From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753416Ab1BKHDF (ORCPT ); Fri, 11 Feb 2011 02:03:05 -0500 Received: from smtp-out.google.com ([216.239.44.51]:25438 "EHLO smtp-out.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752652Ab1BKHDD (ORCPT ); Fri, 11 Feb 2011 02:03:03 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=google.com; s=beta; h=date:from:x-x-sender:to:cc:subject:in-reply-to:message-id :references:user-agent:mime-version:content-type; b=c4I3wgFwEzkBJPIyqOnktfLtvulpPior6RAPz9791MokF3H8FL50vdSW44lwGf/mCj hKZ5ZQfpeeK7abl74Fwg== Date: Thu, 10 Feb 2011 23:02:50 -0800 (PST) From: Hugh Dickins X-X-Sender: hugh@sister.anvils To: Andrea Arcangeli cc: KAMEZAWA Hiroyuki , Daisuke Nishimura , "linux-mm@kvack.org" , "linux-kernel@vger.kernel.org" , "akpm@linux-foundation.org" , "hannes@cmpxchg.org" Subject: Re: [mmotm] BUG: Bad page state in process khugepaged ? In-Reply-To: <20110209200728.GQ3347@random.random> Message-ID: References: <20110209151036.f24a36a6.kamezawa.hiroyu@jp.fujitsu.com> <20110209155001.0e369475.nishimura@mxp.nes.nec.co.jp> <20110209155246.69a7f3a1.kamezawa.hiroyu@jp.fujitsu.com> <20110209200728.GQ3347@random.random> User-Agent: Alpine 2.00 (LSU 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-System-Of-Record: true Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 9 Feb 2011, Andrea Arcangeli wrote: > On Wed, Feb 09, 2011 at 03:52:46PM +0900, KAMEZAWA Hiroyuki wrote: > > On Wed, 9 Feb 2011 15:50:01 +0900 > > Daisuke Nishimura wrote: > > > > > > > > > > In hex, pc->flags was 7A00000000004 and this means PCG_USED bit is set. > > > > This implies page_remove_rmap() may not be called but ->mapping is NULL. Hmm? > > > > (7A is encoding of section number.) > > > > > > > Sigh.. it seems another freed-but-not-uncharged problem.. > > > > > > > Ah, ok, this is maybe caused by this. I'm sorry that I missed this. > > == > > static inline int free_pages_check(struct page *page) > > { > > if (unlikely(page_mapcount(page) | > > (page->mapping != NULL) | > > (atomic_read(&page->_count) != 0) | > > (page->flags & PAGE_FLAGS_CHECK_AT_FREE) | > > (mem_cgroup_bad_page_check(page)))) { <==========(*) > > bad_page(page); > > return 1; > > == > > > > Then, ok, this is a memcgroup and hugepage issue. > > > > I'll look into. > > Yes, the rest of the info on the page looked ok and shouldn't have > triggered a bad_page call. Thanks so much for looking into it. There is a separate little issue here, Andrea. Although we went to some trouble for bad_page() to take the page out of circulation yet let the system continue, your VM_BUG_ON(!PageBuddy) inside __ClearPageBuddy(page), from two callsites in bad_page(), is turning it into a fatal error when CONFIG_DEBUG_VM. You could that only MM developers switch CONFIG_DEBUG_VM=y, and they would like bad_page() to be fatal; maybe, but if so we should do that as an intentional patch, rather than as an unexpected side-effect ;) I noticed this a few days ago, but hadn't quite decided whether just to remove the VM_BUG_ON, or move it to __ClearPageBuddy's third callsite, or... doesn't matter much. I do also wonder if PageBuddy would better be _mapcount -something else: if we've got a miscounted page (itself unlikely of course), there's a chance that its _mapcount will be further decremented after it has been freed: whereupon it will go from -1 to -2, PageBuddy at present. The special avoidance of PageBuddy being that it can pull a whole block of pages into misuse if its mistaken. Hugh