On Tue, Nov 22 2016 at 7:48P -0500,
Mike Snitzer <snit...@redhat.com> wrote:
> But regardless, what certainly needs fixing is the inconsistency of
> inheriting DM_TYPE_MQ_REQUEST_BASED but not setting all_blk_mq to true
> (all_blk_mq == true is implied by DM_TYPE_MQ_REQUEST_BASED).
>
> I'm now questioning why we even need the 'all_blk_mq' state within the
> table. I'll revisit the "why?" on all that in my previous commits.
> I'll likely get back with you on this point tomorrow. And will
> hopefully have a fix for you.
We need 'all_blk_mq' because DM_TYPE_* is only used to establish the
DM device's type of request_queue. It doesn't say anything about the DM
device's underlying devices.
Anyway, this _untested_ patch should hopefully resolve the 'all_blk_mq'
inconsistency you saw:
diff --git a/drivers/md/dm-table.c b/drivers/md/dm-table.c
index 8b013ea..8ce81d0 100644
--- a/drivers/md/dm-table.c
+++ b/drivers/md/dm-table.c
@@ -924,12 +924,6 @@ static int dm_table_determine_type(struct dm_table *t)
BUG_ON(!request_based); /* No targets in this table */
- if (list_empty(devices) && __table_type_request_based(live_md_type)) {
- /* inherit live MD type */
- t->type = live_md_type;
- return 0;
- }
-
/*
* The only way to establish DM_TYPE_MQ_REQUEST_BASED is by
* having a compatible target use dm_table_set_type.
@@ -948,6 +942,19 @@ static int dm_table_determine_type(struct dm_table *t)
return -EINVAL;
}
+ if (list_empty(devices)) {
+ int srcu_idx;
+ struct dm_table *live_table = dm_get_live_table(t->md,
&srcu_idx);
+
+ /* inherit live table's type and all_blk_mq */
+ if (live_table) {
+ t->type = live_table->type;
+ t->all_blk_mq = live_table->all_blk_mq;
+ }
+ dm_put_live_table(t->md, srcu_idx);
+ return 0;
+ }
+
/* Non-request-stackable devices can't be used for request-based dm */
list_for_each_entry(dd, devices, list) {
struct request_queue *q = bdev_get_queue(dd->dm_dev->bdev);
--
dm-devel mailing list
dm-devel@redhat.com
https://www.redhat.com/mailman/listinfo/dm-devel
