Skip to contentSkip to conversation

FAILED: patch "[PATCH] dm-verity: make error counter atomic" failed to apply to 5.10-stable tree

PATCH New [email protected] 2 participants 2 messages Jul 20 – Jul 28

Subsystems (2)
Stable branch Device mapper
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 5.10-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]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y
git checkout FETCH_HEAD
git cherry-pick -x 8ec4d9c5a5cf4b61fc087f871465b1f79b393325
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2026072048-same-empathy-c4b3@gregkh' --subject-prefix 'PATCH 5.10.y' 'HEAD^..'

Possible dependencies:

thanks,

greg k-h

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

From 8ec4d9c5a5cf4b61fc087f871465b1f79b393325 Mon Sep 17 00:00:00 2001
From: Mikulas Patocka <[email protected]>
Date: Thu, 9 Jul 2026 21:37:38 +0200
Subject: [PATCH] dm-verity: make error counter atomic

The error counter "v->corrupted_errs" was not atomic, thus it could be subject to race conditions. The call to dm_audit_log_target("max-corrupted-errors") may be skipped due to the races.

Signed-off-by Mikulas Patocka <[email protected]>
Assisted-by Claude:claude-opus-4.6
Fixes 65ff5b7ddf05 ("dm verity: add error handling modes for corrupted blocks")
Cc [email protected]
drivers/md/dm-verity-target.c +7 -5
index e63a8290bd3a..1b0763091254 100644
@@ -180,14 +180,16 @@ static int verity_handle_err(struct dm_verity *v, enum verity_block_type type,
 	char *envp[] = { verity_env, NULL };
 	const char *type_str = "";
 	struct mapped_device *md = dm_table_get_md(v->ti->table);
+	int ce;
 
 	/* Corruption should be visible in device status in all modes */
 	v->hash_failed = true;
 
-	if (v->corrupted_errs >= DM_VERITY_MAX_CORRUPTED_ERRS)
-		goto out;
-
-	v->corrupted_errs++;
+	ce = atomic_read(&v->corrupted_errs);
+	do {
+		if (ce >= DM_VERITY_MAX_CORRUPTED_ERRS)
+			goto out;
+	} while (!atomic_try_cmpxchg(&v->corrupted_errs, &ce, ce + 1));
 
 	switch (type) {
 	case DM_VERITY_BLOCK_TYPE_DATA:

@@ -203,7 +205,7 @@ static int verity_handle_err(struct dm_verity *v, enum verity_block_type type,
 	DMERR_LIMIT("%s: %s block %llu is corrupted", v->data_dev->name,
 		    type_str, block);
 
-	if (v->corrupted_errs == DM_VERITY_MAX_CORRUPTED_ERRS) {
+	if (ce + 1 == DM_VERITY_MAX_CORRUPTED_ERRS) {
 		DMERR("%s: reached maximum errors", v->data_dev->name);
 		dm_audit_log_target(DM_MSG_PREFIX, "max-corrupted-errors", v->ti, 0);
 	}
drivers/md/dm-verity.h +1 -1
index 2922263501f6..e104a651c657 100644
@@ -68,7 +68,7 @@ struct dm_verity {
 	unsigned int digest_size;	/* digest size for the current hash algorithm */
 	enum verity_mode mode;	/* mode for handling verification errors */
 	enum verity_mode error_mode;/* mode for handling I/O errors */
-	unsigned int corrupted_errs;/* Number of errors for corrupted blocks */
+	atomic_t corrupted_errs;/* Number of errors for corrupted blocks */
 
 	struct workqueue_struct *verify_wq;
maintainer
#1 lore raw

PATCH 1/1 dm-verity: make error counter atomic

From: Mikulas Patocka <[email protected]>

[ Upstream commit 8ec4d9c5a5cf4b61fc087f871465b1f79b393325 ]

The error counter "v->corrupted_errs" was not atomic, thus it could be subject to race conditions. The call to dm_audit_log_target("max-corrupted-errors") may be skipped due to the races.

Signed-off-by Mikulas Patocka <[email protected]>
Assisted-by Claude:claude-opus-4.6
Fixes 65ff5b7ddf05 ("dm verity: add error handling modes for corrupted blocks")
Cc [email protected]
[ kept 5.15's braceless single-statement DMERR body instead of upstream's braced block containing the absent dm_audit_log_target() call ]
Signed-off-by Sasha Levin <[email protected]>
---
 drivers/md/dm-verity-target.c | 12 +++++++-----
 drivers/md/dm-verity.h        |  2 +-
 2 files changed, 8 insertions(+), 6 deletions(-)
drivers/md/dm-verity-target.c +7 -5
index 21ef7168cc0b04..926a8c990163b3 100644
@@ -217,14 +217,16 @@ static int verity_handle_err(struct dm_verity *v, enum verity_block_type type,
 	char *envp[] = { verity_env, NULL };
 	const char *type_str = "";
 	struct mapped_device *md = dm_table_get_md(v->ti->table);
+	int ce;
 
 	/* Corruption should be visible in device status in all modes */
 	v->hash_failed = 1;
 
-	if (v->corrupted_errs >= DM_VERITY_MAX_CORRUPTED_ERRS)
-		goto out;
-
-	v->corrupted_errs++;
+	ce = atomic_read(&v->corrupted_errs);
+	do {
+		if (ce >= DM_VERITY_MAX_CORRUPTED_ERRS)
+			goto out;
+	} while (!atomic_try_cmpxchg(&v->corrupted_errs, &ce, ce + 1));
 
 	switch (type) {
 	case DM_VERITY_BLOCK_TYPE_DATA:

@@ -240,7 +242,7 @@ static int verity_handle_err(struct dm_verity *v, enum verity_block_type type,
 	DMERR_LIMIT("%s: %s block %llu is corrupted", v->data_dev->name,
 		    type_str, block);
 
-	if (v->corrupted_errs == DM_VERITY_MAX_CORRUPTED_ERRS)
+	if (ce + 1 == DM_VERITY_MAX_CORRUPTED_ERRS)
 		DMERR("%s: reached maximum errors", v->data_dev->name);
 
 	snprintf(verity_env, DM_VERITY_ENV_LENGTH, "%s=%d,%llu",
drivers/md/dm-verity.h +1 -1
index f61c89c79cf5b9..ad5104daddb54a 100644
@@ -55,7 +55,7 @@ struct dm_verity {
 	unsigned int ahash_reqsize;/* the size of temporary space for crypto */
 	int hash_failed;	/* set to 1 if hash of any block failed */
 	enum verity_mode mode;	/* mode for handling verification errors */
-	unsigned corrupted_errs;/* Number of errors for corrupted blocks */
+	atomic_t corrupted_errs;/* Number of errors for corrupted blocks */
 
 	struct workqueue_struct *verify_wq;
Signature · 1 line
2.53.0