linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2 00/10] Chunk allocator unification
@ 2018-02-09  7:44 Qu Wenruo
  2018-02-09  7:44 ` [PATCH v2] btrfs-progs: ctree: Add extra level check for read_node_slot() Qu Wenruo
                   ` (10 more replies)
  0 siblings, 11 replies; 24+ messages in thread
From: Qu Wenruo @ 2018-02-09  7:44 UTC (permalink / raw)
  To: linux-btrfs, dsterba

This patchset can be fetched from github:
https://github.com/adam900710/btrfs-progs/tree/libbtrfs_prepare

This patchset unified a large part of chunk allocator (free device
extent search) between kernel and btrfs-progs.
And reuses kernel function structures like btrfs_finish_chunk_alloc()
and btrfs_alloc_dev_extent().

Before the unification:
               Kernel                |              Btrfs-progs
 btrfs_alloc_chunk()                 | btrfs_alloc_chunk()
 |- __btrfs_alloc_chunk()            | |- Do all the work
                                     |
 btrfs_create_pending_block_groups() |
 |- btrfs_insert_item()              |
 |- btrfs_finish_chunk_alloc()       |
    |- btrfs_alloc_dev_extent()      |

After the unification:
               Kernel                |              Btrfs-progs
 btrfs_alloc_chunk()                 | btrfs_alloc_chunk()
 |- __btrfs_alloc_chunk()            | |- __btrfs_alloc_chunk()
                                     | |- btrfs_finish_chunk_alloc()
 btrfs_create_pending_block_groups() |    |- btrfs_alloc_dev_extent()
 |- btrfs_insert_item()              |
 |- btrfs_finish_chunk_alloc()       |

And the similiar functions are share the same code base, with minor
member/functions change.

This update only modifies patches 7 and after.

Changelog:
v2:
  Make error handler in patch 7 better.
  New patches to unify more functions used in btrfs_alloc_chunk()

Qu Wenruo (10):
  btrfs-progs: Refactor parameter of BTRFS_MAX_DEVS() from root to
    fs_info
  btrfs-progs: Merge btrfs_alloc_data_chunk into btrfs_alloc_chunk
  btrfs-progs: Make btrfs_alloc_chunk to handle block group creation
  btrfs-progs: Introduce btrfs_raid_array and related infrastructures
  btrfs-progs: volumes: Allow find_free_dev_extent() to return maximum
    hole size
  btrfs-progs: kernel-lib: Port kernel sort() to btrfs-progs
  btrfs-progs: volumes: Unify free dev extent search behavior between
    kernel and btrfs-progs.
  btrfs-progs: Move chunk stripe size calcution function to volumes.h
  btrfs-progs: Use btrfs_device->fs_info to replace
    btrfs_device->dev_root
  btrfs-progs: Refactor btrfs_alloc_chunk to mimic kernel structure and
    behavior

 Makefile          |   3 +-
 check/main.c      |  22 --
 convert/main.c    |  10 +-
 ctree.h           |  12 +-
 extent-tree.c     |  12 +-
 kerncompat.h      |   5 +
 kernel-lib/sort.c | 104 ++++++
 kernel-lib/sort.h |  16 +
 mkfs/main.c       |  27 +-
 utils.c           |   2 +-
 volumes.c         | 927 ++++++++++++++++++++++++++++--------------------------
 volumes.h         |  66 +++-
 12 files changed, 695 insertions(+), 511 deletions(-)
 create mode 100644 kernel-lib/sort.c
 create mode 100644 kernel-lib/sort.h

-- 
2.16.1


^ permalink raw reply	[flat|nested] 24+ messages in thread
* [PATCH v2] btrfs-progs: ctree: Add extra level check for read_node_slot()
@ 2018-02-08  0:59 Qu Wenruo
  2018-02-08  7:38 ` Nikolay Borisov
  2018-03-19 12:36 ` David Sterba
  0 siblings, 2 replies; 24+ messages in thread
From: Qu Wenruo @ 2018-02-08  0:59 UTC (permalink / raw)
  To: linux-btrfs, dsterba; +Cc: ralphgauges

Strangely, we have level check in btrfs_print_tree() while we don't have
the same check in read_node_slot().

That's to say, for the following corruption, btrfs_search_slot() or
btrfs_next_leaf() can return invalid leaf:

Parent eb:
  node XXXXXX level 1
              ^^^^^^^
              Child should be leaf (level 0)
  ...
  key (XXX XXX XXX) block YYYYYY

Child eb:
  leaf YYYYYY level 1
              ^^^^^^^
              Something went wrong now

And for the corrupted leaf returned, later caller can be screwed up
easily.

