From 3056e3aec8d8ba61a0710fb78b2d562600aa2ea7 Mon Sep 17 00:00:00 2001 From: Alex Lyakas Date: Tue, 4 Jun 2013 20:42:21 +0300 Subject: [PATCH] md/raid1: consider WRITE as successful only if at least one non-Faulty and non-rebuilding drive completed it. Without that fix, the following scenario could happen: - RAID1 with drives A and B; drive B was freshly-added and is rebuilding - Drive A fails - WRITE request arrives to the array. It is failed by drive A, so r1_bio is marked as R1BIO_WriteError, but the rebuilding drive B succeeds in writing it, so the same r1_bio is marked as R1BIO_Uptodate. - r1_bio arrives to handle_write_finished, badblocks are disabled, md_error()->error() does nothing because we don't fail the last drive of raid1 - raid_end_bio_io() calls call_bio_endio() - As a result, in call_bio_endio(): if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) clear_bit(BIO_UPTODATE, &bio->bi_flags); this code doesn't clear the BIO_UPTODATE flag, and the whole master WRITE succeeds, back to the upper layer. So we returned success to the upper layer, even though we had written the data onto the rebuilding drive only. But when we want to read the data back, we would not read from the rebuilding drive, so this data is lost. [neilb - applied identical change to raid10 as well] This bug can result in lost data, so it is suitable for any -stable kernel. Cc: stable@vger.kernel.org Signed-off-by: Alex Lyakas Signed-off-by: NeilBrown --- drivers/md/raid1.c | 12 +++++++++++- drivers/md/raid10.c | 12 +++++++++++- 2 files changed, 22 insertions(+), 2 deletions(-) diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 851023e2ba5d..f2db7a9d5964 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -427,7 +427,17 @@ static void raid1_end_write_request(struct bio *bio, int error) r1_bio->bios[mirror] = NULL; to_put = bio; - set_bit(R1BIO_Uptodate, &r1_bio->state); + /* + * Do not set R1BIO_Uptodate if the current device is + * rebuilding or Faulty. This is because we cannot use + * such device for properly reading the data back (we could + * potentially use it, if the current write would have felt + * before rdev->recovery_offset, but for simplicity we don't + * check this here. + */ + if (test_bit(In_sync, &conf->mirrors[mirror].rdev->flags) && + !test_bit(Faulty, &conf->mirrors[mirror].rdev->flags)) + set_bit(R1BIO_Uptodate, &r1_bio->state); /* Maybe we can clear some bad blocks. */ if (is_badblock(conf->mirrors[mirror].rdev, diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 018741ba9310..8000ee25650d 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -490,7 +490,17 @@ static void raid10_end_write_request(struct bio *bio, int error) sector_t first_bad; int bad_sectors; - set_bit(R10BIO_Uptodate, &r10_bio->state); + /* + * Do not set R10BIO_Uptodate if the current device is + * rebuilding or Faulty. This is because we cannot use + * such device for properly reading the data back (we could + * potentially use it, if the current write would have felt + * before rdev->recovery_offset, but for simplicity we don't + * check this here. + */ + if (test_bit(In_sync, &rdev->flags) && + !test_bit(Faulty, &rdev->flags)) + set_bit(R10BIO_Uptodate, &r10_bio->state); /* Maybe we can clear some bad blocks. */ if (is_badblock(rdev, -- 2.20.1