Skip to contentSkip to conversation

Btrfs fixes for 4.15-rc2

GIT PULL David Sterba 4 participants 5 messages Nov 29, 2017 – Feb 18, 2019 Btrfs file system

lore
Share & tools
lore

David Sterba

author maintainer
#0 lore raw

Hi,

we've collected some fixes in since the pre-merge window freeze. There's technically only one regression fix for 4.15, but the rest seems important and candidates for stable. No merge conflicts, please pull, thanks.

- fix missing flush bio puts in error cases (is serious, but rarely happens)

- fix reporting stat::st_blocks for buffered append writes

- fix space cache invalidation

- fix out of bound memory access when setting zlib level

- fix potential memory corruption when fsync fails in the middle

- fix crash in integrity checker

- incremetnal send fix, path mixup for certain unlink/rename combination

- pass flags to writeback so compressed writes can be throttled properly

- error handling fixes

With signed tag: for-4.15-rc2-tag

----------------------------------------------------------------
The following changes since commit d28e649a5c58b779b303c252c66ee84a0f2c3b32:

btrfs: Fix bug for misused dev_t when lookup in dev state hash table. (2017-11-01 20:45:36 +0100)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux.git for-4.15-rc2

for you to fetch changes up to ea37d5998b50a72b9045ba60a132eeb20e1c4230:

Btrfs: incremental send, fix wrong unlink path after renaming file (2017-11-28 17:15:30 +0100)

----------------------------------------------------------------
David Sterba (2):
btrfs: add missing device::flush_bio puts
btrfs: dev_alloc_list is not protected by RCU, use normal list_del

Filipe Manana (3):
Btrfs: move definition of the function btrfs_find_new_delalloc_bytes
Btrfs: fix reported number of inode blocks after buffered append writes
Btrfs: incremental send, fix wrong unlink path after renaming file

Josef Bacik (2):
btrfs: clear space cache inode generation always
btrfs: fix deadlock when writing out space cache

Liu Bo (3):
Btrfs: add write_flags for compression bio
Btrfs: bail out gracefully rather than BUG_ON
Btrfs: fix list_add corruption and soft lockups in fsync

Nikolay Borisov (1):
btrfs: Fix transaction abort during failure in btrfs_rm_dev_item

Qu Wenruo (2):
btrfs: Fix wild memory access in compression level parser
btrfs: tree-checker: Fix false panic for sanity test

 fs/btrfs/compression.c           |   9 +--
 fs/btrfs/compression.h           |   5 +-
 fs/btrfs/ctree.h                 |   1 +
 fs/btrfs/disk-io.c               |  10 ++-
 fs/btrfs/extent-tree.c           |  14 ++---
 fs/btrfs/extent_io.c             |   2 +-
 fs/btrfs/extent_io.h             |   8 ++-
 fs/btrfs/file.c                  | 130 +++++++++++++++++++++------------------
 fs/btrfs/free-space-cache.c      |   3 +-
 fs/btrfs/inode.c                 |  34 +++++++---
 fs/btrfs/relocation.c            |   3 +-
 fs/btrfs/send.c                  | 124 +++++++++++++++++++++++++++++++------
 fs/btrfs/super.c                 |  13 +++-
 fs/btrfs/tests/extent-io-tests.c |   6 +-
 fs/btrfs/tests/inode-tests.c     |  12 ++--
 fs/btrfs/tree-checker.c          |  27 ++++++--
 fs/btrfs/tree-checker.h          |  14 ++++-
 fs/btrfs/tree-log.c              |   2 +-
 fs/btrfs/volumes.c               |  32 +++++++---
 19 files changed, 314 insertions(+), 135 deletions(-)
#1 lore raw
Quoted 4 lines · David Sterba

With signed tag: for-4.15-rc2-tag

git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux.git for-4.15-rc2

Oh, please actually ask me to pull the signed tag (exact same pull-request, just point git request-pull at the tag), because now what happened was that first I just pulled that branch you mentioned, and only noticed that "With signed tag:" notice after I had already pulled and was filling in the merge message.

Anyway, I redid the pull with the proper signed tag, but it was just annoying extra work.

And I wonder how many times I _hadn't_ noticed that, because I didn't have your key in my keyring either. Or maybe I caught it the first time.

Linus

#2 lore raw
Quoted 8 lines · Linus Torvalds
Quoted 4 lines · David Sterba

With signed tag: for-4.15-rc2-tag

git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux.git for-4.15-rc2

Oh, please actually ask me to pull the signed tag (exact same pull-request, just point git request-pull at the tag),

Will do next time.

Quoted 11 lines

because now
what happened was that first I just pulled that branch you mentioned, and only noticed that "With signed tag:" notice after I had already pulled and was filling in the merge message.

Anyway, I redid the pull with the proper signed tag, but it was just annoying extra work.

And I wonder how many times I _hadn't_ noticed that, because I didn't have your key in my keyring either. Or maybe I caught it the first time.

All my previous pull requests were like that. I did the split branch/tag beacuse the ambiguous name for branch and brings some hassle to push or remove them. I thought a separate tag with same top commit as the branch plus mentioning the tag in the mail would be enough to verify the pulled branch. But apparently was not, sorry for the trouble.

#3 lore raw
In reply to David Sterba

Hi David,

Quoted 5 lines

Btrfs: incremental send, fix wrong unlink path after renaming file (2017-11-28 17:15:30 +0100)

----------------------------------------------------------------
David Sterba (2):
btrfs: add missing device::flush_bio puts

Is there a reason that this one should not be tagged as "stable"? At least the missing bio_put in free_fs_devices is not an error case, and would happen every time.

Thanks,
Alex.

#4 lore raw
Quoted 10 lines · Alex Lyakas

Hi David,

Quoted 5 lines

Btrfs: incremental send, fix wrong unlink path after renaming file (2017-11-28 17:15:30 +0100)

----------------------------------------------------------------
David Sterba (2):
btrfs: add missing device::flush_bio puts

Is there a reason that this one should not be tagged as "stable"? At least the missing bio_put in free_fs_devices is not an error case, and would happen every time.

2 years old, hard to remember if there was a reason or not. At that time, not all patches got cc:stable because patches with Fixes: tag were usually picked by stable team and that's what we relied on, as long as the patch applied. This worked and still works, though I've started to do more thorough review of various fixes against stable trees and add the tag eventually.

The fix applies only to 4.14 stable, so it could be submitted there for inclusion.