FAILED: patch "[PATCH] dm table: fix iterate_devices based device capability checks" failed to apply to 4.4-stable tree

PATCH In review [email protected] 2 participants 9 messages Mar 1, 2021 – Mar 8, 2021

Subsystems (2)

Stable branch Device mapper

lore

Share & tools

lore

[email protected]

author maintainer

2021-03-01 13:45 UTC #0 lore raw

The patch below does not apply to the 4.4-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From a4c8dd9c2d0987cf542a2a0c42684c9c6d78a04e Mon Sep 17 00:00:00 2001
From: Jeffle Xu <[email protected]>
Date: Tue, 2 Feb 2021 11:35:28 +0800
Subject: [PATCH] dm table: fix iterate_devices based device capability checks

According to the definition of dm_iterate_devices_fn:
* This function must iterate through each section of device used by the
* target until it encounters a non-zero return code, which it then returns.
* Returns zero if no callout returned non-zero.

For some target type (e.g. dm-stripe), one call of iterate_devices() may iterate multiple underlying devices internally, in which case a non-zero return code returned by iterate_devices_callout_fn will stop the iteration in advance. No iterate_devices_callout_fn should return non-zero unless device iteration should stop.

Rename dm_table_requires_stable_pages() to dm_table_any_dev_attr() and elevate it for reuse to stop iterating (and return non-zero) on the first device that causes iterate_devices_callout_fn to return non-zero. Use dm_table_any_dev_attr() to properly iterate through devices.

Rename device_is_nonrot() to device_is_rotational() and invert logic accordingly to fix improper disposition.

Fixes c3c4555edd10 ("dm table: clear add_random unless all devices have it set")

Fixes 4693c9668fdc ("dm table: propagate non rotational flag")

Cc [email protected]

Signed-off-by Jeffle Xu <[email protected]>

Signed-off-by Mike Snitzer <[email protected]>

drivers/md/dm-table.c +51 ~~-46~~

index 4acf2342f7ad..3e211e64f348 100644
@@ -1295,6 +1295,46 @@ struct dm_target *dm_table_find_target(struct dm_table *t, sector_t sector)
 	return &t->targets[(KEYS_PER_NODE * n) + k];
 }
 
+/*
+ * type->iterate_devices() should be called when the sanity check needs to
+ * iterate and check all underlying data devices. iterate_devices() will
+ * iterate all underlying data devices until it encounters a non-zero return
+ * code, returned by whether the input iterate_devices_callout_fn, or
+ * iterate_devices() itself internally.
+ *
+ * For some target type (e.g. dm-stripe), one call of iterate_devices() may
+ * iterate multiple underlying devices internally, in which case a non-zero
+ * return code returned by iterate_devices_callout_fn will stop the iteration
+ * in advance.
+ *
+ * Cases requiring _any_ underlying device supporting some kind of attribute,
+ * should use the iteration structure like dm_table_any_dev_attr(), or call
+ * it directly. @func should handle semantics of positive examples, e.g.
+ * capable of something.
+ *
+ * Cases requiring _all_ underlying devices supporting some kind of attribute,
+ * should use the iteration structure like dm_table_supports_nowait() or
+ * dm_table_supports_discards(). Or introduce dm_table_all_devs_attr() that
+ * uses an @anti_func that handle semantics of counter examples, e.g. not
+ * capable of something. So: return !dm_table_any_dev_attr(t, anti_func);
+ */
+static bool dm_table_any_dev_attr(struct dm_table *t,
+				  iterate_devices_callout_fn func)
+{
+	struct dm_target *ti;
+	unsigned int i;
+
+	for (i = 0; i < dm_table_get_num_targets(t); i++) {
+		ti = dm_table_get_target(t, i);
+
+		if (ti->type->iterate_devices &&
+		    ti->type->iterate_devices(ti, func, NULL))
+			return true;
+        }
+
+	return false;
+}
+
 static int count_device(struct dm_target *ti, struct dm_dev *dev,
 			sector_t start, sector_t len, void *data)
 {

