From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lj1-f195.google.com (mail-lj1-f195.google.com [209.85.208.195]) (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 0A4F646970E for ; Thu, 19 Dec 2019 18:09:08 +0300 (MSK) Received: by mail-lj1-f195.google.com with SMTP id j6so6621596lja.2 for ; Thu, 19 Dec 2019 07:09:08 -0800 (PST) Date: Thu, 19 Dec 2019 18:09:07 +0300 From: Konstantin Osipov Message-ID: <20191219150907.GD30445@atlas> References: <20191219134126.25839-1-arkholga@tarantool.org> <20191219141445.GA26708@atlas> <4350dbf5-6fcd-634b-fc28-6eb2df288ca0@tarantool.org> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <4350dbf5-6fcd-634b-fc28-6eb2df288ca0@tarantool.org> Subject: Re: [Tarantool-patches] [PATCH v3] memtx: fix out of memory handling for rtree List-Id: Tarantool development patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Olga Arkhangelskaia Cc: tarantool-patches@dev.tarantool.org * Olga Arkhangelskaia [19/12/19 17:45]: > I can't agree - everything that has memtx_ should be in memtx, because > generic is used by vinyl, and sysview. Neither vinyl nor sysview build secondary indexes using build_next. This name - generic_* is a conversion mistake introduced during C++ -> C transition, it should be memtx_default_build_next. The generic build next should assert(false), it should be never called. If, for whatever reason, it's called, you still can split have a generic_memtx_build_next > And the only way I see do it generic - is to use index_reserve? Did you mean > index_reserve, that has  memtx_index_extent_reserve( inside? > > In this case it is not shorter. > -- Konstantin Osipov, Moscow, Russia