From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx164.postini.com [74.125.245.164]) by kanga.kvack.org (Postfix) with SMTP id 21E156B00FF for ; Wed, 12 Sep 2012 17:32:41 -0400 (EDT) Date: Wed, 12 Sep 2012 14:32:39 -0700 From: Andrew Morton Subject: Re: [PATCH 2/2] memory-hotplug: don't replace lowmem pages with highmem Message-Id: <20120912143239.65fa8b58.akpm@linux-foundation.org> In-Reply-To: <1347414231-31451-2-git-send-email-minchan@kernel.org> References: <1347414231-31451-1-git-send-email-minchan@kernel.org> <1347414231-31451-2-git-send-email-minchan@kernel.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Minchan Kim Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org, Kamezawa Hiroyuki , Yasuaki Ishimatsu , Michal Nazarewicz , Marek Szyprowski , Wen Congyang On Wed, 12 Sep 2012 10:43:51 +0900 Minchan Kim wrote: > [1] reporeted that lowmem pages could be replaced by > highmem pages during migration of CMA and fixed. > > Quote from [1]'s description > " > The filesystem layer expects pages in the block device's mapping to not > be in highmem (the mapping's gfp mask is set in bdget()), but CMA can > currently replace lowmem pages with highmem pages, leading to crashes in > filesystem code such as the one below: > > Unable to handle kernel NULL pointer dereference at virtual address 00000400 > pgd = c0c98000 > [00000400] *pgd=00c91831, *pte=00000000, *ppte=00000000 > Internal error: Oops: 817 [#1] PREEMPT SMP ARM > CPU: 0 Not tainted (3.5.0-rc5+ #80) > PC is at __memzero+0x24/0x80 > ... > Process fsstress (pid: 323, stack limit = 0xc0cbc2f0) > Backtrace: > [] (ext4_getblk+0x0/0x180) from [] (ext4_bread+0x1c/0x98) > [] (ext4_bread+0x0/0x98) from [] (ext4_mkdir+0x160/0x3bc) > r4:c15337f0 > [] (ext4_mkdir+0x0/0x3bc) from [] (vfs_mkdir+0x8c/0x98) > [] (vfs_mkdir+0x0/0x98) from [] (sys_mkdirat+0x74/0xac) > r6:00000000 r5:c152eb40 r4:000001ff r3:c14b43f0 > [] (sys_mkdirat+0x0/0xac) from [] (sys_mkdir+0x20/0x24) > r6:beccdcf0 r5:00074000 r4:beccdbbc > [] (sys_mkdir+0x0/0x24) from [] (ret_fast_syscall+0x0/0x30) > " > > Memory-hotplug has same problem with CMA so [1]'s fix could be applied > with memory-hotplug, too. > > Fix it by reusing. Do we think this issue should be fixed in 3.6? Earlier? > @@ -809,8 +802,12 @@ do_migrate_range(unsigned long start_pfn, unsigned long end_pfn) > putback_lru_pages(&source); > goto out; > } > - /* this function returns # of failed pages */ > - ret = migrate_pages(&source, hotremove_migrate_alloc, 0, > + > + /* > + * alloc_migrate_target should be improooooved!! Not a helpful comment! If you've identified some improvement then please do provide all the details. > + * migrate_pages returns # of failed pages. > + */ > + ret = migrate_pages(&source, alloc_migrate_target, 0, > true, MIGRATE_SYNC); > if (ret) > putback_lru_pages(&source); -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org