From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: <gorcunov@gmail.com> Received: from mail-lf1-f45.google.com (mail-lf1-f45.google.com [209.85.167.45]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by dev.tarantool.org (Postfix) with ESMTPS id 69155469719 for <tarantool-patches@dev.tarantool.org>; Sun, 1 Nov 2020 11:26:24 +0300 (MSK) Received: by mail-lf1-f45.google.com with SMTP id y184so11428763lfa.12 for <tarantool-patches@dev.tarantool.org>; Sun, 01 Nov 2020 01:26:24 -0700 (PDT) Date: Sun, 1 Nov 2020 11:26:20 +0300 From: Cyrill Gorcunov <gorcunov@gmail.com> Message-ID: <20201101082620.GJ198833@grain> References: <20201014133535.224573-1-gorcunov@gmail.com> <20201014133535.224573-4-gorcunov@gmail.com> <6572e8c8-b801-0db9-80c0-31bdaca89355@tarantool.org> <20201030125101.GF198833@grain> <7a6a6dd8-5eb0-0c55-2eeb-8aac3f5b11bd@tarantool.org> <20201031215946.GI198833@grain> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20201031215946.GI198833@grain> Subject: Re: [Tarantool-patches] [PATCH v8 3/4] box/cbox: implement cbox Lua module List-Id: Tarantool development patches <tarantool-patches.dev.tarantool.org> List-Unsubscribe: <https://lists.tarantool.org/mailman/options/tarantool-patches>, <mailto:tarantool-patches-request@dev.tarantool.org?subject=unsubscribe> List-Archive: <https://lists.tarantool.org/pipermail/tarantool-patches/> List-Post: <mailto:tarantool-patches@dev.tarantool.org> List-Help: <mailto:tarantool-patches-request@dev.tarantool.org?subject=help> List-Subscribe: <https://lists.tarantool.org/mailman/listinfo/tarantool-patches>, <mailto:tarantool-patches-request@dev.tarantool.org?subject=subscribe> To: Vladislav Shpilevoy <v.shpilevoy@tarantool.org> Cc: tml <tarantool-patches@dev.tarantool.org> On Sun, Nov 01, 2020 at 12:59:46AM +0300, Cyrill Gorcunov wrote: ... > > And I think we should consider moving to xxhash instead > > git@github.com:Cyan4973/xxHash.git > > which shows better result so far. Still as any noncrypto > hashes it has collissions. FWIW, I changed Murmur3/32 to xxhash/32 and here are results --- murmur3 [cyrill@grain cbox] ./hvsrb --mh mh create diff : 459890 ns mh lookup diff : 194201 ns xxhash [cyrill@grain cbox] ./hvsrb --mh mh create diff : 421618 ns mh lookup diff : 123385 ns lookup works almost twice faster on random data, creation time is comparable.