@@ -1595,12 +1635,12 @@ static int dm_table_supports_dax_write_cache(struct dm_table *t)
 	return false;
 }
 
-static int device_is_nonrot(struct dm_target *ti, struct dm_dev *dev,
-			    sector_t start, sector_t len, void *data)
+static int device_is_rotational(struct dm_target *ti, struct dm_dev *dev,
+				sector_t start, sector_t len, void *data)
 {
 	struct request_queue *q = bdev_get_queue(dev->bdev);
 
-	return q && blk_queue_nonrot(q);
+	return q && !blk_queue_nonrot(q);
 }
 
 static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,

@@ -1611,23 +1651,6 @@ static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
 	return q && !blk_queue_add_random(q);
 }
 
-static bool dm_table_all_devices_attribute(struct dm_table *t,
-					   iterate_devices_callout_fn func)
-{
-	struct dm_target *ti;
-	unsigned i;
-
-	for (i = 0; i < dm_table_get_num_targets(t); i++) {
-		ti = dm_table_get_target(t, i);
-
-		if (!ti->type->iterate_devices ||
-		    !ti->type->iterate_devices(ti, func, NULL))
-			return false;
-	}
-
-	return true;
-}
-
 static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,
 					 sector_t start, sector_t len, void *data)
 {

@@ -1779,27 +1802,6 @@ static int device_requires_stable_pages(struct dm_target *ti,
 	return q && blk_queue_stable_writes(q);
 }
 
-/*
- * If any underlying device requires stable pages, a table must require
- * them as well.  Only targets that support iterate_devices are considered:
- * don't want error, zero, etc to require stable pages.
- */
-static bool dm_table_requires_stable_pages(struct dm_table *t)
-{
-	struct dm_target *ti;
-	unsigned i;
-
-	for (i = 0; i < dm_table_get_num_targets(t); i++) {
-		ti = dm_table_get_target(t, i);
-
-		if (ti->type->iterate_devices &&
-		    ti->type->iterate_devices(ti, device_requires_stable_pages, NULL))
-			return true;
-	}
-
-	return false;
-}
-
 void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 			       struct queue_limits *limits)
 {

@@ -1849,10 +1851,10 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 		dax_write_cache(t->md->dax_dev, true);
 
 	/* Ensure that all underlying devices are non-rotational. */
-	if (dm_table_all_devices_attribute(t, device_is_nonrot))
-		blk_queue_flag_set(QUEUE_FLAG_NONROT, q);
-	else
+	if (dm_table_any_dev_attr(t, device_is_rotational))
 		blk_queue_flag_clear(QUEUE_FLAG_NONROT, q);
+	else
+		blk_queue_flag_set(QUEUE_FLAG_NONROT, q);
 
 	if (!dm_table_supports_write_same(t))
 		q->limits.max_write_same_sectors = 0;

@@ -1864,8 +1866,11 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	/*
 	 * Some devices don't use blk_integrity but still want stable pages
 	 * because they do their own checksumming.
+	 * If any underlying device requires stable pages, a table must require
+	 * them as well.  Only targets that support iterate_devices are considered:
+	 * don't want error, zero, etc to require stable pages.
 	 */
-	if (dm_table_requires_stable_pages(t))
+	if (dm_table_any_dev_attr(t, device_requires_stable_pages))
 		blk_queue_flag_set(QUEUE_FLAG_STABLE_WRITES, q);
 	else
 		blk_queue_flag_clear(QUEUE_FLAG_STABLE_WRITES, q);

@@ -1876,7 +1881,7 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	 * Clear QUEUE_FLAG_ADD_RANDOM if any underlying device does not
 	 * have it set.
 	 */
-	if (blk_queue_add_random(q) && dm_table_all_devices_attribute(t, device_is_not_random))
+	if (blk_queue_add_random(q) && dm_table_any_dev_attr(t, device_is_not_random))
 		blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, q);
 
