From: Artur Paszkiewicz <artur.paszkiew...@intel.com>
[ Upstream commit 07719ff767dcd8cc42050f185d332052f3816546 ]
If starting an array that is undergoing rebuild, make ppl recovery honor
the recovery_offset of a member disk and don't read data that is not yet
in-sync.
Signed-off-by: Artur Paszkiewicz <artur.paszkiew...@intel.com>
Signed-off-by: Shaohua Li <s...@fb.com>
Signed-off-by: Sasha Levin <alexander.le...@verizon.com>
---
drivers/md/raid5-ppl.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/md/raid5-ppl.c b/drivers/md/raid5-ppl.c
index cd026c88f7ef..702b76008886 100644
--- a/drivers/md/raid5-ppl.c
+++ b/drivers/md/raid5-ppl.c
@@ -758,7 +758,8 @@ static int ppl_recover_entry(struct ppl_log *log, struct
ppl_header_entry *e,
(unsigned long long)sector);
rdev = conf->disks[dd_idx].rdev;
- if (!rdev) {
+ if (!rdev || (!test_bit(In_sync, &rdev->flags) &&
+ sector >= rdev->recovery_offset)) {
pr_debug("%s:%*s data member disk %d missing\n",
__func__, indent, "", dd_idx);
update_parity = false;
--
2.11.0
