Re: [PATCH RESEND] locking/lockdep: Replace snprintf with strscpy in seq_stats

[Thorsten Blum]

On Tue, Jun 09, 2026 at 06:54:55AM -0700, Boqun Feng wrote:
> On Wed, Mar 18, 2026 at 09:44:49AM -0700, Boqun Feng wrote:
> > On Wed, Mar 18, 2026 at 11:43:49AM -0400, Waiman Long wrote:
> > > 
> > > On 3/17/26 8:14 PM, Thorsten Blum wrote:
> > > > Replace snprintf("%s", ...) with the faster and more direct strscpy().
> > > > 
> > > > Reviewed-by: Waiman Long <longman@xxxxxxxxxx>
> > > > Signed-off-by: Thorsten Blum <thorsten.blum@xxxxxxxxx>
> > > > ---
> > > >   kernel/locking/lockdep_proc.c | 5 +++--
> > > >   1 file changed, 3 insertions(+), 2 deletions(-)
> > > > 
> > > > diff --git a/kernel/locking/lockdep_proc.c b/kernel/locking/lockdep_proc.c
> > > > index 1916db9aa46b..e458fa258d05 100644
> > > > --- a/kernel/locking/lockdep_proc.c
> > > > +++ b/kernel/locking/lockdep_proc.c
> > > > @@ -19,6 +19,7 @@
> > > >   #include <linux/debug_locks.h>
> > > >   #include <linux/vmalloc.h>
> > > >   #include <linux/sort.h>
> > > > +#include <linux/string.h>
> > > >   #include <linux/uaccess.h>
> > > >   #include <asm/div64.h>
> > > > @@ -488,9 +489,9 @@ static void seq_stats(struct seq_file *m, struct lock_stat_data *data)
> > > >   		const char *key_name;
> > > >   		key_name = __get_key_name(ckey, str);
> > > > -		snprintf(name, namelen, "%s", key_name);
> > > > +		strscpy(name, key_name, namelen);
> > > >   	} else {
> > > > -		snprintf(name, namelen, "%s", cname);
> > > > +		strscpy(name, cname, namelen);
> > > >   	}
> > > >   	rcu_read_unlock_sched();
> > > > 
> > > Acked-by: Waiman Long <longman@xxxxxxxxxx>
> > > 
> > 
> > Queued, thank you both!
> > 
> 
> I was trying to merge it for v7.2 but sashiko found it out this:
> 
> 	https://sashiko.dev/#/patchset/20260605052331.1628-1-boqun%40kernel.org
> 
> Thorsten, could you take a look?

This appears valid, so we should probably drop this patch. I incorrectly
assumed snprintf() also requires the source to be non-NULL for %s.

Handling a failed lookup explicitly could improve this code, but that
requires more work and testing because it would change /proc/lock_stat
output from "(null)" to no output.

Thanks for catching this,
Thorsten