Although the root cause (powerloss, but still something wrong breaking
metadata CoW of btrfs) is still unknown, at least enhance btrfs-progs to
avoid SEGV.

Reported-by: Ralph Gauges <ralphgauges@googlemail.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
---
changlog:
v2:
  Check if the extent buffer is up-to-date before checking its level to
  avoid possible NULL pointer access.
---
 ctree.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/ctree.c b/ctree.c
index 4fc33b14000a..430805e3043f 100644
--- a/ctree.c
+++ b/ctree.c
@@ -22,6 +22,7 @@
 #include "repair.h"
 #include "internal.h"
 #include "sizes.h"
+#include "messages.h"
 
 static int split_node(struct btrfs_trans_handle *trans, struct btrfs_root
 		      *root, struct btrfs_path *path, int level);
@@ -640,7 +641,9 @@ static int bin_search(struct extent_buffer *eb, struct btrfs_key *key,
 struct extent_buffer *read_node_slot(struct btrfs_fs_info *fs_info,
 				   struct extent_buffer *parent, int slot)
 {
+	struct extent_buffer *ret;
 	int level = btrfs_header_level(parent);
+
 	if (slot < 0)
 		return NULL;
 	if (slot >= btrfs_header_nritems(parent))
@@ -649,8 +652,19 @@ struct extent_buffer *read_node_slot(struct btrfs_fs_info *fs_info,
 	if (level == 0)
 		return NULL;
 
-	return read_tree_block(fs_info, btrfs_node_blockptr(parent, slot),
+	ret = read_tree_block(fs_info, btrfs_node_blockptr(parent, slot),
 		       btrfs_node_ptr_generation(parent, slot));
+	if (!extent_buffer_uptodate(ret))
+		return ERR_PTR(-EIO);
+
+	if (btrfs_header_level(ret) != level - 1) {
+		error("child eb corrupted: parent bytenr=%llu item=%d parent level=%d child level=%d",
+		      btrfs_header_bytenr(parent), slot,
+		      btrfs_header_level(parent), btrfs_header_level(ret));
+		free_extent_buffer(ret);
+		return ERR_PTR(-EIO);
+	}
+	return ret;
 }
 
 static int balance_level(struct btrfs_trans_handle *trans,
-- 
2.16.1


^ permalink raw reply related	[flat|nested] 24+ messages in thread

end of thread, other threads:[~2018-03-20  5:24 UTC | newest]

Thread overview: 24+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-02-09  7:44 [PATCH v2 00/10] Chunk allocator unification Qu Wenruo
2018-02-09  7:44 ` [PATCH v2] btrfs-progs: ctree: Add extra level check for read_node_slot() Qu Wenruo
2018-02-09  8:09   ` Nikolay Borisov
2018-02-09 10:02     ` Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 01/10] btrfs-progs: Refactor parameter of BTRFS_MAX_DEVS() from root to fs_info Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 02/10] btrfs-progs: Merge btrfs_alloc_data_chunk into btrfs_alloc_chunk Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 03/10] btrfs-progs: Make btrfs_alloc_chunk to handle block group creation Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 04/10] btrfs-progs: Introduce btrfs_raid_array and related infrastructures Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 05/10] btrfs-progs: volumes: Allow find_free_dev_extent() to return maximum hole size Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 06/10] btrfs-progs: kernel-lib: Port kernel sort() to btrfs-progs Qu Wenruo
2018-02-21 15:40   ` David Sterba
2018-02-22  1:43     ` Qu Wenruo
2018-03-09 18:14       ` David Sterba
2018-02-09  7:44 ` [PATCH v2 07/10] btrfs-progs: volumes: Unify free dev extent search behavior between kernel and btrfs-progs Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 08/10] btrfs-progs: Move chunk stripe size calcution function to volumes.h Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 09/10] btrfs-progs: Use btrfs_device->fs_info to replace btrfs_device->dev_root Qu Wenruo
2018-02-09  7:44 ` [PATCH v2 10/10] btrfs-progs: Refactor btrfs_alloc_chunk to mimic kernel structure and behavior Qu Wenruo
2018-02-09  8:17   ` Nikolay Borisov
2018-02-09 10:01     ` Qu Wenruo
  -- strict thread matches above, loose matches on Subject: below --
2018-02-08  0:59 [PATCH v2] btrfs-progs: ctree: Add extra level check for read_node_slot() Qu Wenruo
2018-02-08  7:38 ` Nikolay Borisov
2018-03-19 12:36 ` David Sterba
2018-03-20  5:16   ` Qu Wenruo
2018-03-20  5:23     ` Qu Wenruo

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).