| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <20250826085205.1061353-10-yukuai1@huaweicloud.com>
Date: Tue, 26 Aug 2025 16:52:03 +0800
From: Yu Kuai <yukuai1@...weicloud.com>
To: hch@...radead.org,
corbet@....net,
agk@...hat.com,
snitzer@...nel.org,
mpatocka@...hat.com,
song@...nel.org,
xni@...hat.com,
hare@...e.de,
linan122@...wei.com,
colyli@...nel.org
Cc: linux-doc@...r.kernel.org,
linux-kernel@...r.kernel.org,
dm-devel@...ts.linux.dev,
linux-raid@...r.kernel.org,
yukuai3@...wei.com,
yukuai1@...weicloud.com,
yi.zhang@...wei.com,
yangerkun@...wei.com,
johnny.chenyi@...wei.com
Subject: [PATCH v6 md-6.18 09/11] md: add a new recovery_flag MD_RECOVERY_LAZY_RECOVER
From: Yu Kuai <yukuai3@...wei.com>
This flag is used by llbitmap in later patches to skip raid456 initial
recover and delay building initial xor data to first write.
Signed-off-by: Yu Kuai <yukuai3@...wei.com>
---
drivers/md/md.c | 47 +++++++++++++++++++++++++++++++++++++++++++++-
drivers/md/md.h | 2 ++
drivers/md/raid5.c | 19 +++++++++++++++----
3 files changed, 63 insertions(+), 5 deletions(-)
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 7196e7f6b2a4..199843356449 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -9199,6 +9199,39 @@ static sector_t md_sync_max_sectors(struct mddev *mddev,
}
}
+/*
+ * If lazy recovery is requested and all rdevs are in sync, select the rdev with
+ * the higest index to perfore recovery to build initial xor data, this is the
+ * same as old bitmap.
+ */
+static bool mddev_select_lazy_recover_rdev(struct mddev *mddev)
+{
+ struct md_rdev *recover_rdev = NULL;
+ struct md_rdev *rdev;
+ bool ret = false;
+
+ rcu_read_lock();
+ rdev_for_each_rcu(rdev, mddev) {
+ if (rdev->raid_disk < 0)
+ continue;
+
+ if (test_bit(Faulty, &rdev->flags) ||
+ !test_bit(In_sync, &rdev->flags))
+ break;
+
+ if (!recover_rdev || recover_rdev->raid_disk < rdev->raid_disk)
+ recover_rdev = rdev;
+ }
+
+ if (recover_rdev) {
+ clear_bit(In_sync, &recover_rdev->flags);
+ ret = true;
+ }
+
+ rcu_read_unlock();
+ return ret;
+}
+
static sector_t md_sync_position(struct mddev *mddev, enum sync_action action)
{
sector_t start = 0;
@@ -9230,6 +9263,14 @@ static sector_t md_sync_position(struct mddev *mddev, enum sync_action action)
start = rdev->recovery_offset;
rcu_read_unlock();
+ /*
+ * If there are no spares, and raid456 lazy initial recover is
+ * requested.
+ */
+ if (test_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery) &&
+ start == MaxSector && mddev_select_lazy_recover_rdev(mddev))
+ start = 0;
+
/* If there is a bitmap, we need to make sure all
* writes that started before we added a spare
* complete before we start doing a recovery.
@@ -9791,6 +9832,7 @@ static bool md_choose_sync_action(struct mddev *mddev, int *spares)
set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
+ clear_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery);
return true;
}
@@ -9799,6 +9841,7 @@ static bool md_choose_sync_action(struct mddev *mddev, int *spares)
remove_spares(mddev, NULL);
set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
+ clear_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery);
return true;
}
@@ -9808,7 +9851,7 @@ static bool md_choose_sync_action(struct mddev *mddev, int *spares)
* re-add.
*/
*spares = remove_and_add_spares(mddev, NULL);
- if (*spares) {
+ if (*spares || test_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery)) {
clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
@@ -10021,6 +10064,7 @@ void md_check_recovery(struct mddev *mddev)
}
clear_bit(MD_RECOVERY_RECOVER, &mddev->recovery);
+ clear_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery);
clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
clear_bit(MD_SB_CHANGE_PENDING, &mddev->sb_flags);
@@ -10131,6 +10175,7 @@ void md_reap_sync_thread(struct mddev *mddev)
clear_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
+ clear_bit(MD_RECOVERY_LAZY_RECOVER, &mddev->recovery);
/*
* We call mddev->cluster_ops->update_size here because sync_size could
* be changed by md_update_sb, and MD_RECOVERY_RESHAPE is cleared,
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 4fa5a3e68a0c..7b6357879a84 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -667,6 +667,8 @@ enum recovery_flags {
MD_RECOVERY_RESHAPE,
/* remote node is running resync thread */
MD_RESYNCING_REMOTE,
+ /* raid456 lazy initial recover */
+ MD_RECOVERY_LAZY_RECOVER,
};
enum md_ro_state {
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 672ab226e43c..5112658ef5f6 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -4705,10 +4705,21 @@ static void analyse_stripe(struct stripe_head *sh, struct stripe_head_state *s)
}
} else if (test_bit(In_sync, &rdev->flags))
set_bit(R5_Insync, &dev->flags);
- else if (sh->sector + RAID5_STRIPE_SECTORS(conf) <= rdev->recovery_offset)
- /* in sync if before recovery_offset */
- set_bit(R5_Insync, &dev->flags);
- else if (test_bit(R5_UPTODATE, &dev->flags) &&
+ else if (sh->sector + RAID5_STRIPE_SECTORS(conf) <=
+ rdev->recovery_offset) {
+ /*
+ * in sync if:
+ * - normal IO, or
+ * - resync IO that is not lazy recovery
+ *
+ * For lazy recovery, we have to mark the rdev without
+ * In_sync as failed, to build initial xor data.
+ */
+ if (!test_bit(STRIPE_SYNCING, &sh->state) ||
+ !test_bit(MD_RECOVERY_LAZY_RECOVER,
+ &conf->mddev->recovery))
+ set_bit(R5_Insync, &dev->flags);
+ } else if (test_bit(R5_UPTODATE, &dev->flags) &&
test_bit(R5_Expanded, &dev->flags))
/* If we've reshaped into here, we assume it is Insync.
* We will shortly update recovery_offset to make
--
2.39.2
Powered by blists - more mailing lists