Skip to contentSkip to conversation

blkcg: add sanity check for blkcg policy operations

PATCH Superseded weiping zhang 3 participants 3 messages Oct 10, 2017 – Oct 11, 2017

Subsystems (2)

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 blkcg: add sanity check for blkcg policy operations

blkcg policy should keep cpd/pd's alloc_fn and free_fn in pairs, otherwise policy would register fail.

---
 block/blk-cgroup.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)
block/blk-cgroup.c +23 -0
index e7ec676..557c122 100644
@@ -1397,6 +1397,26 @@ void blkcg_deactivate_policy(struct request_queue *q,
 EXPORT_SYMBOL_GPL(blkcg_deactivate_policy);
 
 /**
+ * blkcg_policy_check_ops - check policy's operations
+ * @pol: blkcg policy to check
+ *
+ * Make sure cpd/pd_alloc_fn and cpd/pd_free_fn in pairs.
+ * Return true on success and false on failure.
+ */
+static bool blkcg_policy_check_ops(struct blkcg_policy *pol)
+{
+	if ((pol->cpd_alloc_fn && !pol->cpd_free_fn) ||
+		(!pol->cpd_alloc_fn && pol->cpd_free_fn))
+		return false;
+
+	if ((pol->pd_alloc_fn && !pol->pd_free_fn) ||
+		(!pol->pd_alloc_fn && pol->pd_free_fn))
+		return false;
+
+	return true;
+}
+
+/**
  * blkcg_policy_register - register a blkcg policy
  * @pol: blkcg policy to register
  *
@@ -1419,6 +1439,9 @@ int blkcg_policy_register(struct blkcg_policy *pol)
 	if (i >= BLKCG_MAX_POLS)
 		goto err_unlock;
 
+	if (!blkcg_policy_check_ops(pol))
+		goto err_unlock;
+
 	/* register @pol */
 	pol->plid = i;
 	blkcg_policy[pol->plid] = pol;
Signature · 1 line
2.9.4
#1 lore raw
Quoted 24 lines · weiping zhang

+ * blkcg_policy_check_ops - check policy's operations
+ * @pol: blkcg policy to check
+ *
+ * Make sure cpd/pd_alloc_fn and cpd/pd_free_fn in pairs.
+ * Return true on success and false on failure.
+ */

block/blk-cgroup.c +14 -0
+static bool blkcg_policy_check_ops(struct blkcg_policy *pol)
+{
+	if ((pol->cpd_alloc_fn && !pol->cpd_free_fn) ||
+		(!pol->cpd_alloc_fn && pol->cpd_free_fn))
+		return false;
+
+	if ((pol->pd_alloc_fn && !pol->pd_free_fn) ||
+		(!pol->pd_alloc_fn && pol->pd_free_fn))
+		return false;
+
+	return true;
+}
+
+/**
  * blkcg_policy_register - register a blkcg policy
  * @pol: blkcg policy to register
  *

block/blk-cgroup.c int blkcg_policy_register(struct blkcg_policy *pol) · line 1444

Quoted 6 lines
block/blk-cgroup.c +3 -0
@@ -1419,6 +1439,9 @@ int blkcg_policy_register(struct blkcg_policy *pol)
 	if (i >= BLKCG_MAX_POLS)
 		goto err_unlock;
 
+	if (!blkcg_policy_check_ops(pol))
+		goto err_unlock;
+

Can you merge it into the caller?

+ if ((pol->cpd_alloc_fn && !pol->cpd_free_fn) ||
+ (!pol->cpd_alloc_fn && pol->cpd_free_fn))
+ goto err_unlock;
+
+ if ((pol->pd_alloc_fn && !pol->pd_free_fn) ||
+ (!pol->pd_alloc_fn && pol->pd_free_fn))
+ goto err_unlock;

Signature · 6 lines
Johannes Thumshirn                                          Storage
[email protected]                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 N�rnberg
GF: Felix Imend�rffer, Jane Smithard, Graham Norton
HRB 21284 (AG N�rnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
#2 lore raw
Quoted 36 lines · Johannes Thumshirn
Quoted 24 lines · weiping zhang

+ * blkcg_policy_check_ops - check policy's operations
+ * @pol: blkcg policy to check
+ *
+ * Make sure cpd/pd_alloc_fn and cpd/pd_free_fn in pairs.
+ * Return true on success and false on failure.
+ */

block/blk-cgroup.c +14 -0
+static bool blkcg_policy_check_ops(struct blkcg_policy *pol)
+{
+	if ((pol->cpd_alloc_fn && !pol->cpd_free_fn) ||
+		(!pol->cpd_alloc_fn && pol->cpd_free_fn))
+		return false;
+
+	if ((pol->pd_alloc_fn && !pol->pd_free_fn) ||
+		(!pol->pd_alloc_fn && pol->pd_free_fn))
+		return false;
+
+	return true;
+}
+
+/**
  * blkcg_policy_register - register a blkcg policy
  * @pol: blkcg policy to register
  *

block/blk-cgroup.c int blkcg_policy_register(struct blkcg_policy *pol) · line 1444

Quoted 6 lines
block/blk-cgroup.c +3 -0
@@ -1419,6 +1439,9 @@ int blkcg_policy_register(struct blkcg_policy *pol)
 	if (i >= BLKCG_MAX_POLS)
 		goto err_unlock;
 
+	if (!blkcg_policy_check_ops(pol))
+		goto err_unlock;
+

Can you merge it into the caller?

block/blk-cgroup.c void blkcg_deactivate_policy(struct request_queue *q,

Quoted 7 lines
block/blk-cgroup.c +7 -0
+	if ((pol->cpd_alloc_fn && !pol->cpd_free_fn) ||
+		(!pol->cpd_alloc_fn && pol->cpd_free_fn))
+		goto err_unlock;
+
+	if ((pol->pd_alloc_fn && !pol->pd_free_fn) ||
+		(!pol->pd_alloc_fn && pol->pd_free_fn))
+		goto err_unlock;

It's ok for me, I send V2 later.

Thanks
Weiping