 	/*

Jeffle Xu

2021-03-05 06:31 UTC #1 lore raw

cover dm: device capability check fixes

- patch 1 is from upstream
- patch 2 is to fix the code specific to 4.4 (has been removed in upstream)

Jeffle Xu (2):
dm table: fix iterate_devices based device capability checks dm table: fix no_sg_merge iterate_devices based device capability
checks

 drivers/md/dm-table.c | 83 +++++++++++++++++++++++++++----------------
 1 file changed, 53 insertions(+), 30 deletions(-)

Signature · 1 line

2.27.0

Jeffle Xu

2021-03-05 06:30 UTC #2 lore raw

PATCH 1/2 dm table: fix iterate_devices based device capability checks

commit a4c8dd9c2d0987cf542a2a0c42684c9c6d78a04e upstream.

Rename device_is_nonrot() to device_is_rotational() and invert logic accordingly to fix improper disposition.

Fixes c3c4555edd10 ("dm table: clear add_random unless all devices have it set")

Fixes 4693c9668fdc ("dm table: propagate non rotational flag")

Cc [email protected]

Signed-off-by Jeffle Xu <[email protected]>

Signed-off-by Mike Snitzer <[email protected]>

[jeffle: no stable writes]

---
 drivers/md/dm-table.c | 71 ++++++++++++++++++++++++++++---------------
 1 file changed, 47 insertions(+), 24 deletions(-)

drivers/md/dm-table.c +47 ~~-24~~

index a5a6c7f073af..6580de65b81d 100644
@@ -1210,6 +1210,46 @@ struct dm_target *dm_table_find_target(struct dm_table *t, sector_t sector)
 	return &t->targets[(KEYS_PER_NODE * n) + k];
 }
 
+/*
+ * type->iterate_devices() should be called when the sanity check needs to
+ * iterate and check all underlying data devices. iterate_devices() will
+ * iterate all underlying data devices until it encounters a non-zero return
+ * code, returned by whether the input iterate_devices_callout_fn, or
+ * iterate_devices() itself internally.
+ *
+ * For some target type (e.g. dm-stripe), one call of iterate_devices() may
+ * iterate multiple underlying devices internally, in which case a non-zero
+ * return code returned by iterate_devices_callout_fn will stop the iteration
+ * in advance.
+ *
+ * Cases requiring _any_ underlying device supporting some kind of attribute,
+ * should use the iteration structure like dm_table_any_dev_attr(), or call
+ * it directly. @func should handle semantics of positive examples, e.g.
+ * capable of something.
+ *
+ * Cases requiring _all_ underlying devices supporting some kind of attribute,
+ * should use the iteration structure like dm_table_supports_nowait() or
+ * dm_table_supports_discards(). Or introduce dm_table_all_devs_attr() that
+ * uses an @anti_func that handle semantics of counter examples, e.g. not
+ * capable of something. So: return !dm_table_any_dev_attr(t, anti_func);
+ */
+static bool dm_table_any_dev_attr(struct dm_table *t,
+				  iterate_devices_callout_fn func)
+{
+	struct dm_target *ti;
+	unsigned int i;
+
+	for (i = 0; i < dm_table_get_num_targets(t); i++) {
+		ti = dm_table_get_target(t, i);
+
+		if (ti->type->iterate_devices &&
+		    ti->type->iterate_devices(ti, func, NULL))
+			return true;
+        }
+
+	return false;
+}
+
 static int count_device(struct dm_target *ti, struct dm_dev *dev,
 			sector_t start, sector_t len, void *data)
 {

@@ -1380,12 +1420,12 @@ static bool dm_table_discard_zeroes_data(struct dm_table *t)
 	return true;
 }
 
-static int device_is_nonrot(struct dm_target *ti, struct dm_dev *dev,
-			    sector_t start, sector_t len, void *data)
+static int device_is_rotational(struct dm_target *ti, struct dm_dev *dev,
+				sector_t start, sector_t len, void *data)
 {
 	struct request_queue *q = bdev_get_queue(dev->bdev);
 
