kernel/linux.git/drivers/md/md.c, branch v6.18.43

md/raid1: honor REQ_NOWAIT when waiting for behind writes

2026-07-24T14:16:32+00:00

[ Upstream commit a286cb88ddb26c5f4377859d8e77233d9181eb82 ] raid1 supports REQ_NOWAIT reads by avoiding waits in the barrier path through wait_read_barrier(). However, a read can still block on a WriteMostly device when the array uses a bitmap and there are outstanding behind writes. In that case raid1 unconditionally calls wait_behind_writes(), which may sleep until all behind writes complete. As a result, a REQ_NOWAIT read can block despite the caller explicitly requesting non-blocking behavior. This ensures that raid1 consistently honors REQ_NOWAIT reads across all paths that may otherwise wait for behind writes. Fixes: 5aa705039c4f ("md: raid1 add nowait support") Signed-off-by: Abd-Alrhman Masalkhi Link: https://patch.msgid.link/20260611083514.754922-1-abd.masalkhi@gmail.com Signed-off-by: Yu Kuai Signed-off-by: Sasha Levin

md: merge mddev serialize_policy into mddev_flags

2026-07-24T14:16:32+00:00

[ Upstream commit 10787568cc1f3f80afc510b2728751989dfa0ae6 ] There is not need to use a separate field in struct mddev, there are no functional changes. Link: https://lore.kernel.org/linux-raid/20260114171241.3043364-5-yukuai@fnnas.com Signed-off-by: Yu Kuai Reviewed-by: Li Nan Stable-dep-of: a286cb88ddb2 ("md/raid1: honor REQ_NOWAIT when waiting for behind writes") Signed-off-by: Sasha Levin

md: merge mddev faillast_dev into mddev_flags

2026-07-24T14:16:32+00:00

[ Upstream commit 4f6d2e648cbe963b328cb8815290676da3866434 ] There is not need to use a separate field in struct mddev, there are no functional changes. Link: https://lore.kernel.org/linux-raid/20260114171241.3043364-4-yukuai@fnnas.com Signed-off-by: Yu Kuai Reviewed-by: Li Nan Stable-dep-of: a286cb88ddb2 ("md/raid1: honor REQ_NOWAIT when waiting for behind writes") Signed-off-by: Sasha Levin

md: merge mddev has_superblock into mddev_flags

2026-07-24T14:16:32+00:00

[ Upstream commit fba4a980403d2f489bc680dbff7d7d2514e669f9 ] There is not need to use a separate field in struct mddev, there are no functional changes. Link: https://lore.kernel.org/linux-raid/20260114171241.3043364-3-yukuai@fnnas.com Signed-off-by: Yu Kuai Reviewed-by: Li Nan Stable-dep-of: a286cb88ddb2 ("md/raid1: honor REQ_NOWAIT when waiting for behind writes") Signed-off-by: Sasha Levin

md/md-bitmap: add a none backend for bitmap grow

2026-05-23T11:07:10+00:00

[ Upstream commit f2926a533d03fe70d753b512b713e06a2aa174af ] Add a real none bitmap backend that exposes the common bitmap sysfs group and use it to keep bitmap/location available when an array has no bitmap. Then switch the bitmap location sysfs path to move only between none and the classic bitmap backend, using the no-sysfs bitmap helpers while merging or unmerging the internal bitmap sysfs group. This restores mdadm --grow bitmap addition through bitmap/location. Fixes: fb8cc3b0d9db ("md/md-bitmap: delay registration of bitmap_ops until creating bitmap") Reviewed-by: Su Yue Link: https://lore.kernel.org/r/20260425024615.1696892-4-yukuai@fnnas.com Signed-off-by: Yu Kuai Signed-off-by: Sasha Levin

