btrfs_stop_workers(&fs_info->endio_write_workers);
 	btrfs_stop_workers(&fs_info->submit_workers);
 
-#if 0
-	while (!list_empty(&fs_info->hashers)) {
-		struct btrfs_hasher *hasher;
-		hasher = list_entry(fs_info->hashers.next, struct btrfs_hasher,
-				    hashers);
-		list_del(&hasher->hashers);
-		crypto_free_hash(&fs_info->hash_tfm);
-		kfree(hasher);
-	}
-#endif
 	btrfs_close_devices(fs_info->fs_devices);
 	btrfs_mapping_tree_free(&fs_info->mapping_tree);
 
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index c33b54029d7..fe9eb990e44 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -1401,69 +1401,6 @@ out:
 	return total_bytes;
 }
 
-#if 0
-/*
- * helper function to lock both pages and extents in the tree.
- * pages must be locked first.
- */
-static int lock_range(struct extent_io_tree *tree, u64 start, u64 end)
-{
-	unsigned long index = start >> PAGE_CACHE_SHIFT;
-	unsigned long end_index = end >> PAGE_CACHE_SHIFT;
-	struct page *page;
-	int err;
-
-	while (index <= end_index) {
-		page = grab_cache_page(tree->mapping, index);
-		if (!page) {
-			err = -ENOMEM;
-			goto failed;
-		}
-		if (IS_ERR(page)) {
-			err = PTR_ERR(page);
-			goto failed;
-		}
-		index++;
-	}
-	lock_extent(tree, start, end, GFP_NOFS);
-	return 0;
-
-failed:
-	/*
-	 * we failed above in getting the page at 'index', so we undo here
-	 * up to but not including the page at 'index'
-	 */
-	end_index = index;
-	index = start >> PAGE_CACHE_SHIFT;
-	while (index < end_index) {
-		page = find_get_page(tree->mapping, index);
-		unlock_page(page);
-		page_cache_release(page);
-		index++;
-	}
-	return err;
-}
-
-/*
- * helper function to unlock both pages and extents in the tree.
- */
-static int unlock_range(struct extent_io_tree *tree, u64 start, u64 end)
-{
-	unsigned long index = start >> PAGE_CACHE_SHIFT;
-	unsigned long end_index = end >> PAGE_CACHE_SHIFT;
-	struct page *page;
-
-	while (index <= end_index) {
-		page = find_get_page(tree->mapping, index);
-		unlock_page(page);
-		page_cache_release(page);
-		index++;
-	}
-	unlock_extent(tree, start, end, GFP_NOFS);
-	return 0;
-}
-#endif
-
 /*
  * set the private field for a given byte offset in the tree.  If there isn't
  * an extent_state there already, this does nothing.
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index da3ed965c95..1d51dc38bb4 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -272,83 +272,6 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 	return 0;
 }
 
-int btrfs_check_file(struct btrfs_root *root, struct inode *inode)
-{
-	return 0;
-#if 0
-	struct btrfs_path *path;
-	struct btrfs_key found_key;
-	struct extent_buffer *leaf;
-	struct btrfs_file_extent_item *extent;
-	u64 last_offset = 0;
-	int nritems;
-	int slot;
-	int found_type;
-	int ret;
-	int err = 0;
-	u64 extent_end = 0;
-
-	path = btrfs_alloc_path();
-	ret = btrfs_lookup_file_extent(NULL, root, path, inode->i_ino,
-				       last_offset, 0);
-	while (1) {
-		nritems = btrfs_header_nritems(path->nodes[0]);
-		if (path->slots[0] >= nritems) {
-			ret = btrfs_next_leaf(root, path);
-			if (ret)
-				goto out;
-			nritems = btrfs_header_nritems(path->nodes[0]);
-		}
-		slot = path->slots[0];
-		leaf = path->nodes[0];
-		btrfs_item_key_to_cpu(leaf, &found_key, slot);
-		if (found_key.objectid != inode->i_ino)
-			break;
-		if (found_key.type != BTRFS_EXTENT_DATA_KEY)
-			goto out;
-
-		if (found_key.offset < last_offset) {
-			WARN_ON(1);
-			btrfs_print_leaf(root, leaf);
-			printk(KERN_ERR "inode %lu found offset %llu "
-			       "expected %llu\n", inode->i_ino,
-			       (unsigned long long)found_key.offset,
-			       (unsigned long long)last_offset);
-			err = 1;
-			goto out;
-		}
-		extent = btrfs_item_ptr(leaf, slot,
-					struct btrfs_file_extent_item);
-		found_type = btrfs_file_extent_type(leaf, extent);
-		if (found_type == BTRFS_FILE_EXTENT_REG) {
-			extent_end = found_key.offset +
-			     btrfs_file_extent_num_bytes(leaf, extent);
-		} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
-			struct btrfs_item *item;
-			item = btrfs_item_nr(leaf, slot);
-			extent_end = found_key.offset +
-			     btrfs_file_extent_inline_len(leaf, extent);
-			extent_end = (extent_end + root->sectorsize - 1) &
-				~((u64)root->sectorsize - 1);
-		}
-		last_offset = extent_end;
-		path->slots[0]++;
-	}
-	if (0 && last_offset < inode->i_size) {
-		WARN_ON(1);
-		btrfs_print_leaf(root, leaf);
-		printk(KERN_ERR "inode %lu found offset %llu size %llu\n",
-		       inode->i_ino, (unsigned long long)last_offset,
-		       (unsigned long long)inode->i_size);
-		err = 1;
-
-	}
-out:
-	btrfs_free_path(path);
-	return err;
-#endif
-}
-
 /*
  * this is very complex, but the basic idea is to drop all extents
  * in the range start - end.  hint_block is filled in with a block number
@@ -689,7 +612,6 @@ out:
 		unlock_extent(&BTRFS_I(inode)->io_tree, orig_locked_end,
 			      locked_end - 1, GFP_NOFS);
 	}
-	btrfs_check_file(root, inode);
 	return ret;
 }
 
-- 
cgit v1.2.3-70-g09d2


From 193f284d4985db0370a8a1bbdfb20df548cf9ffb Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Mon, 27 Apr 2009 07:29:05 -0400
Subject: Btrfs: ratelimit IO error printks

Btrfs has printks for various IO errors, including bad checksums and
mismatches between what we expect the block headers to contain and what
we actually find on the disk.

Longer term we need a real reporting mechanism for this, but for now
printk is going to have to do.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c | 38 +++++++++++++++++++++++++-------------
 fs/btrfs/inode.c   | 10 ++++++----
 2 files changed, 31 insertions(+), 17 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 77f9a3b824b..aa0c259b9c2 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -232,10 +232,14 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
 			memcpy(&found, result, csum_size);
 
