Btrfs: take i_mutex before generic_write_checks

[linux-2.6-block.git] / fs / btrfs / file.c

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 7c3cd248d8d6151ce5a89d876c2541b2abf73179..f155179877a6ff81cb741a8a47b13a6aebc992a4 100644 (file)
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -22,7 +22,6 @@
 #include <linux/time.h>
 #include <linux/init.h>
 #include <linux/string.h>
-#include <linux/smp_lock.h>
 #include <linux/backing-dev.h>
 #include <linux/mpage.h>
 #include <linux/swap.h>
@@ -113,8 +112,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
        int err = 0;
        int i;
        struct inode *inode = fdentry(file)->d_inode;
-       struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
-       u64 hint_byte;
        u64 num_bytes;
        u64 start_pos;
        u64 end_of_last_block;
@@ -126,23 +123,10 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
                    root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
 
        end_of_last_block = start_pos + num_bytes - 1;
+       err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
+       if (err)
+               return err;
 
-       lock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
-       trans = btrfs_join_transaction(root, 1);
-       if (!trans) {
-               err = -ENOMEM;
-               goto out_unlock;
-       }
-       btrfs_set_trans_block_group(trans, inode);
-       hint_byte = 0;
-
-       set_extent_uptodate(io_tree, start_pos, end_of_last_block, GFP_NOFS);
-
-       /* check for reserved extents on each page, we don't want
-        * to reset the delalloc bit on things that already have
-        * extents reserved.
-        */
-       btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
        for (i = 0; i < num_pages; i++) {
                struct page *p = pages[i];
                SetPageUptodate(p);
@@ -156,9 +140,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
                 * at this time.
                 */
        }
-       err = btrfs_end_transaction(trans, root);
-out_unlock:
-       unlock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
        return err;
 }
 
@@ -190,18 +171,18 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
                if (!split2)
                        split2 = alloc_extent_map(GFP_NOFS);
 
-               spin_lock(&em_tree->lock);
+               write_lock(&em_tree->lock);
                em = lookup_extent_mapping(em_tree, start, len);
                if (!em) {
-                       spin_unlock(&em_tree->lock);
+                       write_unlock(&em_tree->lock);
                        break;
                }
                flags = em->flags;
                if (skip_pinned && test_bit(EXTENT_FLAG_PINNED, &em->flags)) {
-                       spin_unlock(&em_tree->lock);
                        if (em->start <= start &&
                            (!testend || em->start + em->len >= start + len)) {
                                free_extent_map(em);
+                               write_unlock(&em_tree->lock);
                                break;
                        }
                        if (start < em->start) {
@@ -211,6 +192,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
                                start = em->start + em->len;
                        }
                        free_extent_map(em);
+                       write_unlock(&em_tree->lock);
                        continue;
                }
                compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
@@ -261,7 +243,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
                        free_extent_map(split);
                        split = NULL;
                }
-               spin_unlock(&em_tree->lock);
+               write_unlock(&em_tree->lock);
 
                /* once for us */
                free_extent_map(em);
@@ -290,7 +272,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans,
                       struct btrfs_root *root, struct inode *inode,
                       u64 start, u64 end, u64 locked_end,
-                      u64 inline_limit, u64 *hint_byte)
+                      u64 inline_limit, u64 *hint_byte, int drop_cache)
 {
        u64 extent_end = 0;
        u64 search_start = start;
@@ -315,7 +297,8 @@ noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans,
        int ret;
 
        inline_limit = 0;
-       btrfs_drop_extent_cache(inode, start, end - 1, 0);
+       if (drop_cache)
+               btrfs_drop_extent_cache(inode, start, end - 1, 0);
 
        path = btrfs_alloc_path();
        if (!path)
@@ -937,21 +920,35 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf,
        start_pos = pos;
 
        vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+
+       /* do the reserve before the mutex lock in case we have to do some
+        * flushing.  We wouldn't deadlock, but this is more polite.
+        */
+       err = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+       if (err)
+               goto out_nolock;
+
+       mutex_lock(&inode->i_mutex);
+
        current->backing_dev_info = inode->i_mapping->backing_dev_info;
        err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode));
        if (err)
-               goto out_nolock;
+               goto out;
+
        if (count == 0)
-               goto out_nolock;
+               goto out;
 
        err = file_remove_suid(file);
        if (err)
-               goto out_nolock;
+               goto out;
+
        file_update_time(file);
 
        pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
 
-       mutex_lock(&inode->i_mutex);
+       /* generic_write_checks can change our pos */
+       start_pos = pos;
+
        BTRFS_I(inode)->sequence++;
        first_index = pos >> PAGE_CACHE_SHIFT;
        last_index = (pos + count) >> PAGE_CACHE_SHIFT;
@@ -1048,6 +1045,7 @@ out:
        mutex_unlock(&inode->i_mutex);
        if (ret)
                err = ret;
+       btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
 
 out_nolock:
        kfree(pages);