Hi Linus,
Please pull MD changes for 4.10. This update includes:
- A raid5 writeback cache feature. The goal is to aggregate writes to make full
stripe write and reduce read-modify-write. It's helpful for workload which
does sequential write and follows fsync for example. This feature is
experimental and off by default right now.
- FAILFAST support. This fails IOs to broken raid disks quickly, so can improve
latency. It's mainly for DASD storage, but some patches help normal raid
array too.
- Support bad block for raid array with external metadata
- AVX2 instruction support for raid6 parity calculation
- Normalize MD info output
- Add missing blktrace
- Other bug fixes
Thanks,
Shaohua
The following changes since commit b78b499a67c3f77aeb6cd0b54724bc38b141255d:
Merge tag 'char-misc-4.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc (2016-12-13 12:11:01 -0800)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/shli/md.git for-linus
for you to fetch changes up to 20737738d397dfadbca1ea50dcc00d7259f500cf:
Merge branch 'md-next' into md-linus (2016-12-13 12:40:15 -0800)
----------------------------------------------------------------
Dan Carpenter (1):
md/r5cache: enable IRQs on error path
Gayatri Kammela (1):
lib/raid6: Add AVX2 optimized xor_syndrome functions
Guoqing Jiang (1):
md/bitmap: call bitmap_file_unmap once bitmap_storage_alloc returns -ENOMEM
JackieLiu (8):
raid5-cache: restrict the use area of the log_offset variable
md/raid5-cache: remove unnecessary function parameters
md/raid5-cache: use ring add to prevent overflow
md/raid5-cache: release the stripe_head at the appropriate location
md/raid5-cache: remove the unnecessary next_cp_seq field from the r5l_log
md/raid5-cache: do not need to set STRIPE_PREREAD_ACTIVE repeatedly
md/raid5-cache: adjust the write position of the empty block if no data blocks
md/raid5-cache: no recovery is required when create super-block
Konstantin Khlebnikov (1):
md/raid5: limit request size according to implementation limits
NeilBrown (26):
md: fix some issues with alloc_disk_sb()
md: change all printk() to pr_err() or pr_warn() etc.
md/bitmap: change all printk() to pr_*()
md/linear: replace printk() with pr_*()
md/multipath: replace printk() with pr_*()
md/raid0: replace printk() with pr_*()
md/raid1: change printk() to pr_*()
md/raid10: change printk() to pr_*()
md/raid5: change printk() to pr_*()
md: perform async updates for metadata where possible.
md/raid1: abort delayed writes when device fails.
md/raid10: abort delayed writes when device fails.
md/bitmap: Don't write bitmap while earlier writes might be in-flight
md/raid1: fix: IO can block resync indefinitely
md: define mddev flags, recovery flags and r1bio state bits using enums
md: remove md_super_wait() call after bitmap_flush()
md: add block tracing for bio_remapping
md/bitmap: add blktrace event for writes to the bitmap
md/raid1, raid10: add blktrace records when IO is delayed
md/failfast: add failfast flag for md to be used by some personalities.
md: Use REQ_FAILFAST_* on metadata writes where appropriate
md/raid1: add failfast handling for reads.
md/raid1: add failfast handling for writes.
md/raid10: add failfast handling for reads.
md/raid10: add failfast handling for writes.
md: fix refcount problem on mddev when stopping array.
Shaohua Li (8):
raid5-cache: fix lockdep warning
md: add blktrace event for writes to superblock
raid5-cache: suspend reclaim thread instead of shutdown
md: stop write should stop journal reclaim
md: takeover should clear unrelated bits
md: MD_RECOVERY_NEEDED is set for mddev->recovery
md: separate flags for superblock changes
Merge branch 'md-next' into md-linus
Song Liu (14):
md/r5cache: Check array size in r5l_init_log
md/r5cache: move some code to raid5.h
md/r5cache: State machine for raid5-cache write back mode
md/r5cache: caching phase of r5cache
md/r5cache: write-out phase and reclaim support
md/r5cache: sysfs entry journal_mode
md/r5cache: refactoring journal recovery code
md/r5cache: r5cache recovery: part 1
md/r5cache: r5cache recovery: part 2
md/r5cache: handle FLUSH and FUA
md/r5cache: handle alloc_page failure
md/r5cache: run_no_space_stripes() when R5C_LOG_CRITICAL == 0
md/raid5-cache: fix crc in rewrite_data_only_stripes()
md/r5cache: after recovery, increase journal seq by 10000
Tomasz Majchrzak (4):
md: add bad block support for external metadata
md: don't fail an array if there are unacknowledged bad blocks
md: wake up personality thread after array state update
raid5: revert commit 11367799f3d1
Zhengyuan Liu (3):
raid5-cache: add another check conditon before replaying one stripe
raid5-cache: don't set STRIPE_R5C_PARTIAL_STRIPE flag while load stripe into cache
md/r5cache: do r5c_update_log_state after log recovery
drivers/md/bitmap.c | 166 ++--
drivers/md/dm-raid.c | 4 +-
drivers/md/linear.c | 31 +-
drivers/md/md.c | 701 ++++++++-------
drivers/md/md.h | 108 ++-
drivers/md/multipath.c | 92 +-
drivers/md/raid0.c | 107 ++-
drivers/md/raid1.c | 247 ++++--
drivers/md/raid1.h | 19 +-
drivers/md/raid10.c | 295 ++++---
drivers/md/raid10.h | 2 +
drivers/md/raid5-cache.c | 1885 +++++++++++++++++++++++++++++++++++-----
drivers/md/raid5.c | 623 +++++++------
drivers/md/raid5.h | 172 +++-
include/uapi/linux/raid/md_p.h | 7 +-
lib/raid6/avx2.c | 232 ++++-
16 files changed, 3429 insertions(+), 1262 deletions(-)