]> git.ipfire.org Git - thirdparty/kernel/linux.git/commitdiff
ratelimit: Avoid atomic decrement under lock if already rate-limited
authorPaul E. McKenney <paulmck@kernel.org>
Wed, 16 Apr 2025 16:34:14 +0000 (09:34 -0700)
committerPaul E. McKenney <paulmck@kernel.org>
Thu, 8 May 2025 23:13:27 +0000 (16:13 -0700)
Currently, if the lock is acquired, the code unconditionally does
an atomic decrement on ->rs_n_left, even if that atomic operation is
guaranteed to return a limit-rate verdict.  A limit-rate verdict will
in fact be the common case when something is spewing into a rate limit.
This unconditional atomic operation incurs needless overhead and also
raises the spectre of counter wrap.

Therefore, do the atomic decrement only if there is some chance that
rates won't be limited.

Link: https://lore.kernel.org/all/fbe93a52-365e-47fe-93a4-44a44547d601@paulmck-laptop/
Link: https://lore.kernel.org/all/20250423115409.3425-1-spasswolf@web.de/
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Kuniyuki Iwashima <kuniyu@amazon.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: John Ogness <john.ogness@linutronix.de>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
lib/ratelimit.c

index a7aaebb7a7189ff594b4b79ea39468bf862c114b..ab8472edeb1d2d49590499be759674511b6c5732 100644 (file)
@@ -103,13 +103,16 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func)
                }
        }
        if (burst) {
-               int n_left;
+               int n_left = atomic_read(&rs->rs_n_left);
 
                /* The burst might have been taken by a parallel call. */
-               n_left = atomic_dec_return(&rs->rs_n_left);
-               if (n_left >= 0) {
-                       ret = 1;
-                       goto unlock_ret;
+
+               if (n_left > 0) {
+                       n_left = atomic_dec_return(&rs->rs_n_left);
+                       if (n_left >= 0) {
+                               ret = 1;
+                               goto unlock_ret;
+                       }
                }
        }