Nothing terribly exciting, things have been stabilizing nicely...
the btree_trans_peek_key_cache fix finishes fixing an oops that I've had
reports of from a few people
'journal stuck timeout' change addresses another source of SRCU lock
hold time warnings; the main remaining one? I'm not seeing it in the CI
anymore with that fix.
'improve journal pin flushing' addresses slow & hanging unmounts
'trace_move_extent_finish' improvement is to help us debug the
'rebalance not making progress' issues
Kent Overstreet (11):
bcachefs: Fix btree_trans_peek_key_cache()
bcachefs: Reduce stack frame size of __bch2_str_hash_check_key()
bcachefs: "Journal stuck" timeout now takes into account device
latency
bcachefs: bset_blacklisted_journal_seq is now AUTOFIX
bcachefs: Improve decompression error messages
bcachefs: rebalance, copygc enabled are runtime opts
bcachefs: fix bch2_btree_node_flags
bcachefs: Improve journal pin flushing
bcachefs: Journal writes are now IOPRIO_CLASS_RT
bcachefs: Fix trace_copygc
bcachefs: Improve trace_move_extent_finish
fs/bcachefs/btree_cache.c | 5 +-
fs/bcachefs/btree_iter.c | 3 +-
fs/bcachefs/btree_key_cache.c | 4 +-
fs/bcachefs/btree_trans_commit.c | 2 +-
fs/bcachefs/compress.c | 31 +++++--
fs/bcachefs/compress.h | 4 +-
fs/bcachefs/data_update.c | 50 +++++++----
fs/bcachefs/debug.c | 1 +
fs/bcachefs/io_write.c | 4 +-
fs/bcachefs/io_write.h | 2 +
fs/bcachefs/journal.c | 92 +++++++-------------
fs/bcachefs/journal.h | 9 +-
fs/bcachefs/journal_io.c | 2 +
fs/bcachefs/journal_reclaim.c | 142 ++++++++++++++++++++++++++-----
fs/bcachefs/journal_reclaim.h | 3 +
fs/bcachefs/journal_types.h | 13 +--
fs/bcachefs/movinggc.c | 11 ++-
fs/bcachefs/opts.h | 4 +-
fs/bcachefs/sb-errors_format.h | 2 +-
fs/bcachefs/str_hash.c | 24 +++---
fs/bcachefs/trace.h | 26 +++---
21 files changed, 275 insertions(+), 159 deletions(-)
--
2.45.2
