Btrfs: fix snapshot vs nocow writting

While running fsstress and snapshots concurrently, we will hit something
like followings:

Thread 1			Thread 2

|->fallocate
  |->write pages
    |->join transaction
       |->add ordered extent
    |->end transaction
				|->flushing data
				  |->creating pending snapshots
|->write data into src root's
   fallocated space

After above work flows finished, we will get a state that source and
snapshot root share same space, but source root have written data into
fallocated space, this will make fsck fail to verify checksums for
snapshot root's preallocating file extent data.Nocow writting also
has this same problem.

Fix this problem by syncing snapshots with nocow writting:

 1.for nocow writting,if there are pending snapshots, we will
 fall into COW way.

 2.if there are pending nocow writes, snapshots for this root
 will be blocked until nocow writting finish.

Reported-by: Gui Hecheng <guihc.fnst@cn.fujitsu.com>
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Signed-off-by: Chris Mason <clm@fb.com>

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 0ec876657923511ff21184f8dcfbc191e0c5b9d5..251db68148b23a39187653f7f77e979fd282435d 100644 (file)
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1270,6 +1270,15 @@ next_slot:
                        disk_bytenr += extent_offset;
                        disk_bytenr += cur_offset - found_key.offset;
                        num_bytes = min(end + 1, extent_end) - cur_offset;
+                       /*
+                        * if there are pending snapshots for this root,
+                        * we fall into common COW way.
+                        */
+                       if (!nolock) {
+                               err = btrfs_start_nocow_write(root);
+                               if (!err)
+                                       goto out_check;
+                       }
                        /*
                         * force cow if csum exists in the range.
                         * this ensure that csum for a given extent are
@@ -1289,6 +1298,8 @@ next_slot:
 out_check:
                if (extent_end <= start) {
                        path->slots[0]++;
+                       if (!nolock && nocow)
+                               btrfs_end_nocow_write(root);
                        goto next_slot;
                }
                if (!nocow) {
@@ -1306,8 +1317,11 @@ out_check:
                        ret = cow_file_range(inode, locked_page,
                                             cow_start, found_key.offset - 1,
                                             page_started, nr_written, 1);
-                       if (ret)
+                       if (ret) {
+                               if (!nolock && nocow)
+                                       btrfs_end_nocow_write(root);
                                goto error;
+                       }
                        cow_start = (u64)-1;
                }
 
@@ -1354,8 +1368,11 @@ out_check:
                    BTRFS_DATA_RELOC_TREE_OBJECTID) {
                        ret = btrfs_reloc_clone_csums(inode, cur_offset,
                                                      num_bytes);
-                       if (ret)
+                       if (ret) {
+                               if (!nolock && nocow)
+                                       btrfs_end_nocow_write(root);
                                goto error;
+                       }
                }
 
                extent_clear_unlock_delalloc(inode, cur_offset,
@@ -1363,6 +1380,8 @@ out_check:
                                             locked_page, EXTENT_LOCKED |
                                             EXTENT_DELALLOC, PAGE_UNLOCK |
                                             PAGE_SET_PRIVATE2);
+               if (!nolock && nocow)
+                       btrfs_end_nocow_write(root);
                cur_offset = extent_end;
                if (cur_offset > end)
                        break;