 			read_extent_buffer(buf, &val, 0, csum_size);
-			printk(KERN_INFO "btrfs: %s checksum verify failed "
-			       "on %llu wanted %X found %X level %d\n",
-			       root->fs_info->sb->s_id,
-			       buf->start, val, found, btrfs_header_level(buf));
+			if (printk_ratelimit()) {
+				printk(KERN_INFO "btrfs: %s checksum verify "
+				       "failed on %llu wanted %X found %X "
+				       "level %d\n",
+				       root->fs_info->sb->s_id,
+				       (unsigned long long)buf->start, val, found,
+				       btrfs_header_level(buf));
+			}
 			if (result != (char *)&inline_result)
 				kfree(result);
 			return 1;
@@ -268,10 +272,13 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
 		ret = 0;
 		goto out;
 	}
-	printk("parent transid verify failed on %llu wanted %llu found %llu\n",
-	       (unsigned long long)eb->start,
-	       (unsigned long long)parent_transid,
-	       (unsigned long long)btrfs_header_generation(eb));
+	if (printk_ratelimit()) {
+		printk("parent transid verify failed on %llu wanted %llu "
+		       "found %llu\n",
+		       (unsigned long long)eb->start,
+		       (unsigned long long)parent_transid,
+		       (unsigned long long)btrfs_header_generation(eb));
+	}
 	ret = 1;
 	clear_extent_buffer_uptodate(io_tree, eb);
 out:
@@ -415,9 +422,12 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 
 	found_start = btrfs_header_bytenr(eb);
 	if (found_start != start) {
-		printk(KERN_INFO "btrfs bad tree block start %llu %llu\n",
-		       (unsigned long long)found_start,
-		       (unsigned long long)eb->start);
+		if (printk_ratelimit()) {
+			printk(KERN_INFO "btrfs bad tree block start "
+			       "%llu %llu\n",
+			       (unsigned long long)found_start,
+			       (unsigned long long)eb->start);
+		}
 		ret = -EIO;
 		goto err;
 	}
@@ -429,8 +439,10 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 		goto err;
 	}
 	if (check_tree_block_fsid(root, eb)) {
-		printk(KERN_INFO "btrfs bad fsid on block %llu\n",
-		       (unsigned long long)eb->start);
+		if (printk_ratelimit()) {
+			printk(KERN_INFO "btrfs bad fsid on block %llu\n",
+			       (unsigned long long)eb->start);
+		}
 		ret = -EIO;
 		goto err;
 	}
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 2fdb2995be6..552e08afc7f 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1823,10 +1823,12 @@ good:
 	return 0;
 
 zeroit:
-	printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
-	       "private %llu\n", page->mapping->host->i_ino,
-	       (unsigned long long)start, csum,
-	       (unsigned long long)private);
+	if (printk_ratelimit()) {
+		printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
+		       "private %llu\n", page->mapping->host->i_ino,
+		       (unsigned long long)start, csum,
+		       (unsigned long long)private);
+	}
 	memset(kaddr + offset, 1, end - start + 1);
 	flush_dcache_page(page);
 	kunmap_atomic(kaddr, KM_USER0);
-- 
cgit v1.2.3-70-g09d2


From 21380931eb4da4e29ac663d0221581282cbba208 Mon Sep 17 00:00:00 2001
From: Joel Becker <joel.becker@oracle.com>
Date: Tue, 21 Apr 2009 12:38:29 -0700
Subject: Btrfs: Fix a bunch of printk() warnings.

Just happened to notice a bunch of %llu vs u64 warnings.  Here's a patch
to cast them all.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c          |  6 +++---
 fs/btrfs/extent-tree.c      | 21 ++++++++++++++-------
 fs/btrfs/free-space-cache.c | 15 ++++++++++-----
 fs/btrfs/ioctl.c            |  6 ++++--
 fs/btrfs/super.c            | 15 +++++++++------
 5 files changed, 40 insertions(+), 23 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index aa0c259b9c2..0ff16d3331d 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1671,7 +1671,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	if (features) {
 		printk(KERN_ERR "BTRFS: couldn't mount because of "
 		       "unsupported optional features (%Lx).\n",
-		       features);
+		       (unsigned long long)features);
 		err = -EINVAL;
 		goto fail_iput;
 	}
@@ -1681,7 +1681,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	if (!(sb->s_flags & MS_RDONLY) && features) {
 		printk(KERN_ERR "BTRFS: couldn't mount RDWR because of "
 		       "unsupported option features (%Lx).\n",
-		       features);
+		       (unsigned long long)features);
 		err = -EINVAL;
 		goto fail_iput;
 	}
@@ -2273,7 +2273,7 @@ int close_ctree(struct btrfs_root *root)
 
