Skip to contentSkip to conversation

dm: fix stacking of geometry limits for thinp and cache

PATCH In review Mike Snitzer 2 participants 2 messages Aug 15, 2013 Device mapper

lore
Share & tools
lore

Mike Snitzer

author
#0 lore raw

PATCH 1/1 dm: fix stacking of geometry limits for thinp and cache

Do not blindly override the queue limits (specifically io_min and io_opt). Allow traditional stacking of these limits if io_opt is a factor of each target's block size.

Without this patch mkfs.xfs does not recognize these target provided limits as a useful geometry (e.g. raid) so these hints are ignored. This was due to setting io_min to a useless value.

Signed-off-by Mike Snitzer <[email protected]>
---
 drivers/md/dm-cache-target.c |   12 ++++++++++--
 drivers/md/dm-thin.c         |   12 ++++++++++--
 2 files changed, 20 insertions(+), 4 deletions(-)
drivers/md/dm-cache-target.c +10 -2
Index: linux/drivers/md/dm-cache-target.c
===================================================================
@@ -2609,9 +2609,17 @@ static void set_discard_limits(struct ca
 static void cache_io_hints(struct dm_target *ti, struct queue_limits *limits)
 {
 	struct cache *cache = ti->private;
+	unsigned io_opt_sectors = limits->io_opt >> SECTOR_SHIFT;
 
-	blk_limits_io_min(limits, 0);
-	blk_limits_io_opt(limits, cache->sectors_per_block << SECTOR_SHIFT);
+	/*
+	 * If the system-determined stacked limits are compatible with the
+	 * cache's blocksize (io_opt is a factor) do not override them.
+	 */
+	if (io_opt_sectors < cache->sectors_per_block ||
+	    do_div(io_opt_sectors, cache->sectors_per_block)) {
+		blk_limits_io_min(limits, 0);
+		blk_limits_io_opt(limits, cache->sectors_per_block << SECTOR_SHIFT);
+	}
 	set_discard_limits(cache, limits);
 }
drivers/md/dm-thin.c +10 -2
Index: linux/drivers/md/dm-thin.c
===================================================================
@@ -2648,9 +2648,17 @@ static void pool_io_hints(struct dm_targ
 {
 	struct pool_c *pt = ti->private;
 	struct pool *pool = pt->pool;
+	unsigned io_opt_sectors = limits->io_opt >> SECTOR_SHIFT;
 
-	blk_limits_io_min(limits, 0);
-	blk_limits_io_opt(limits, pool->sectors_per_block << SECTOR_SHIFT);
+	/*
+	 * If the system-determined stacked limits are compatible with the
+	 * pool's blocksize (io_opt is a factor) do not override them.
+	 */
+	if (io_opt_sectors < pool->sectors_per_block ||
+	    do_div(io_opt_sectors, pool->sectors_per_block)) {
+		blk_limits_io_min(limits, 0);
+		blk_limits_io_opt(limits, pool->sectors_per_block << SECTOR_SHIFT);
+	}
 
 	/*
 	 * pt->adjusted_pf is a staging area for the actual features to use.
#1 lore raw
Quoted 3 lines · Mike Snitzer

Do not blindly override the queue limits (specifically io_min and io_opt). Allow traditional stacking of these limits if io_opt is a factor of each target's block size.

Ack