netdev - Re: [PATCH v2] net: cache for same cpu skb_attempt_defer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <457b4869-8f35-4619-8807-f79fc0122313@gmail.com>
Date: Tue, 13 Feb 2024 14:07:24 +0000
From: Pavel Begunkov <asml.silence@...il.com>
To: Eric Dumazet <edumazet@...gle.com>
Cc: netdev@...r.kernel.org, davem@...emloft.net, dsahern@...nel.org,
 pabeni@...hat.com, kuba@...nel.org
Subject: Re: [PATCH v2] net: cache for same cpu skb_attempt_defer_free

On 2/13/24 13:53, Eric Dumazet wrote:
> On Tue, Feb 13, 2024 at 2:42 PM Pavel Begunkov <asml.silence@...il.com> wrote:
>>
>> Optimise skb_attempt_defer_free() executed by the CPU the skb was
>> allocated on. Instead of __kfree_skb() -> kmem_cache_free() we can
>> disable softirqs and put the buffer into cpu local caches.
>>
>> Trying it with a TCP CPU bound ping pong benchmark (i.e. netbench), it
>> showed a 1% throughput improvement (392.2 -> 396.4 Krps). Cross checking
>> with profiles, the total CPU share of skb_attempt_defer_free() dropped by
>> 0.6%. Note, I'd expect the win doubled with rx only benchmarks, as the
>> optimisation is for the receive path, but the test spends >55% of CPU
>> doing writes.
>>
>> Signed-off-by: Pavel Begunkov <asml.silence@...il.com>
>> ---
>>
>> v2: remove in_hardirq()
>>
>>   net/core/skbuff.c | 16 +++++++++++++++-
>>   1 file changed, 15 insertions(+), 1 deletion(-)
>>
>> diff --git a/net/core/skbuff.c b/net/core/skbuff.c
>> index 9b790994da0c..f32f358ef1d8 100644
>> --- a/net/core/skbuff.c
>> +++ b/net/core/skbuff.c
>> @@ -6947,6 +6947,20 @@ void __skb_ext_put(struct skb_ext *ext)
>>   EXPORT_SYMBOL(__skb_ext_put);
>>   #endif /* CONFIG_SKB_EXTENSIONS */
>>
>> +static void kfree_skb_napi_cache(struct sk_buff *skb)
>> +{
>> +       /* if SKB is a clone, don't handle this case */
>> +       if (skb->fclone != SKB_FCLONE_UNAVAILABLE) {
>> +               __kfree_skb(skb);
>> +               return;
>> +       }
>> +
>> +       local_bh_disable();
>> +       skb_release_all(skb, SKB_DROP_REASON_NOT_SPECIFIED, false);
> 
> I am trying to understand why we use false instead of true here ?
> Or if you prefer:
> local_bh_disable();
> __napi_kfree_skb(skb, SKB_DROP_REASON_NOT_SPECIFIED);
> local_bh_enable();

Maybe it's my misunderstanding but disabled bh != "napi safe",
e.g. the napi_struct we're interested in might be scheduled for
another CPU. Which is also why "napi" prefix in percpu
napi_alloc_cache sounds a bit misleading to me.

The second reason is that it shouldn't change anything
performance wise

napi_pp_put_page(napi_safe) {
     ...
     if (napi_safe || in_softirq()) { ... }
}


>> +       napi_skb_cache_put(skb);
>> +       local_bh_enable();
>> +}
>> +
>>   /**
>>    * skb_attempt_defer_free - queue skb for remote freeing
>>    * @skb: buffer
>> @@ -6965,7 +6979,7 @@ void skb_attempt_defer_free(struct sk_buff *skb)
>>          if (WARN_ON_ONCE(cpu >= nr_cpu_ids) ||
>>              !cpu_online(cpu) ||
>>              cpu == raw_smp_processor_id()) {
>> -nodefer:       __kfree_skb(skb);
>> +nodefer:       kfree_skb_napi_cache(skb);
>>                  return;
>>          }
>>
>> --
>> 2.43.0
>>

-- 
Pavel Begunkov