-	return q && blk_queue_nonrot(q);
+	return q && !blk_queue_nonrot(q);
 }
 
 static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,

@@ -1404,23 +1444,6 @@ static int queue_supports_sg_merge(struct dm_target *ti, struct dm_dev *dev,
 	return q && !test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
 }
 
-static bool dm_table_all_devices_attribute(struct dm_table *t,
-					   iterate_devices_callout_fn func)
-{
-	struct dm_target *ti;
-	unsigned i = 0;
-
-	while (i < dm_table_get_num_targets(t)) {
-		ti = dm_table_get_target(t, i++);
-
-		if (!ti->type->iterate_devices ||
-		    !ti->type->iterate_devices(ti, func, NULL))
-			return false;
-	}
-
-	return true;
-}
-
 static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,
 					 sector_t start, sector_t len, void *data)
 {

@@ -1511,10 +1534,10 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 		q->limits.discard_zeroes_data = 0;
 
 	/* Ensure that all underlying devices are non-rotational. */
-	if (dm_table_all_devices_attribute(t, device_is_nonrot))
-		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
-	else
+	if (dm_table_any_dev_attr(t, device_is_rotational))
 		queue_flag_clear_unlocked(QUEUE_FLAG_NONROT, q);
+	else
+		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);
 
 	if (!dm_table_supports_write_same(t))
 		q->limits.max_write_same_sectors = 0;

@@ -1532,7 +1555,7 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	 * Clear QUEUE_FLAG_ADD_RANDOM if any underlying device does not
 	 * have it set.
 	 */
-	if (blk_queue_add_random(q) && dm_table_all_devices_attribute(t, device_is_not_random))
+	if (blk_queue_add_random(q) && dm_table_any_dev_attr(t, device_is_not_random))
 		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, q);
 
 	/*

Signature · 1 line

2.27.0

Jeffle Xu

2021-03-05 06:30 UTC #3 lore raw

PATCH 2/2 dm table: fix no_sg_merge iterate_devices based device capability checks

In reply to Jeffle Xu

Similar to commit a4c8dd9c2d09 ("dm table: fix iterate_devices based device capability checks"), fix NO_SG_MERGE capability check and invert logic of the corresponding iterate_devices_callout_fn so that all devices' NO_SG_MERGE capabilities are properly checked.

Signed-off-by Jeffle Xu <[email protected]>

Fixes 200612ec33e5 ("dm table: propagate QUEUE_FLAG_NO_SG_MERGE")

---
 drivers/md/dm-table.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

drivers/md/dm-table.c +6 -6

index 6580de65b81d..7ee520d4d216 100644
@@ -1436,12 +1436,12 @@ static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
 	return q && !blk_queue_add_random(q);
 }
 
-static int queue_supports_sg_merge(struct dm_target *ti, struct dm_dev *dev,
-				   sector_t start, sector_t len, void *data)
+static int queue_no_sg_merge(struct dm_target *ti, struct dm_dev *dev,
+			     sector_t start, sector_t len, void *data)
 {
 	struct request_queue *q = bdev_get_queue(dev->bdev);
 
-	return q && !test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
+	return q && test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
 }
 
 static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,

@@ -1542,10 +1542,10 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	if (!dm_table_supports_write_same(t))
 		q->limits.max_write_same_sectors = 0;
 
-	if (dm_table_all_devices_attribute(t, queue_supports_sg_merge))
-		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
-	else
+	if (dm_table_any_dev_attr(t, queue_no_sg_merge))
 		queue_flag_set_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
+	else
+		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
 
 	dm_table_verify_integrity(t);

Signature · 1 line

2.27.0

Greg KH

author maintainer

2021-03-05 06:46 UTC #4 lore raw

Re: [PATCH 4.4.y 2/2] dm table: fix no_sg_merge iterate_devices based device capability checks

