seccomp: don't leave dangling ->notif if file allocation fails
authorTycho Andersen <tycho@tycho.pizza>
Wed, 2 Sep 2020 14:09:53 +0000 (08:09 -0600)
committerKees Cook <keescook@chromium.org>
Tue, 8 Sep 2020 18:30:16 +0000 (11:30 -0700)
Christian and Kees both pointed out that this is a bit sloppy to open-code
both places, and Christian points out that we leave a dangling pointer to
->notif if file allocation fails. Since we check ->notif for null in order
to determine if it's ok to install a filter, this means people won't be
able to install a filter if the file allocation fails for some reason, even
if they subsequently should be able to.

To fix this, let's hoist this free+null into its own little helper and use
it.

Reported-by: Kees Cook <keescook@chromium.org>
Reported-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Tycho Andersen <tycho@tycho.pizza>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20200902140953.1201956-1-tycho@tycho.pizza
Signed-off-by: Kees Cook <keescook@chromium.org>
kernel/seccomp.c

index bb0dd9a..676d4af 100644 (file)
@@ -1109,6 +1109,12 @@ out:
 }
 
 #ifdef CONFIG_SECCOMP_FILTER
+static void seccomp_notify_free(struct seccomp_filter *filter)
+{
+       kfree(filter->notif);
+       filter->notif = NULL;
+}
+
 static void seccomp_notify_detach(struct seccomp_filter *filter)
 {
        struct seccomp_knotif *knotif;
@@ -1138,8 +1144,7 @@ static void seccomp_notify_detach(struct seccomp_filter *filter)
                complete(&knotif->ready);
        }
 
-       kfree(filter->notif);
-       filter->notif = NULL;
+       seccomp_notify_free(filter);
        mutex_unlock(&filter->notify_lock);
 }
 
@@ -1494,7 +1499,7 @@ static struct file *init_listener(struct seccomp_filter *filter)
 
 out_notif:
        if (IS_ERR(ret))
-               kfree(filter->notif);
+               seccomp_notify_free(filter);
 out:
        return ret;
 }