 	if (fs_info->delalloc_bytes) {
 		printk(KERN_INFO "btrfs: at unmount delalloc count %llu\n",
-		       fs_info->delalloc_bytes);
+		       (unsigned long long)fs_info->delalloc_bytes);
 	}
 	if (fs_info->total_ref_cache_size) {
 		printk(KERN_INFO "btrfs: at umount reference cache size %llu\n",
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 2895a837323..e4966444811 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -1844,10 +1844,14 @@ again:
 		printk(KERN_ERR "no space left, need %llu, %llu delalloc bytes"
 		       ", %llu bytes_used, %llu bytes_reserved, "
 		       "%llu bytes_pinned, %llu bytes_readonly, %llu may use"
-		       "%llu total\n", bytes, data_sinfo->bytes_delalloc,
-		       data_sinfo->bytes_used, data_sinfo->bytes_reserved,
-		       data_sinfo->bytes_pinned, data_sinfo->bytes_readonly,
-		       data_sinfo->bytes_may_use, data_sinfo->total_bytes);
+		       "%llu total\n", (unsigned long long)bytes,
+		       (unsigned long long)data_sinfo->bytes_delalloc,
+		       (unsigned long long)data_sinfo->bytes_used,
+		       (unsigned long long)data_sinfo->bytes_reserved,
+		       (unsigned long long)data_sinfo->bytes_pinned,
+		       (unsigned long long)data_sinfo->bytes_readonly,
+		       (unsigned long long)data_sinfo->bytes_may_use,
+		       (unsigned long long)data_sinfo->total_bytes);
 		return -ENOSPC;
 	}
 	data_sinfo->bytes_may_use += bytes;
@@ -2824,9 +2828,12 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes)
 				    info->bytes_pinned - info->bytes_reserved),
 	       (info->full) ? "" : "not ");
 	printk(KERN_INFO "space_info total=%llu, pinned=%llu, delalloc=%llu,"
-	       " may_use=%llu, used=%llu\n", info->total_bytes,
-	       info->bytes_pinned, info->bytes_delalloc, info->bytes_may_use,
-	       info->bytes_used);
+	       " may_use=%llu, used=%llu\n",
+	       (unsigned long long)info->total_bytes,
+	       (unsigned long long)info->bytes_pinned,
+	       (unsigned long long)info->bytes_delalloc,
+	       (unsigned long long)info->bytes_may_use,
+	       (unsigned long long)info->bytes_used);
 
