From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lf1-f68.google.com (mail-lf1-f68.google.com [209.85.167.68]) (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 2729D46970E for ; Tue, 4 Feb 2020 01:05:24 +0300 (MSK) Received: by mail-lf1-f68.google.com with SMTP id 9so10793599lfq.10 for ; Mon, 03 Feb 2020 14:05:24 -0800 (PST) Date: Tue, 4 Feb 2020 01:05:20 +0300 From: Cyrill Gorcunov Message-ID: <20200203220520.GB12445@uranus> References: <20200115170524.20471-1-gorcunov@gmail.com> <20200115170524.20471-2-gorcunov@gmail.com> <20200203215627.l55b6rikwubeaqvn@tkn_work_nb> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20200203215627.l55b6rikwubeaqvn@tkn_work_nb> Subject: Re: [Tarantool-patches] [PATCH v2 1/2] fiber: use diag_ logger in fiber_madvise/mprotect failures List-Id: Tarantool development patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alexander Turenko Cc: tml On Tue, Feb 04, 2020 at 12:56:27AM +0300, Alexander Turenko wrote: > Everything looks correct. > > However, I don't believe my eyes, so I wrote a test case (see below). > > Don't sure whether it is good to have error injections for such > functions. Maybe LD_PRELOAD way is better (see [1]). Anyway, I checked > that the behaviour was incorrect using the test and that it becomes > correct. > > Please, consider the shared test and decide whether it worth to add it > to the repository (I don't sure). I think having errinj is better in a long term, it is controllable by our code and doesn't depend on any other hooks. Except - I would call it just FIBER_MPROTECT (without prot-none suffix) - iirc there was some test in app/ errinj or similar which should be updated everytime new errinj introduced. With this two things addressed feel free to put my Ack. And thanks a huge, Sasha, for this test