Hi,
在 2024/10/30 17:49, John Garry 写道:
Set BLK_FEAT_ATOMIC_WRITES_STACKED to enable atomic writes.
For an attempt to atomic write to a region which has bad blocks, error
the write as we just cannot do this. It is unlikely to find devices which
support atomic writes and bad blocks.
Signed-off-by: John Garry <john.g.garry@xxxxxxxxxx>
---
drivers/md/raid1.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index a10018282629..b57f69e3e8a7 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1524,6 +1524,13 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
blocked_rdev = rdev;
break;
}
+
+ if (is_bad && bio->bi_opf & REQ_ATOMIC) {
+ /* We just cannot atomically write this ... */
+ error = -EFAULT;
+ goto err_handle;
+ }
One nit here. If the write range are all badblocks, then this rdev is
skipped, and bio won't be splited, so I think atomic write is still fine
in this case. Perhaps move this conditon below?
Same for raid10.
Thanks,
Kuai
+
if (is_bad && first_bad <= r1_bio->sector) {
/* Cannot write here at all */
bad_sectors -= (r1_bio->sector - first_bad);
@@ -3220,6 +3227,7 @@ static int raid1_set_limits(struct mddev *mddev)
md_init_stacking_limits(&lim);
lim.max_write_zeroes_sectors = 0;
+ lim.features |= BLK_FEAT_ATOMIC_WRITES_STACKED;
err = mddev_stack_rdev_limits(mddev, &lim, MDDEV_STACK_INTEGRITY);
if (err) {
queue_limits_cancel_update(mddev->gendisk->queue);