 	down_read(&info->groups_sem);
 	list_for_each_entry(cache, &info->block_groups, list) {
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 768b9523662..0bc93657b46 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -332,13 +332,17 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group,
 			printk(KERN_ERR "couldn't find space %llu to free\n",
 			       (unsigned long long)offset);
 			printk(KERN_ERR "cached is %d, offset %llu bytes %llu\n",
-			       block_group->cached, block_group->key.objectid,
-			       block_group->key.offset);
+			       block_group->cached,
+			       (unsigned long long)block_group->key.objectid,
+			       (unsigned long long)block_group->key.offset);
 			btrfs_dump_free_space(block_group, bytes);
 		} else if (info) {
 			printk(KERN_ERR "hmm, found offset=%llu bytes=%llu, "
 			       "but wanted offset=%llu bytes=%llu\n",
-			       info->offset, info->bytes, offset, bytes);
+			       (unsigned long long)info->offset,
+			       (unsigned long long)info->bytes,
+			       (unsigned long long)offset,
+			       (unsigned long long)bytes);
 		}
 		WARN_ON(1);
 	}
@@ -357,8 +361,9 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group,
 		info = rb_entry(n, struct btrfs_free_space, offset_index);
 		if (info->bytes >= bytes)
 			count++;
-		printk(KERN_ERR "entry offset %llu, bytes %llu\n", info->offset,
-		       info->bytes);
+		printk(KERN_ERR "entry offset %llu, bytes %llu\n",
+		       (unsigned long long)info->offset,
+		       (unsigned long long)info->bytes);
 	}
 	printk(KERN_INFO "%d blocks of free space at or bigger than bytes is"
 	       "\n", count);
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index f4e5d2e5ece..48762aa1e94 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -483,11 +483,13 @@ static int btrfs_ioctl_resize(struct btrfs_root *root, void __user *arg)
 		*devstr = '\0';
 		devstr = vol_args->name;
 		devid = simple_strtoull(devstr, &end, 10);
-		printk(KERN_INFO "resizing devid %llu\n", devid);
+		printk(KERN_INFO "resizing devid %llu\n",
+		       (unsigned long long)devid);
 	}
 	device = btrfs_find_device(root, devid, NULL, NULL);
 	if (!device) {
-		printk(KERN_INFO "resizer unable to find device %llu\n", devid);
+		printk(KERN_INFO "resizer unable to find device %llu\n",
+		       (unsigned long long)devid);
 		ret = -EINVAL;
 		goto out_unlock;
 	}
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 30c9a8ca2a5..bf0e84c7560 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -196,7 +196,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				info->max_extent = max_t(u64,
 					info->max_extent, root->sectorsize);
 				printk(KERN_INFO "btrfs: max_extent at %llu\n",
-				       info->max_extent);
+				       (unsigned long long)info->max_extent);
 			}
 			break;
 		case Opt_max_inline:
@@ -211,7 +211,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 						root->sectorsize);
 				}
 				printk(KERN_INFO "btrfs: max_inline at %llu\n",
-					info->max_inline);
+					(unsigned long long)info->max_inline);
 			}
 			break;
 		case Opt_alloc_start:
@@ -221,7 +221,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 				kfree(num);
 				printk(KERN_INFO
 					"btrfs: allocations start at %llu\n",
-					info->alloc_start);
+					(unsigned long long)info->alloc_start);
 			}
 			break;
 		case Opt_noacl:
@@ -420,11 +420,14 @@ static int btrfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
 	if (btrfs_test_opt(root, NOBARRIER))
 		seq_puts(seq, ",nobarrier");
 	if (info->max_extent != (u64)-1)
-		seq_printf(seq, ",max_extent=%llu", info->max_extent);
+		seq_printf(seq, ",max_extent=%llu",
+			   (unsigned long long)info->max_extent);
 	if (info->max_inline != 8192 * 1024)
-		seq_printf(seq, ",max_inline=%llu", info->max_inline);
+		seq_printf(seq, ",max_inline=%llu",
+			   (unsigned long long)info->max_inline);
 	if (info->alloc_start != 0)
-		seq_printf(seq, ",alloc_start=%llu", info->alloc_start);
+		seq_printf(seq, ",alloc_start=%llu",
+			   (unsigned long long)info->alloc_start);
 	if (info->thread_pool_size !=  min_t(unsigned long,
 					     num_online_cpus() + 2, 8))
 		seq_printf(seq, ",thread_pool=%d", info->thread_pool_size);
-- 
cgit v1.2.3-70-g09d2


From cc7b0c9b701a079016183f3546b4d720194b367f Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Thu, 14 May 2009 13:31:21 -0400
Subject: Btrfs: remove some WARN_ONs in the IO failure path

These debugging WARN_ONs make too much console noise during regular
IO failures.  An IO failure will still generate a number of messages
as we verify checksums etc, but these two are not needed.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c | 2 --
 fs/btrfs/inode.c   | 1 -
 2 files changed, 3 deletions(-)

(limited to 'fs/btrfs/disk-io.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 0ff16d3331d..4b0ea0b80c2 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -848,8 +848,6 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
 
 	if (ret == 0)
 		set_bit(EXTENT_BUFFER_UPTODATE, &buf->bflags);
-	else
-		WARN_ON(1);
 	return buf;
 
 }
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 22450bd972d..1c8b0190d03 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -4296,7 +4296,6 @@ out:
 	}
 	if (err) {
 		free_extent_map(em);
-		WARN_ON(1);
 		return ERR_PTR(err);
 	}
 	return em;
-- 
cgit v1.2.3-70-g09d2