dm thin: stay in out-of-data-space mode once no_space_timeout expires

This fixes an issue where running out of data space would cause the
thin-pool's metadata to become read-only.  There was no reason to make
metadata read-only -- calling set_pool_mode() with PM_READ_ONLY was a
misguided way to error all queued and future write IOs.  We can
accomplish the same by degrading from PM_OUT_OF_DATA_SPACE to
PM_OUT_OF_DATA_SPACE with error_if_no_space enabled.

Otherwise, the use of PM_READ_ONLY could cause a race where commit() was
started before the PM_READ_ONLY transition but dm_pool_commit_metadata()
would go on to fail because the block manager had transitioned to
read-only.  The return of -EPERM from dm_pool_commit_metadata(), due to
attempting to commit while in read-only mode, caused the thin-pool to
set 'needs_check' because a metadata_operation_failed().  This needless
cascade of failures makes life for users more difficult than needed.

Reported-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>

diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c
index 8f015d924a242493639a18bf2c6411e336c9c2f0..34e79531ea3ffda98664278c8bf3e27c0d23aba4 100644 (file)
--- a/drivers/md/dm-thin.c
+++ b/drivers/md/dm-thin.c
@@ -2282,18 +2282,23 @@ static void do_waker(struct work_struct *ws)
        queue_delayed_work(pool->wq, &pool->waker, COMMIT_PERIOD);
 }
 
+static void notify_of_pool_mode_change_to_oods(struct pool *pool);
+
 /*
  * We're holding onto IO to allow userland time to react.  After the
  * timeout either the pool will have been resized (and thus back in
- * PM_WRITE mode), or we degrade to PM_READ_ONLY and start erroring IO.
+ * PM_WRITE mode), or we degrade to PM_OUT_OF_DATA_SPACE w/ error_if_no_space.
  */
 static void do_no_space_timeout(struct work_struct *ws)
 {
        struct pool *pool = container_of(to_delayed_work(ws), struct pool,
                                         no_space_timeout);
 
-       if (get_pool_mode(pool) == PM_OUT_OF_DATA_SPACE && !pool->pf.error_if_no_space)
-               set_pool_mode(pool, PM_READ_ONLY);
+       if (get_pool_mode(pool) == PM_OUT_OF_DATA_SPACE && !pool->pf.error_if_no_space) {
+               pool->pf.error_if_no_space = true;
+               notify_of_pool_mode_change_to_oods(pool);
+               error_retry_list(pool);
+       }
 }
 
 /*----------------------------------------------------------------*/
@@ -2371,6 +2376,14 @@ static void notify_of_pool_mode_change(struct pool *pool, const char *new_mode)
               dm_device_name(pool->pool_md), new_mode);
 }
 
+static void notify_of_pool_mode_change_to_oods(struct pool *pool)
+{
+       if (!pool->pf.error_if_no_space)
+               notify_of_pool_mode_change(pool, "out-of-data-space (queue IO)");
+       else
+               notify_of_pool_mode_change(pool, "out-of-data-space (error IO)");
+}
+
 static bool passdown_enabled(struct pool_c *pt)
 {
        return pt->adjusted_pf.discard_passdown;
@@ -2455,7 +2468,7 @@ static void set_pool_mode(struct pool *pool, enum pool_mode new_mode)
                 * frequently seeing this mode.
                 */
                if (old_mode != new_mode)
-                       notify_of_pool_mode_change(pool, "out-of-data-space");
+                       notify_of_pool_mode_change_to_oods(pool);
                pool->process_bio = process_bio_read_only;
                pool->process_discard = process_discard_bio;
                pool->process_cell = process_cell_read_only;