Skip to contentSkip to conversation

bcache changes for 3.13

GIT PULL Kent Overstreet 2 participants 4 messages Nov 4, 2013 bcache (block layer cache)

lore
Share & tools
lore
#0 lore raw

Hey Jens, sorry for being late with this - anyways, it's roughly the same set of patches you had queued up before plus a few minor fixes, but it's been rebased onto your for-3.13/core branch.

The following changes since commit febca1baea1cfe2d7a0271385d89b03d5fb34f94:

block: setup bi_vcnt on clones (2013-10-31 13:32:42 -0600)

are available in the git repository at:

git://evilpiepirate.org/~kent/linux-bcache.git bcache-for-3.13

for you to fetch changes up to 0bb953dba5d295b03bb035bfc2d28ed4155df377:

bcache: defensively handle format strings (2013-11-01 18:03:20 -0700)

----------------------------------------------------------------
Kees Cook (1):
bcache: defensively handle format strings

Kent Overstreet (53):
bcache: Fixed incorrect order of arguments to bio_alloc_bioset()
bcache: Convert bch_data_verify() to immutable biovecs
bcache: Fix a journalling performance bug
bcache: Fix a lockdep splat
bcache: Use blkdev_issue_discard()
bcache: Add on error panic/unregister setting
bcache: Stripe size isn't necessarily a power of two
bcache: Remove unnecessary check in should_split()
bcache: Explicitly track btree node's parent
bcache: Add btree_insert_node()
bcache: Insert multiple keys at a time
bcache: Convert btree_insert_check_key() to btree_insert_node()
bcache: Add explicit keylist arg to btree_insert()
bcache: Clean up keylist code
bcache: Refactor request_write()
bcache: Refactor read request code a bit
bcache: Refactor journalling flow control
bcache: Move keylist out of btree_op
bcache: Convert try_wait to wait_queue_head_t
bcache: Convert bucket_wait to wait_queue_head_t
bcache: Convert gc to a kthread
bcache: Convert writeback to a kthread
bcache: Add btree_map() functions
bcache: Move some stuff to btree.c
bcache: Convert bch_btree_read_async() to bch_btree_map_keys()
bcache: Clean up cache_lookup_fn
bcache: Prune struct btree_op
bcache: Kill op->cl
bcache: Drop some closure stuff
bcache: Kill op->replace
bcache: Don't use op->insert_collision
bcache: Convert bch_btree_insert() to bch_btree_map_leaf_nodes()
bcache: Break up struct search
bcache: Move sector allocator to alloc.c
bcache: Pull on disk data structures out into a separate header
bcache: Fix bch_ptr_bad()
bcache: Debug code improvements
bcache: Don't bother with bucket refcount for btree node allocations
bcache: bch_(btree|extent)_ptr_invalid()
bcache: PRECEDING_KEY()
bcache: Add btree_node_write_sync()
bcache: Add make_btree_freeing_key()
bcache: Incremental gc
bcache: Avoid deadlocking in garbage collection
bcache: Kill bch_next_recurse_key()
bcache: Kill sequential_merge option
bcache: Move spinlock into struct time_stats
bcache: Have btree_split() insert into parent directly
bcache: Better full stripe scanning
bcache: Fix sysfs splat on shutdown with flash only devs
bcache: Use ida for bcache block dev minor
bcache: Delete some slower inline asm
bcache: Bypass torture test

 drivers/md/bcache/Kconfig     |   11 +-
 drivers/md/bcache/alloc.c     |  383 +++++++-----
 drivers/md/bcache/bcache.h    |  327 +---------
 drivers/md/bcache/bset.c      |  289 ++++-----
 drivers/md/bcache/bset.h      |   93 +--
 drivers/md/bcache/btree.c     | 1386 ++++++++++++++++++++++-------------------
 drivers/md/bcache/btree.h     |  195 ++----
 drivers/md/bcache/closure.c   |  103 +--
 drivers/md/bcache/closure.h   |  183 +-----
 drivers/md/bcache/debug.c     |  202 +++---
 drivers/md/bcache/debug.h     |   50 +-
 drivers/md/bcache/journal.c   |  293 +++++----
 drivers/md/bcache/journal.h   |   52 +-
 drivers/md/bcache/movinggc.c  |   85 ++-
 drivers/md/bcache/request.c   | 1106 ++++++++++++++++----------------
 drivers/md/bcache/request.h   |   42 +-
 drivers/md/bcache/stats.c     |   26 +-
 drivers/md/bcache/stats.h     |   13 +-
 drivers/md/bcache/super.c     |  190 +++---
 drivers/md/bcache/sysfs.c     |   42 +-
 drivers/md/bcache/trace.c     |    1 -
 drivers/md/bcache/util.c      |   12 +-
 drivers/md/bcache/util.h      |   15 +-
 drivers/md/bcache/writeback.c |  455 +++++++-------
 drivers/md/bcache/writeback.h |   46 +-
 include/trace/events/bcache.h |   47 +-
 include/uapi/linux/bcache.h   |  373 +++++++++++
 27 files changed, 3006 insertions(+), 3014 deletions(-)
 create mode 100644 include/uapi/linux/bcache.h
#1 lore raw
Quoted 3 lines · Kent Overstreet

Hey Jens, sorry for being late with this - anyways, it's roughly the same set of patches you had queued up before plus a few minor fixes, but it's been rebased onto your for-3.13/core branch.

Merge window is a little later this time, and since it was already in for the previous release, we can make it happen.

But can you please base it on for-3.13/drivers however? That's where I'd like to pull in the driver bits, and if I pull this one, then things get even more tangled since for-3.13/core is holding the blk-mq changes too.

Signature · 1 line
Jens Axboe
#2 lore raw
Quoted 11 lines · Jens Axboe
Quoted 3 lines · Kent Overstreet

Hey Jens, sorry for being late with this - anyways, it's roughly the same set of patches you had queued up before plus a few minor fixes, but it's been rebased onto your for-3.13/core branch.

Merge window is a little later this time, and since it was already in for the previous release, we can make it happen.

But can you please base it on for-3.13/drivers however? That's where I'd like to pull in the driver bits, and if I pull this one, then things get even more tangled since for-3.13/core is holding the blk-mq changes too.

I'd like to do the merge with immutable biovecs myself though - there's some tricky bits in there. How will that work?

#3 lore raw
Quoted 9 lines · Kent Overstreet
Quoted 8 lines · Jens Axboe
Quoted 3 lines · Kent Overstreet

Hey Jens, sorry for being late with this - anyways, it's roughly the same set of patches you had queued up before plus a few minor fixes, but it's been rebased onto your for-3.13/core branch.

Merge window is a little later this time, and since it was already in for the previous release, we can make it happen.

Quoted 6 lines
Quoted 3 lines

But can you please base it on for-3.13/drivers however? That's where I'd like to pull in the driver bits, and if I pull this one, then things get even more tangled since for-3.13/core is holding the blk-mq changes too.

I'd like to do the merge with immutable biovecs myself though - there's some tricky bits in there. How will that work?

OK, in that case, yeah it might be easier to just pull it in wholesale.

Signature · 1 line
Jens Axboe