From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stephen Hemminger <stephen@networkplumber.org>
Subject: Re: rte_hash thread safe
Date: Mon, 23 Apr 2018 22:02:24 -0700
Message-ID: <20180423220224.1bd37f5b@xeon-e3>
References: <CAAQ1jEGBzgdLG=2g=3nbMHNCUT4W09UChpXeTG4tNR1z+FFHCQ@mail.gmail.com>
 <CAAQ1jEFF_MWsM2tSLG+9weh3JQNQkkEk2USJbi3QQxj6wG7C5g@mail.gmail.com>
 <20180423165039.51393aad@xeon-e3>
 <CAJC5fi33YBnE0H-Ug0DONx_QFtTcNO-ga+3k9t9_9b=i+1xr0A@mail.gmail.com>
 <20180423173034.7086b772@xeon-e3>
 <CAJC5fi1+2iJtyUPj7X_f-5HFVeF8u0F+TzjeGfOtUNTZUCYnwA@mail.gmail.com>
 <20180424034853.GA4546@jerin>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: Jim Murphy <jmurphy@arista.com>, Brijesh Singh
 <brijesh.s.singh@gmail.com>, dev@dpdk.org
To: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Return-path: <dev-bounces@dpdk.org>
Received: from mail-pf0-f195.google.com (mail-pf0-f195.google.com
 [209.85.192.195]) by dpdk.org (Postfix) with ESMTP id 77290E5D
 for <dev@dpdk.org>; Tue, 24 Apr 2018 07:02:27 +0200 (CEST)
Received: by mail-pf0-f195.google.com with SMTP id f15so11234739pfn.0
 for <dev@dpdk.org>; Mon, 23 Apr 2018 22:02:27 -0700 (PDT)
In-Reply-To: <20180424034853.GA4546@jerin>
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

On Tue, 24 Apr 2018 09:18:54 +0530
Jerin Jacob <jerin.jacob@caviumnetworks.com> wrote:

> -----Original Message-----
> > Date: Mon, 23 Apr 2018 17:48:50 -0700
> > From: Jim Murphy <jmurphy@arista.com>
> > To: Stephen Hemminger <stephen@networkplumber.org>
> > Cc: Brijesh Singh <brijesh.s.singh@gmail.com>, dev@dpdk.org
> > Subject: Re: [dpdk-dev] rte_hash thread safe
> > 
> > Anecdotally I've heard that the urcu hash implementation is slower than
> > rte_hash based on pure lookup performance. Has anyone considered adding RCU
> > hooks into rte_hash?  
> 
> 
> For one of our internal project on arm64, we did try rte_hash vs URCU hash.
> Based on our results URCU lookup was much better. By default, URCU
> library does not allocate the memory from huge page, But it has some
> plugin based scheme to override the memory allocation scheme to choose
> hugepage using DPDK backend.
> 

Also URCU hash table can be dynamically resized. With DPDK the application
has to guess at the number of hash heads which makes for bad behavior under
small and large workloads.