From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lf0-f72.google.com (mail-lf0-f72.google.com [209.85.215.72]) by kanga.kvack.org (Postfix) with ESMTP id BE5516B0024 for ; Tue, 10 Apr 2018 08:28:09 -0400 (EDT) Received: by mail-lf0-f72.google.com with SMTP id g13-v6so2436699lfl.15 for ; Tue, 10 Apr 2018 05:28:09 -0700 (PDT) Received: from mail-sor-f65.google.com (mail-sor-f65.google.com. [209.85.220.65]) by mx.google.com with SMTPS id x13sor675283ljj.87.2018.04.10.05.28.07 for (Google Transport Security); Tue, 10 Apr 2018 05:28:08 -0700 (PDT) Date: Tue, 10 Apr 2018 15:28:04 +0300 From: Cyrill Gorcunov Subject: Re: [v3 PATCH] mm: introduce arg_lock to protect arg_start|end and env_start|end in mm_struct Message-ID: <20180410122804.GD2041@uranus.lan> References: <1523310774-40300-1-git-send-email-yang.shi@linux.alibaba.com> <20180410090917.GZ21835@dhcp22.suse.cz> <20180410094047.GB2041@uranus.lan> <20180410104215.GB21835@dhcp22.suse.cz> <20180410110242.GC2041@uranus.lan> <20180410111001.GD21835@dhcp22.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180410111001.GD21835@dhcp22.suse.cz> Sender: owner-linux-mm@kvack.org List-ID: To: Michal Hocko Cc: Yang Shi , adobriyan@gmail.com, willy@infradead.org, mguzik@redhat.com, akpm@linux-foundation.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org On Tue, Apr 10, 2018 at 01:10:01PM +0200, Michal Hocko wrote: > > > > Because do_brk does vma manipulations, for this reason it's > > running under down_write_killable(&mm->mmap_sem). Or you > > mean something else? > > Yes, all we need the new lock for is to get a consistent view on brk > values. I am simply asking whether there is something fundamentally > wrong by doing the update inside the new lock while keeping the original > mmap_sem locking in the brk path. That would allow us to drop the > mmap_sem lock in the proc path when looking at brk values. Michal gimme some time. I guess we might do so, but I need some spare time to take more precise look into the code, hopefully today evening. Also I've a suspicion that we've wracked check_data_rlimit with this new lock in prctl. Need to verify it again.