md/md-bitmap: split bitmap sysfs groups

2026-05-23T11:07:10+00:00

[ Upstream commit aba3d6d6cb55c6e1116d1215140559dd7ecdf9a9 ] Split the classic bitmap sysfs files into a common bitmap group with the location attribute and a separate internal bitmap group for the remaining files. At the same time, convert bitmap operations from a single sysfs group to a sysfs group array so backends can share part of their sysfs layout while adding backend-specific attributes separately. Switch the bitmap sysfs helpers to use sysfs_update_groups() for the add and update path, and remove groups in reverse order so shared named groups are unmerged before the last group removes the directory. Also make bitmap operation lookup depend only on the currently selected bitmap id matching the installed backend. This prepares the lookup path for a later registered none backend. Reviewed-by: Su Yue Link: https://lore.kernel.org/r/20260425024615.1696892-3-yukuai@fnnas.com Signed-off-by: Yu Kuai Stable-dep-of: f2926a533d03 ("md/md-bitmap: add a none backend for bitmap grow") Signed-off-by: Sasha Levin

md: factor bitmap creation away from sysfs handling

2026-05-23T11:07:10+00:00

[ Upstream commit 8776d342cf8fa0b98ca5e6fb2d956966fb5ca364 ] Factor bitmap creation and destruction into helpers that do not touch bitmap sysfs registration. This prepares the bitmap sysfs rework so callers such as the sysfs bitmap location path can create or destroy a bitmap backend without coupling that to sysfs group lifetime management. Reviewed-by: Su Yue Link: https://lore.kernel.org/r/20260425024615.1696892-2-yukuai@fnnas.com Signed-off-by: Yu Kuai Stable-dep-of: f2926a533d03 ("md/md-bitmap: add a none backend for bitmap grow") Signed-off-by: Sasha Levin

md: add fallback to correct bitmap_ops on version mismatch

2026-05-23T11:07:10+00:00

[ Upstream commit 09af773650024279a60348e7319d599e6571b15c ] If default bitmap version and on-disk version doesn't match, and mdadm is not the latest version to set bitmap_type, set bitmap_ops based on the disk version. Link: https://lore.kernel.org/linux-raid/20260323054644.3351791-2-yukuai@fnnas.com/ Signed-off-by: Yu Kuai Stable-dep-of: f2926a533d03 ("md/md-bitmap: add a none backend for bitmap grow") Signed-off-by: Sasha Levin

md: wake raid456 reshape waiters before suspend

2026-05-23T11:06:23+00:00

[ Upstream commit cf86bb53b9c92354904a328e947a05ffbfdd1840 ] During raid456 reshape, direct IO across the reshape position can sleep in raid5_make_request() waiting for reshape progress while still holding an active_io reference. If userspace then freezes reshape and writes md/suspend_lo or md/suspend_hi, mddev_suspend() kills active_io and waits for all in-flight IO to drain. This can deadlock: the IO needs reshape progress to continue, but the reshape thread is already frozen, so the active_io reference is never dropped and suspend never completes. raid5_prepare_suspend() already wakes wait_for_reshape for dm-raid. Do the same for normal md suspend when reshape is already interrupted, so waiting raid456 IO can abort, drop its reference, and let suspend finish. The mdadm test tests/25raid456-reshape-deadlock reproduces the hang. Fixes: 714d20150ed8 ("md: add new helpers to suspend/resume array") Link: https://lore.kernel.org/linux-raid/20260327140729.2030564-1-yukuai@fnnas.com/ Signed-off-by: Yu Kuai Signed-off-by: Sasha Levin

md: remove unused static md_wq workqueue

2026-05-23T11:06:23+00:00

[ Upstream commit e4979f4fac4d6bbe757be50441b45e28e6bf7360 ] The md_wq workqueue is defined as static and initialized in md_init(), but it is not used anywhere within md.c. All asynchronous and deferred work in this file is handled via md_misc_wq or dedicated md threads. Fixes: b75197e86e6d3 ("md: Remove flush handling") Signed-off-by: Abd-Alrhman Masalkhi Link: https://lore.kernel.org/linux-raid/20260328193522.3624-1-abd.masalkhi@gmail.com/ Signed-off-by: Yu Kuai Signed-off-by: Sasha Levin