Quoted 10 lines · Jeffle Xu

Similar to commit a4c8dd9c2d09 ("dm table: fix iterate_devices based device capability checks"), fix NO_SG_MERGE capability check and invert logic of the corresponding iterate_devices_callout_fn so that all devices' NO_SG_MERGE capabilities are properly checked.
Signed-off-by Jeffle Xu <[email protected]>
Fixes 200612ec33e5 ("dm table: propagate QUEUE_FLAG_NO_SG_MERGE")
---
 drivers/md/dm-table.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

What is the git commit id of this patch in Linus's tree?

thanks,

greg k-h

JeffleXu

2021-03-05 07:00 UTC #5 lore raw

Re: [PATCH 4.4.y 2/2] dm table: fix no_sg_merge iterate_devices based device capability checks

Quoted 12 lines · Greg KH

Quoted 10 lines · Jeffle Xu
Similar to commit a4c8dd9c2d09 ("dm table: fix iterate_devices based device capability checks"), fix NO_SG_MERGE capability check and invert logic of the corresponding iterate_devices_callout_fn so that all devices' NO_SG_MERGE capabilities are properly checked.
Signed-off-by Jeffle Xu <[email protected]>
Fixes 200612ec33e5 ("dm table: propagate QUEUE_FLAG_NO_SG_MERGE")
---
 drivers/md/dm-table.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

Quoted 6 lines

What is the git commit id of this patch in Linus's tree?
thanks,
greg k-h

The code this patch fixes, i.e., commit 200612ec33e5 ("dm table:
propagate QUEUE_FLAG_NO_SG_MERGE"), was removed since commit 2705c93742e9 ("block: kill QUEUE_FLAG_NO_SG_MERGE") in v5.1. Thus the code base doesn't exist in the latest master branch.

Signature · 2 lines

Thanks,
Jeffle

JeffleXu

2021-03-08 03:22 UTC #6 lore raw

Re: [PATCH 4.4.y 2/2] dm table: fix no_sg_merge iterate_devices based device capability checks

In reply to Jeffle Xu

Hi, Mike,

Would you please spare some time to review the following patches for stable tree?

- [1] for 4.4.y
- [2] for 4.9.y
- [3] for 4.14.y
- [4] for 4.19.y

While backporting [5] for stable tree, there's some extra code specific to stable tree needs to be fixed, see [6] for the background info.

[1] https://www.spinics.net/lists/stable/msg448728.html
[2] https://www.spinics.net/lists/stable/msg448737.html
[3] https://www.spinics.net/lists/stable/msg448740.html
[4] https://www.spinics.net/lists/stable/msg448748.html
[5]
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.12-rc2&id=a4c8dd9c2d0987cf542a2a0c42684c9c6d78a04e
[6] https://www.spinics.net/lists/stable/msg448757.html

Thanks,
Jeffle

Quoted 46 lines · Jeffle Xu

Signed-off-by Jeffle Xu <[email protected]>

Fixes 200612ec33e5 ("dm table: propagate QUEUE_FLAG_NO_SG_MERGE")

---
 drivers/md/dm-table.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

drivers/md/dm-table.c +6 -6

index 6580de65b81d..7ee520d4d216 100644
@@ -1436,12 +1436,12 @@ static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
 	return q && !blk_queue_add_random(q);
 }
 
-static int queue_supports_sg_merge(struct dm_target *ti, struct dm_dev *dev,
-				   sector_t start, sector_t len, void *data)
+static int queue_no_sg_merge(struct dm_target *ti, struct dm_dev *dev,
+			     sector_t start, sector_t len, void *data)
 {
 	struct request_queue *q = bdev_get_queue(dev->bdev);
 
-	return q && !test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
+	return q && test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
 }
 
 static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,

@@ -1542,10 +1542,10 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	if (!dm_table_supports_write_same(t))
 		q->limits.max_write_same_sectors = 0;
 
-	if (dm_table_all_devices_attribute(t, queue_supports_sg_merge))
-		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
-	else
+	if (dm_table_any_dev_attr(t, queue_no_sg_merge))
 		queue_flag_set_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
+	else
+		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
 
 	dm_table_verify_integrity(t);

