From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753838AbcARIRl (ORCPT ); Mon, 18 Jan 2016 03:17:41 -0500 Received: from LGEAMRELO12.lge.com ([156.147.23.52]:54754 "EHLO lgeamrelo12.lge.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750988AbcARIRj (ORCPT ); Mon, 18 Jan 2016 03:17:39 -0500 X-Original-SENDERIP: 156.147.1.127 X-Original-MAILFROM: minchan@kernel.org X-Original-SENDERIP: 165.244.98.76 X-Original-MAILFROM: minchan@kernel.org X-Original-SENDERIP: 10.177.223.161 X-Original-MAILFROM: minchan@kernel.org Date: Mon, 18 Jan 2016 17:20:00 +0900 From: Minchan Kim To: Vlastimil Babka CC: Sergey Senozhatsky , Junil Lee , ngupta@vflare.org, akpm@linux-foundation.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH v3] zsmalloc: fix migrate_zspage-zs_free race condition Message-ID: <20160118082000.GA20244@bbox> References: <1453095596-44055-1-git-send-email-junil0814.lee@lge.com> <20160118063611.GC7453@bbox> <20160118065434.GB459@swordfish> <20160118071157.GD7453@bbox> <20160118073939.GA30668@swordfish> <569C9A1F.2020303@suse.cz> MIME-Version: 1.0 In-Reply-To: <569C9A1F.2020303@suse.cz> User-Agent: Mutt/1.5.21 (2010-09-15) X-MIMETrack: Itemize by SMTP Server on LGEKRMHUB08/LGE/LG Group(Release 8.5.3FP6|November 21, 2013) at 2016/01/18 17:17:36, Serialize by Router on LGEKRMHUB08/LGE/LG Group(Release 8.5.3FP6|November 21, 2013) at 2016/01/18 17:17:36, Serialize complete at 2016/01/18 17:17:36 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Jan 18, 2016 at 08:54:07AM +0100, Vlastimil Babka wrote: > On 18.1.2016 8:39, Sergey Senozhatsky wrote: > > On (01/18/16 16:11), Minchan Kim wrote: > > [..] > >>> so, even if clear_bit_unlock/test_and_set_bit_lock do smp_mb or > >>> barrier(), there is no corresponding barrier from record_obj()->WRITE_ONCE(). > >>> so I don't think WRITE_ONCE() will help the compiler, or am I missing > >>> something? > >> > >> We need two things > >> 2. memory barrier. > >> > >> As compiler barrier, WRITE_ONCE works to prevent store tearing here > >> by compiler. > >> However, if we omit unpin_tag here, we lose memory barrier(e,g, smp_mb) > >> so another CPU could see stale data caused CPU memory reordering. > > > > oh... good find! lost release semantic of unpin_tag()... > > Ah, release semantic, good point indeed. OK then we need the v2 approach again, > with WRITE_ONCE() in record_obj(). Or some kind of record_obj_release() with > release semantic, which would be a bit more effective, but I guess migration is > not that critical path to be worth introducing it. WRITE_ONCE in record_obj would add more memory operations in obj_malloc but I don't feel it's too heavy in this phase so, How about this? Junil, Could you resend patch if others agree this? Thanks. +/* + * record_obj updates handle's value to free_obj and it shouldn't + * invalidate lock bit(ie, HANDLE_PIN_BIT) of handle, otherwise + * it breaks synchronization using pin_tag(e,g, zs_free) so let's + * keep the lock bit. + */ static void record_obj(unsigned long handle, unsigned long obj) { - *(unsigned long *)handle = obj; + int locked = (*(unsigned long *)handle) & (1< > Thanks, > Vlastimil > > > > > -ss > > >