[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <72d6418b-5f37-c527-1b0a-be5e0cc88f5a@redhat.com>
Date: Thu, 3 Oct 2019 16:37:03 -0400
From: Waiman Long <longman@...hat.com>
To: Ingo Molnar <mingo@...nel.org>,
"Peter Zijlstra (Intel)" <peterz@...radead.org>,
Thomas Gleixner <tglx@...utronix.de>
Cc: linux-kernel@...r.kernel.org,
Masami Hiramatsu <mhiramat@...nel.org>,
Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
Juri Lelli <jlelli@...hat.com>
Subject: Re: [PATCH] lib/smp_processor_id: Don't use cpumask_equal()
On 10/3/19 4:18 PM, Waiman Long wrote:
> The check_preemption_disabled() function uses cpumask_equal() to see
> if the task is bounded to the current CPU only. cpumask_equal() calls
> memcmp() to do the comparison. As x86 doesn't have __HAVE_ARCH_MEMCMP,
> the slow memcmp() function in lib/string.c is used.
>
> On a RT kernel that call check_preemption_disabled() very frequently,
> below is the perf-record output of a certain microbenchmark:
>
> 42.75% 2.45% testpmd [kernel.kallsyms] [k] check_preemption_disabled
> 40.01% 39.97% testpmd [kernel.kallsyms] [k] memcmp
>
> We should avoid calling memcmp() in performance critical path. So the
> cpumask_equal() call is now replaced with an equivalent check that
> makes no external function call.
>
> Signed-off-by: Waiman Long <longman@...hat.com>
> ---
> lib/smp_processor_id.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c
> index 60ba93fc42ce..3fee05ac92f8 100644
> --- a/lib/smp_processor_id.c
> +++ b/lib/smp_processor_id.c
> @@ -23,7 +23,8 @@ unsigned int check_preemption_disabled(const char *what1, const char *what2)
> * Kernel threads bound to a single CPU can safely use
> * smp_processor_id():
> */
> - if (cpumask_equal(current->cpus_ptr, cpumask_of(this_cpu)))
> + if ((current->nr_cpus_allowed == 1) &&
> + cpumask_test_cpu(this_cpu, current->cpus_ptr))
> goto out;
>
> /*
My bad. The second check isn't really necessary.
Cheers,
Longman
Powered by blists - more mailing lists