From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756068AbaEOWgX (ORCPT ); Thu, 15 May 2014 18:36:23 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:39062 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755475AbaEOWgV (ORCPT ); Thu, 15 May 2014 18:36:21 -0400 Date: Thu, 15 May 2014 15:36:20 -0700 From: Andrew Morton To: Michal Hocko Cc: Jianyu Zhan , iamjoonsoo.kim@lge.com, aneesh.kumar@linux.vnet.ibm.com, n-horiguchi@ah.jp.nec.com, aarcange@redhat.com, steve.capper@linaro.org, davidlohr@hp.com, kirill.shutemov@linux.intel.com, dave.hansen@linux.intel.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] mm, hugetlb: move the error handle logic out of normal code path Message-Id: <20140515153620.344fe054b6b8d054a28fbf82@linux-foundation.org> In-Reply-To: <20140515090142.GB3938@dhcp22.suse.cz> References: <1400051459-20578-1-git-send-email-nasa4836@gmail.com> <20140515090142.GB3938@dhcp22.suse.cz> X-Mailer: Sylpheed 3.2.0beta5 (GTK+ 2.24.10; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 15 May 2014 11:01:42 +0200 Michal Hocko wrote: > On Wed 14-05-14 15:10:59, Jianyu Zhan wrote: > > alloc_huge_page() now mixes normal code path with error handle logic. > > This patches move out the error handle logic, to make normal code > > path more clean and redue code duplicate. > > I don't know. Part of the function returns and cleans up on its own and > other part relies on clean up labels. This is not so much nicer than the > previous state. That's actually a common pattern: foo() { if (check which doesn't change any state) return -Efoo; if (another check which doesn't change any state) return -Ebar; do_something_which_changes_state() if (another check) goto undo_that_state_chage; ... undo_that_state_change: ... } This ties into the main reason why we use all these gotos: to support evolution of the code. With multiple return points we risk later adding resource leaks and locking errors. Plus the code becomes more and more duplicative and spaghettified.