Skip to contentSkip to conversation

Btrfs fixes for 5.13-rc5

GIT PULL Merged David Sterba 4 participants 4 messages Jun 3, 2021 – Jun 4, 2021 Btrfs file system

lore
Share & tools
lore

David Sterba

author maintainer
#0 lore raw

Hi,

this is a batch from last week, I wanted to give it more testing because last pull request introduced a bug, interacting zoned and subpage features. Otherwise there are error handling improvements and bug fixes. The last commit is from today, adding IRC link to maintainers file.

Please pull, thanks.

- error handling improvements, caught by error injection
- handle errors during checksum deletion
- set error on mapping when ordered extent io cannot be finished
- inode link count fixup in tree-log
- missing return value checks for inode updates in tree-log
- abort transaction in rename exchange if adding second reference fails

- fixes
- fix fsync failure after writes to prealloc extents
- fix deadlock when cloning inline extents and low on available space
- fix compressed writes that cross stripe boundary

----------------------------------------------------------------
The following changes since commit 764c7c9a464b68f7c6a5a9ec0b923176a05e8e8f:

btrfs: zoned: fix parallel compressed writes (2021-05-20 15:51:07 +0200)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux.git for-5.13-rc4-tag

for you to fetch changes up to 503d1acb01826b42e5afb496dfcc32751bec9478:

MAINTAINERS: add btrfs IRC link (2021-06-03 15:40:38 +0200)

----------------------------------------------------------------
David Sterba (1):
MAINTAINERS: add btrfs IRC link

Filipe Manana (2):
btrfs: fix fsync failure and transaction abort after writes to prealloc extents
btrfs: fix deadlock when cloning inline extents and low on available space

Josef Bacik (6):
btrfs: fix error handling in btrfs_del_csums
btrfs: return errors from btrfs_del_csums in cleanup_ref_head
btrfs: mark ordered extent and inode with error if we fail to finish
btrfs: fixup error handling in fixup_inode_link_counts
btrfs: check error value from btrfs_update_inode in tree log
btrfs: abort in rename_exchange if we fail to insert the second ref

Qu Wenruo (1):
btrfs: fix compressed writes that cross stripe boundary

 MAINTAINERS            |   1 +
 fs/btrfs/compression.c |  17 +++++---
 fs/btrfs/extent-tree.c |   2 +-
 fs/btrfs/file-item.c   | 108 ++++++++++++++++++++++++++++++++++++-------------
 fs/btrfs/inode.c       |  19 ++++++++-
 fs/btrfs/reflink.c     |  38 +++++++++--------
 fs/btrfs/tree-log.c    |  21 ++++++----
 7 files changed, 148 insertions(+), 58 deletions(-)
#2 lore raw
In reply to David Sterba
Quoted 17 lines · David Sterba

Hi,

this is a batch from last week, I wanted to give it more testing because last pull request introduced a bug, interacting zoned and subpage features. Otherwise there are error handling improvements and bug fixes. The last commit is from today, adding IRC link to maintainers file.

Please pull, thanks.

- error handling improvements, caught by error injection
- handle errors during checksum deletion
- set error on mapping when ordered extent io cannot be finished
- inode link count fixup in tree-log
- missing return value checks for inode updates in tree-log
- abort transaction in rename exchange if adding second reference fails

Quoted 5 lines

- fixes
- fix fsync failure after writes to prealloc extents
- fix deadlock when cloning inline extents and low on available space
- fix compressed writes that cross stripe boundary

David,

Could you also add commit "btrfs: zoned: fix zone number to sector/physical calculation" for pull? Without this commit, on a device larger than 4 TB, zoned btrfs will overwrite the primary superblock with the 2nd copy and causes a mount failure after the first mount/umount.

https://lore.kernel.org/linux-btrfs/[email protected]/

Thanks,

#3 lore raw
Quoted 5 lines · Naohiro Aota

Could you also add commit "btrfs: zoned: fix zone number to sector/physical calculation" for pull? Without this commit, on a device larger than 4 TB, zoned btrfs will overwrite the primary superblock with the 2nd copy and causes a mount failure after the first mount/umount.

Yeah, a few more fixes plus that one are in queue for the next pull.