From mboxrd@z Thu Jan  1 00:00:00 1970
From: Eric Dumazet <eric.dumazet@gmail.com>
Subject: Re: [PATCH v2 net-next 08/12] inet: frags: use rhashtables for
 reassembly units
Date: Fri, 30 Mar 2018 15:43:55 -0700
Message-ID: <cf37cf36-adca-3a1d-ff62-983d232f8bd1@gmail.com>
References: <20180330204236.43675-1-edumazet@google.com>
 <20180330204236.43675-9-edumazet@google.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Cc: netdev <netdev@vger.kernel.org>, Florian Westphal <fw@strlen.de>,
        Herbert Xu <herbert@gondor.apana.org.au>,
        Thomas Graf <tgraf@suug.ch>,
        Jesper Dangaard Brouer <brouer@redhat.com>,
        Alexander Aring <alex.aring@gmail.com>,
        Stefan Schmidt <stefan@osg.samsung.com>,
        Kirill Tkhai <ktkhai@virtuozzo.com>,
        Nikolay Aleksandrov <nikolay@redhat.com>
To: Eric Dumazet <edumazet@google.com>,
        "David S . Miller" <davem@davemloft.net>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-pg0-f52.google.com ([74.125.83.52]:35128 "EHLO
        mail-pg0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1752384AbeC3Wn6 (ORCPT
        <rfc822;netdev@vger.kernel.org>); Fri, 30 Mar 2018 18:43:58 -0400
Received: by mail-pg0-f52.google.com with SMTP id j3so5803801pgf.2
        for <netdev@vger.kernel.org>; Fri, 30 Mar 2018 15:43:58 -0700 (PDT)
In-Reply-To: <20180330204236.43675-9-edumazet@google.com>
Content-Language: en-US
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>



On 03/30/2018 01:42 PM, Eric Dumazet wrote:
> Some applications still rely on IP fragmentation, and to be fair linux
> reassembly unit is not working under any serious load.

...

> -
>  static struct inet_frag_queue *inet_frag_alloc(struct netns_frags *nf,
>  					       struct inet_frags *f,
>  					       void *arg)
>  {
>  	struct inet_frag_queue *q;
>  
> -	if (!nf->high_thresh || frag_mem_limit(nf) > nf->high_thresh) {
> -		inet_frag_schedule_worker(f);
> +	if (!nf->high_thresh || frag_mem_limit(nf) > nf->high_thresh)
>  		return NULL;
> -	}
>  
>  	q = kmem_cache_zalloc(f->frags_cachep, GFP_ATOMIC);
>  	if (!q)
> @@ -374,59 +167,53 @@ static struct inet_frag_queue *inet_frag_alloc(struct netns_frags *nf,
>  
>  	timer_setup(&q->timer, f->frag_expire, 0);
>  	spin_lock_init(&q->lock);
> -	refcount_set(&q->refcnt, 1);
> +	refcount_set(&q->refcnt, 3);
>  
>  	return q;
>  }
>  
>  static struct inet_frag_queue *inet_frag_create(struct netns_frags *nf,
> -						struct inet_frags *f,
>  						void *arg)
>  {
> +	struct inet_frags *f = nf->f;
>  	struct inet_frag_queue *q;
> +	int err;
>  
>  	q = inet_frag_alloc(nf, f, arg);
>  	if (!q)
>  		return NULL;
>  
> -	return inet_frag_intern(nf, q, f, arg);
> +	mod_timer(&q->timer, jiffies + nf->timeout);
> +
> +	err = rhashtable_insert_fast(&nf->rhashtable, &q->node,
> +				     f->rhash_params);
> +	add_frag_mem_limit(nf, f->qsize);

When adding back inet_frag_alloc() to address Kirill feedback,
it looks like I forgot to remove this add_frag_mem_limit() :/