[V2] scsi: core: reallocate scsi device's budget map if default queue depth is changed

Commit Message

Ming Lei Jan. 27, 2022, 3:37 p.m. UTC

Martin reported that sdev->queue_depth can often be changed in
->slave_configure(), and now we uses ->cmd_per_lun as initial queue
depth for setting up sdev->budget_map. And some extreme ->cmd_per_lun
or ->can_queue won't be used at default actually, if they are used to
allocate sdev->budget_map, huge memory may be consumed just because
of bad ->cmd_per_lun.

Fix the issue by reallocating sdev->budget_map after ->slave_configure()
returns, at that time, queue_depth should be much more reasonable.

Cc: Bart Van Assche <[e-mail]>
Reported-by: Martin Wilck <[e-mail]>
Suggested-by: Martin Wilck <[e-mail]>
Tested-by: Martin Wilck <[e-mail]>
Reviewed-by: Martin Wilck <[e-mail]>
Signed-off-by: Ming Lei <[e-mail]>
---
V2:
	- rename one local variable, and fix a comment grammar issue, as
	reported by Bart

 drivers/scsi/scsi_scan.c | 56 ++++++++++++++++++++++++++++++++++++----
 1 file changed, 51 insertions(+), 5 deletions(-)

Comments

Bart Van Assche Jan. 27, 2022, 5:19 p.m. UTC | #1

On 1/27/22 07:37, Ming Lei wrote:
> Martin reported that sdev->queue_depth can often be changed in
> ->slave_configure(), and now we uses ->cmd_per_lun as initial queue
> depth for setting up sdev->budget_map. And some extreme ->cmd_per_lun
> or ->can_queue won't be used at default actually, if they are used to
> allocate sdev->budget_map, huge memory may be consumed just because
> of bad ->cmd_per_lun.
> 
> Fix the issue by reallocating sdev->budget_map after ->slave_configure()
> returns, at that time, queue_depth should be much more reasonable.

Reviewed-by: Bart Van Assche <[e-mail]>

Martin K. Petersen Jan. 31, 2022, 5:56 p.m. UTC | #2

Ming,

> Martin reported that sdev->queue_depth can often be changed in
> ->slave_configure(), and now we uses ->cmd_per_lun as initial queue
> depth for setting up sdev->budget_map. And some extreme ->cmd_per_lun
> or ->can_queue won't be used at default actually, if they are used to
> allocate sdev->budget_map, huge memory may be consumed just because of
> bad ->cmd_per_lun.

Applied to 5.17/scsi-fixes, thanks!

Martin K. Petersen Feb. 1, 2022, 2:03 a.m. UTC | #3

On Thu, 27 Jan 2022 23:37:33 +0800, Ming Lei wrote:

> Martin reported that sdev->queue_depth can often be changed in
> ->slave_configure(), and now we uses ->cmd_per_lun as initial queue
> depth for setting up sdev->budget_map. And some extreme ->cmd_per_lun
> or ->can_queue won't be used at default actually, if they are used to
> allocate sdev->budget_map, huge memory may be consumed just because
> of bad ->cmd_per_lun.
> 
> [...]

Applied to 5.17/scsi-fixes, thanks!

[1/1] scsi: core: reallocate scsi device's budget map if default queue depth is changed
      https://git.kernel.org/mkp/scsi/c/edb854a3680b

Patch

diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 3520b9384428..4c7fb5a5ea4a 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -214,6 +214,48 @@  static void scsi_unlock_floptical(struct scsi_device *sdev,
 			 SCSI_TIMEOUT, 3, NULL);
 }
 
+static int scsi_realloc_sdev_budget_map(struct scsi_device *sdev,
+					unsigned int depth)
+{
+	int new_shift = sbitmap_calculate_shift(depth);
+	bool need_alloc = !sdev->budget_map.map;
+	bool need_free = false;
+	int ret;
+	struct sbitmap sb_backup;
+
+	/*
+	 * realloc if new shift is calculated, which is caused by setting
+	 * up one new default queue depth after calling ->slave_configure
+	 */
+	if (!need_alloc && new_shift != sdev->budget_map.shift)
+		need_alloc = need_free = true;
+
+	if (!need_alloc)
+		return 0;
+
+	/*
+	 * Request queue has to be frozen for reallocating budget map,
+	 * and here disk isn't added yet, so freezing is pretty fast
+	 */
+	if (need_free) {
+		blk_mq_freeze_queue(sdev->request_queue);
+		sb_backup = sdev->budget_map;
+	}
+	ret = sbitmap_init_node(&sdev->budget_map,
+				scsi_device_max_queue_depth(sdev),
+				new_shift, GFP_KERNEL,
+				sdev->request_queue->node, false, true);
+	if (need_free) {
+		if (ret)
+			sdev->budget_map = sb_backup;
+		else
+			sbitmap_free(&sb_backup);
+		ret = 0;
+		blk_mq_unfreeze_queue(sdev->request_queue);
+	}
+	return ret;
+}
+
 /**
  * scsi_alloc_sdev - allocate and setup a scsi_Device
  * @starget: which target to allocate a &scsi_device for
@@ -306,11 +348,7 @@  static struct scsi_device *scsi_alloc_sdev(struct scsi_target *starget,
 	 * default device queue depth to figure out sbitmap shift
 	 * since we use this queue depth most of times.
 	 */
-	if (sbitmap_init_node(&sdev->budget_map,
-				scsi_device_max_queue_depth(sdev),
-				sbitmap_calculate_shift(depth),
-				GFP_KERNEL, sdev->request_queue->node,
-				false, true)) {
+	if (scsi_realloc_sdev_budget_map(sdev, depth)) {
 		put_device(&starget->dev);
 		kfree(sdev);
 		goto out;
@@ -1017,6 +1055,14 @@  static int scsi_add_lun(struct scsi_device *sdev, unsigned char *inq_result,
 			}
 			return SCSI_SCAN_NO_RESPONSE;
 		}
+
+		/*
+		 * queue_depth is often changed in ->slave_configure, so
+		 * setup budget map again for getting better memory uses
+		 * since memory consumption of the map depends on queue
+		 * depth heavily
+		 */
+		scsi_realloc_sdev_budget_map(sdev, sdev->queue_depth);
 	}
 
 	if (sdev->scsi_level >= SCSI_3)