From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <vdavydov.dev@gmail.com>
Date: Tue, 12 Mar 2019 16:00:05 +0300
From: Vladimir Davydov <vdavydov.dev@gmail.com>
Subject: Re: [PATCH v5 2/4] memtx: introduce tuple compare hint
Message-ID: <20190312130005.wotntpiyq53yowm7@esperanza>
References: <cover.1551951540.git.kshcherbatov@tarantool.org>
 <53e165cf566611452821c692d79d621628e839ff.1551951540.git.kshcherbatov@tarantool.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <53e165cf566611452821c692d79d621628e839ff.1551951540.git.kshcherbatov@tarantool.org>
To: Kirill Shcherbatov <kshcherbatov@tarantool.org>
Cc: tarantool-patches@freelists.org
List-ID: <tarantool-patches.dev.tarantool.org>

On Thu, Mar 07, 2019 at 12:44:06PM +0300, Kirill Shcherbatov wrote:
> +/* {{{ tuple_aux_compare_with_key */
> +
> +static int
> +tuple_hinted_compare_with_key(const struct tuple *tuple, cmp_aux_t tuple_cmp_aux,
> +			      const char *key, uint32_t part_count,
> +			      cmp_aux_t key_cmp_aux, struct key_def *key_def)
> +{
> +	uint64_t tuple_hint = tuple_cmp_aux.hint;
> +	uint64_t key_hint = key_cmp_aux.hint;
> +	if (likely(tuple_hint != key_hint && tuple_hint != CMP_HINT_INVALID &&
> +		   key_hint != CMP_HINT_INVALID))
> +		return tuple_hint < key_hint ? -1 : 1;
> +	else
> +		return tuple_compare_with_key(tuple, key, part_count, key_def);

This will lead to double indirection - first you call
key_def->tuple_hinted_compare_with_key, which in turn calls
key_def->tuple_compare_with_key. This will introduce some
performance penalty, I guess. Avoid that, please.