diff mbox

nbd: set discard granularity properly

Message ID 1511990024-2742-1-git-send-email-josef@toxicpanda.com (mailing list archive)
State New, archived
Headers show

Commit Message

Josef Bacik Nov. 29, 2017, 9:13 p.m. UTC
From: Josef Bacik <jbacik@fb.com>

For some reason we had discard granularity set to 512 always even when
discards were disabled.  Fix this by having the default be 0, and then
if we turn it on set the discard granularity to the blocksize.

Signed-off-by: Josef Bacik <jbacik@fb.com>
---
 drivers/block/nbd.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

Comments

Martin K. Petersen Nov. 30, 2017, 3:58 p.m. UTC | #1
Josef,

> For some reason we had discard granularity set to 512 always even when
> discards were disabled.  Fix this by having the default be 0, and then
> if we turn it on set the discard granularity to the blocksize.

Originally, discard_max_bytes was the queue limit used to indicate
whether discards were supported or not. Callers were supposed to check
that for a value bigger than 0 before interpreting granularity and
alignment. However, many callers triggered on discard_granularity > 0
instead, and as a result things have been muddled somewhat.

So I don't have a problem with only reporting granularity when the
feature is actually enabled...

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
diff mbox

Patch

diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 5f2a4240a204..32b6f66ea5bf 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -231,6 +231,11 @@  static void nbd_size_clear(struct nbd_device *nbd)
 static void nbd_size_update(struct nbd_device *nbd)
 {
 	struct nbd_config *config = nbd->config;
+
+	if (config->flags & NBD_FLAG_SEND_TRIM) {
+		nbd->disk->queue->limits.discard_granularity = config->blksize;
+		blk_queue_max_discard_sectors(nbd->disk->queue, UINT_MAX);
+	}
 	blk_queue_logical_block_size(nbd->disk->queue, config->blksize);
 	blk_queue_physical_block_size(nbd->disk->queue, config->blksize);
 	set_capacity(nbd->disk, config->bytesize >> 9);
@@ -1040,6 +1045,8 @@  static void nbd_config_put(struct nbd_device *nbd)
 		nbd->config = NULL;
 
 		nbd->tag_set.timeout = 0;
+		nbd->disk->queue->limits.discard_granularity = 0;
+		blk_queue_max_discard_sectors(nbd->disk->queue, 0);
 		queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, nbd->disk->queue);
 
 		mutex_unlock(&nbd->config_lock);
@@ -1490,8 +1497,8 @@  static int nbd_dev_add(int index)
 	 */
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, disk->queue);
 	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, disk->queue);
-	disk->queue->limits.discard_granularity = 512;
-	blk_queue_max_discard_sectors(disk->queue, UINT_MAX);
+	disk->queue->limits.discard_granularity = 0;
+	blk_queue_max_discard_sectors(disk->queue, 0);
 	blk_queue_max_segment_size(disk->queue, UINT_MAX);
 	blk_queue_max_segments(disk->queue, USHRT_MAX);
 	blk_queue_max_hw_sectors(disk->queue, 65536);