From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752214Ab2DQU53 (ORCPT ); Tue, 17 Apr 2012 16:57:29 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:48459 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751024Ab2DQU52 (ORCPT ); Tue, 17 Apr 2012 16:57:28 -0400 Date: Tue, 17 Apr 2012 13:57:26 -0700 From: Andrew Morton To: Konstantin Khlebnikov Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org, Mel Gorman Subject: Re: [PATCH linux-next] mm/hugetlb: fix warning in alloc_huge_page/dequeue_huge_page_vma Message-Id: <20120417135726.05de2546.akpm@linux-foundation.org> In-Reply-To: <20120417122819.7438.26117.stgit@zurg> References: <20120417122819.7438.26117.stgit@zurg> X-Mailer: Sylpheed 3.0.2 (GTK+ 2.20.1; 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 Tue, 17 Apr 2012 16:28:19 +0400 Konstantin Khlebnikov wrote: > This patch fixes gcc warning (and bug?) introduced in linux-next commit cc9a6c877 > ("cpuset: mm: reduce large amounts of memory barrier related damage v3") > > Local variable "page" can be uninitialized if nodemask from vma policy does not > intersects with nodemask from cpuset. Even if it wouldn't happens it's better to > initialize this variable explicitly than to introduce kernel oops on weird corner case. > > mm/hugetlb.c: In function ___alloc_huge_page___: > mm/hugetlb.c:1135:5: warning: ___page___ may be used uninitialized in this function > > ... > > --- a/mm/hugetlb.c > +++ b/mm/hugetlb.c > @@ -532,7 +532,7 @@ static struct page *dequeue_huge_page_vma(struct hstate *h, > struct vm_area_struct *vma, > unsigned long address, int avoid_reserve) > { > - struct page *page; > + struct page *page = NULL; > struct mempolicy *mpol; > nodemask_t *nodemask; > struct zonelist *zonelist; hm, that's a pretty blatant use-uninitialised bug. I wonder why so few gcc versions report it. Mine doesn't.