Skip to contentSkip to conversation

drivers/net/bonding: User strscpy() to copy device name

PATCH In review [email protected] 2 participants 3 messages Jun 6 – Jun 10

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drivers/net/bonding: User strscpy() to copy device name

From: David Laight <[email protected]>

---
This is one of a group of patches that remove potentially unbounded strcpy() calls.

They are mostly replaced by strscpy() or, when strlen() has just been called, with memcpy() (usually including the '\0').

Calls with copy string literals into arrays are left unchanged. They are safe and easily detected as such.

The changes were made by getting the compiler to detect the calls and then fixing the code by hand.

Note that all the changes are only compile tested.

Some Makefiles were changed to allow files to contain strcpy(). As well as 'difficult to fix' files, this included 'show' functions as they really need to use sysfs_emit() or seq_printf().

All the patches are being sent individually to avoid very long cc lists. Apologies for the terse commit messages and likely unexpected tags. (There are about 100 patches in total.)

 drivers/net/bonding/bond_options.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/bonding/bond_options.c +1 -1
index 7380cc4ee75a..c57b7d6af043 100644
@@ -1525,7 +1525,7 @@ static int bond_option_primary_set(struct bonding *bond,
 		if (strncmp(slave->dev->name, primary, IFNAMSIZ) == 0) {
 			slave_dbg(bond->dev, slave->dev, "Setting as primary slave\n");
 			rcu_assign_pointer(bond->primary_slave, slave);
-			strcpy(bond->params.primary, slave->dev->name);
+			strscpy(bond->params.primary, slave->dev->name);
 			bond->force_primary = true;
 			bond_select_active_slave(bond);
 			goto out;
Signature · 1 line
2.39.5
maintainer
#1 lore raw
Quoted 1 line · On Sat, 6 Jun 2026 21:26:08 +0100 [email protected] wrote:

From: David Laight <[email protected]>

Commit message is required. Please explain why you think this patch is needed. AFAICT it copies data between two well formed IFNAMSIZ strings.

Quoted 40 lines

---
This is one of a group of patches that remove potentially unbounded strcpy() calls.

They are mostly replaced by strscpy() or, when strlen() has just been called, with memcpy() (usually including the '\0').

Calls with copy string literals into arrays are left unchanged. They are safe and easily detected as such.

The changes were made by getting the compiler to detect the calls and then fixing the code by hand.

Note that all the changes are only compile tested.

Some Makefiles were changed to allow files to contain strcpy(). As well as 'difficult to fix' files, this included 'show' functions as they really need to use sysfs_emit() or seq_printf().

All the patches are being sent individually to avoid very long cc lists. Apologies for the terse commit messages and likely unexpected tags. (There are about 100 patches in total.)

 drivers/net/bonding/bond_options.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/bonding/bond_options.c +1 -1
index 7380cc4ee75a..c57b7d6af043 100644
@@ -1525,7 +1525,7 @@ static int bond_option_primary_set(struct bonding *bond,
 		if (strncmp(slave->dev->name, primary, IFNAMSIZ) == 0) {
 			slave_dbg(bond->dev, slave->dev, "Setting as primary slave\n");
 			rcu_assign_pointer(bond->primary_slave, slave);
-			strcpy(bond->params.primary, slave->dev->name);
+			strscpy(bond->params.primary, slave->dev->name);
 			bond->force_primary = true;
 			bond_select_active_slave(bond);
 			goto out;
Signature · 1 line
pw-bot: cr
#2 lore raw
Quoted 5 lines · On Tue, 9 Jun 2026 17:58:41 -0700 Jakub Kicinski <[email protected]> wrote:
Quoted 1 line · On Sat, 6 Jun 2026 21:26:08 +0100 [email protected] wrote:

From: David Laight <[email protected]>

Commit message is required. Please explain why you think this patch is needed. AFAICT it copies data between two well formed IFNAMSIZ strings.

Thinks...
I let strcpy(xx->array, "constant") through provided the array is big enough. This gets converted to memcpy().
I could also check the array sizes for strcpy(xx->array, yy->array) and allow provided the destination isn't shorter.
That would remove some of the 'annoying false positives'.
Unlike strscpy() this could be converted to a memcpy() (with or without explicitly writing the terminating '\0') for short (say <= 32 byte) lengths.

So this patch (and a few like it) can be dropped.
Some one else may want to remove strcpy() completely, but I was only trying to remove the ones that a simple compile-time test couldn't show were safe.

-- David

Quoted 41 lines
Quoted 41 lines

---
This is one of a group of patches that remove potentially unbounded strcpy() calls.

They are mostly replaced by strscpy() or, when strlen() has just been called, with memcpy() (usually including the '\0').

Calls with copy string literals into arrays are left unchanged. They are safe and easily detected as such.

The changes were made by getting the compiler to detect the calls and then fixing the code by hand.

Note that all the changes are only compile tested.

Some Makefiles were changed to allow files to contain strcpy(). As well as 'difficult to fix' files, this included 'show' functions as they really need to use sysfs_emit() or seq_printf().

All the patches are being sent individually to avoid very long cc lists. Apologies for the terse commit messages and likely unexpected tags. (There are about 100 patches in total.)

 drivers/net/bonding/bond_options.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/bonding/bond_options.c +1 -1
index 7380cc4ee75a..c57b7d6af043 100644
@@ -1525,7 +1525,7 @@ static int bond_option_primary_set(struct bonding *bond,
 		if (strncmp(slave->dev->name, primary, IFNAMSIZ) == 0) {
 			slave_dbg(bond->dev, slave->dev, "Setting as primary slave\n");
 			rcu_assign_pointer(bond->primary_slave, slave);
-			strcpy(bond->params.primary, slave->dev->name);
+			strscpy(bond->params.primary, slave->dev->name);
 			bond->force_primary = true;
 			bond_select_active_slave(bond);
 			goto out;