Signature · 2 lines

Thanks,
Jeffle

JeffleXu

2021-03-08 13:46 UTC #7 lore raw

Re: [PATCH 4.4.y 2/2] dm table: fix no_sg_merge iterate_devices based device capability checks

Quoted 13 lines · JeffleXu

Hi, Mike,
Would you please spare some time to review the following patches for stable tree?
- [1] for 4.4.y
- [2] for 4.9.y
- [3] for 4.14.y
- [4] for 4.19.y
While backporting [5] for stable tree, there's some extra code specific to stable tree needs to be fixed, see [6] for the background info.

Quoted 8 lines

[1] https://www.spinics.net/lists/stable/msg448728.html
[2] https://www.spinics.net/lists/stable/msg448737.html
[3] https://www.spinics.net/lists/stable/msg448740.html
[4] https://www.spinics.net/lists/stable/msg448748.html
[5]
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v5.12-rc2&id=a4c8dd9c2d0987cf542a2a0c42684c9c6d78a04e
[6] https://www.spinics.net/lists/stable/msg448757.html

Sorry, it seems that this patch should not be separated from patch [5], as the NO_SG_MERGE capability check also calls dm_table_all_devices_attribute(), which has been removed in patch [5]. It will break the per-patch build-test when leave this patch not applied.

I will send a new version later. Please review on the new version directly.

Quoted 50 lines · Jeffle Xu

Quoted 46 lines

Signed-off-by Jeffle Xu <[email protected]>

Fixes 200612ec33e5 ("dm table: propagate QUEUE_FLAG_NO_SG_MERGE")

---
 drivers/md/dm-table.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

drivers/md/dm-table.c +6 -6

index 6580de65b81d..7ee520d4d216 100644
@@ -1436,12 +1436,12 @@ static int device_is_not_random(struct dm_target *ti, struct dm_dev *dev,
 	return q && !blk_queue_add_random(q);
 }
 
-static int queue_supports_sg_merge(struct dm_target *ti, struct dm_dev *dev,
-				   sector_t start, sector_t len, void *data)
+static int queue_no_sg_merge(struct dm_target *ti, struct dm_dev *dev,
+			     sector_t start, sector_t len, void *data)
 {
 	struct request_queue *q = bdev_get_queue(dev->bdev);
 
-	return q && !test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
+	return q && test_bit(QUEUE_FLAG_NO_SG_MERGE, &q->queue_flags);
 }
 
 static int device_not_write_same_capable(struct dm_target *ti, struct dm_dev *dev,

@@ -1542,10 +1542,10 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	if (!dm_table_supports_write_same(t))
 		q->limits.max_write_same_sectors = 0;
 
-	if (dm_table_all_devices_attribute(t, queue_supports_sg_merge))
-		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
-	else
+	if (dm_table_any_dev_attr(t, queue_no_sg_merge))
 		queue_flag_set_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
+	else
+		queue_flag_clear_unlocked(QUEUE_FLAG_NO_SG_MERGE, q);
 
 	dm_table_verify_integrity(t);

Signature · 2 lines

Thanks,
Jeffle

Greg KH

author maintainer

2021-03-05 06:47 UTC #8 lore raw

Re: [PATCH 4.4.y 0/2] dm: device capability check fixes

In reply to Jeffle Xu

Quoted 7 lines · Jeffle Xu

- patch 1 is from upstream
- patch 2 is to fix the code specific to 4.4 (has been removed in upstream)
Jeffle Xu (2):
dm table: fix iterate_devices based device capability checks dm table: fix no_sg_merge iterate_devices based device capability
checks

Why only 4.4.y for these? What about all of the missing kernels in the middle (4.9, 4.14, 4.19, 5.4)? I can't apply fixes to one older branch and ignore others, as someone might upgrade and have a regression.

Please provide fixes for those other trees too, and then I will be glad to merge these.

thanks,

greg k-h