Linux RAID subsystem development
 help / color / mirror / Atom feed
From: Yu Kuai <yukuai@fnnas.com>
To: linux-raid@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, Li Nan <linan122@huawei.com>,
	Yu Kuai <yukuai@fnnas.com>, Cheng Cheng <chencheng@fnnas.com>
Subject: [PATCH 00/19] md: support llbitmap reshape for raid10 and raid5
Date: Sun, 19 Apr 2026 11:09:23 +0800	[thread overview]
Message-ID: <20260419030942.824195-1-yukuai@fnnas.com> (raw)

llbitmap currently tracks one live bitmap geometry, but reshape needs old and
new mapping to coexist until reshape completion. That breaks normal bitmap
accounting in three places:

- bios that cross reshape_position cannot be accounted as one bitmap range
- checkpointed dirty state must move forward as reshape advances
- interrupted reshape must resume with the same llbitmap interpretation after
  reassembly

The main pieces are:

- add bitmap-side range preparation so md core uses one bitmap API
- split reshape-crossing bios in raid10/raid5 before bitmap accounting
- let llbitmap stage target geometry and remap checkpointed bits in one bitmap
- teach raid10/raid5 to provide exact old/new mapping and drive llbitmap
  reshape progress

Validation

Runtime:
- custom QEMU llbitmap reshape matrix passed for raid10 and raid5:
  - expanded/changed bit-state checks with query_bit
  - cross-boundary writes during reshape
  - heavier raid5 write path that previously warned/crashed
  - degraded stop/assemble data verification after reshape
  - interrupted reshape stop/assemble/continue for raid10 and raid5
- mdadm test 02r5grow passed in disk mode

The temporary query_bit validation hook is intentionally kept out of this
submission series.

Yu Kuai (19):
  md: add exact bitmap mapping and reshape hooks
  md: add helper to split bios at reshape offset
  md/md-llbitmap: track bitmap sync_size explicitly
  md/md-llbitmap: allocate page controls independently
  md/md-llbitmap: grow the page cache in place for reshape
  md/md-llbitmap: track target reshape geometry fields
  md/md-llbitmap: finish reshape geometry
  md/md-llbitmap: refuse reshape while llbitmap still needs sync
  md/md-llbitmap: add reshape range mapping helpers
  md/md-llbitmap: don't skip reshape ranges from bitmap state
  md/md-llbitmap: remap checkpointed bits as reshape progresses
  md/md-llbitmap: clamp state-machine walks to tracked bits
  md/raid10: reject llbitmap chunk shrink during reshape
  md/raid10: wire llbitmap reshape lifecycle
  md/raid10: split reshape bios before bitmap accounting
  md/raid5: add exact old and new llbitmap mapping helpers
  md/raid5: reject llbitmap chunk shrink during reshape
  md/raid5: wire llbitmap reshape lifecycle
  md/raid5: split reshape bios before bitmap accounting

 drivers/md/md-bitmap.c   |   8 +
 drivers/md/md-bitmap.h   |   7 +
 drivers/md/md-llbitmap.c | 602 +++++++++++++++++++++++++++++++++++---
 drivers/md/md.c          |  51 +++
 drivers/md/md.h          |   8 +
 drivers/md/raid10.c      |  44 ++-
 drivers/md/raid5.c       | 114 ++++++--
 7 files changed, 760 insertions(+), 74 deletions(-)

-- 
2.51.0

             reply	other threads:[~2026-04-19  3:09 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-19  3:09 Yu Kuai [this message]
2026-04-19  3:09 ` [PATCH] md: add exact bitmap mapping and reshape hooks Yu Kuai
2026-04-19  3:09 ` [PATCH] md: add helper to split bios at reshape offset Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: track bitmap sync_size explicitly Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: allocate page controls independently Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: grow the page cache in place for reshape Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: track target reshape geometry fields Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: finish reshape geometry Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: refuse reshape while llbitmap still needs sync Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: add reshape range mapping helpers Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: don't skip reshape ranges from bitmap state Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: remap checkpointed bits as reshape progresses Yu Kuai
2026-04-19  3:09 ` [PATCH] md/md-llbitmap: clamp state-machine walks to tracked bits Yu Kuai
2026-04-19  3:09 ` [PATCH] md/raid10: reject llbitmap chunk shrink during reshape Yu Kuai
2026-04-19  3:09 ` [PATCH] md/raid10: wire llbitmap reshape lifecycle Yu Kuai
2026-04-30  2:37   ` kernel test robot
2026-04-19  3:09 ` [PATCH] md/raid10: split reshape bios before bitmap accounting Yu Kuai
2026-04-19  3:09 ` [PATCH] md/raid5: add exact old and new llbitmap mapping helpers Yu Kuai
2026-05-01 18:51   ` kernel test robot
2026-04-19  3:09 ` [PATCH] md/raid5: reject llbitmap chunk shrink during reshape Yu Kuai
2026-04-19  3:09 ` [PATCH] md/raid5: wire llbitmap reshape lifecycle Yu Kuai
2026-04-19  3:09 ` [PATCH] md/raid5: split reshape bios before bitmap accounting Yu Kuai
2026-04-30  0:59   ` kernel test robot
2026-04-30  4:07   ` kernel test robot
2026-04-30 19:48   ` kernel test robot

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260419030942.824195-1-yukuai@fnnas.com \
    --to=yukuai@fnnas.com \
    --cc=chencheng@fnnas.com \
    --cc=linan122@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-raid@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox