Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751466Ab2KOVrg (ORCPT ); Thu, 15 Nov 2012 16:47:36 -0500 Received: from mail-pa0-f46.google.com ([209.85.220.46]:46758 "EHLO mail-pa0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751007Ab2KOVrf (ORCPT ); Thu, 15 Nov 2012 16:47:35 -0500 Date: Thu, 15 Nov 2012 13:47:33 -0800 (PST) From: David Rientjes X-X-Sender: rientjes@chino.kir.corp.google.com To: "Kirill A. Shutemov" cc: Andrew Morton , Andrea Arcangeli , linux-mm@kvack.org, Andi Kleen , "H. Peter Anvin" , linux-kernel@vger.kernel.org, "Kirill A. Shutemov" Subject: Re: [PATCH v5 05/11] thp: change_huge_pmd(): keep huge zero page write-protected In-Reply-To: <20121115084635.GC9676@otc-wbsnb-06> Message-ID: References: <1352300463-12627-1-git-send-email-kirill.shutemov@linux.intel.com> <1352300463-12627-6-git-send-email-kirill.shutemov@linux.intel.com> <20121115084635.GC9676@otc-wbsnb-06> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1471 Lines: 37 On Thu, 15 Nov 2012, Kirill A. Shutemov wrote: > > > diff --git a/mm/huge_memory.c b/mm/huge_memory.c > > > index d767a7c..05490b3 100644 > > > --- a/mm/huge_memory.c > > > +++ b/mm/huge_memory.c > > > @@ -1259,6 +1259,8 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, > > > pmd_t entry; > > > entry = pmdp_get_and_clear(mm, addr, pmd); > > > entry = pmd_modify(entry, newprot); > > > + if (is_huge_zero_pmd(entry)) > > > + entry = pmd_wrprotect(entry); > > > set_pmd_at(mm, addr, pmd, entry); > > > spin_unlock(&vma->vm_mm->page_table_lock); > > > ret = 1; > > > > Nack, this should be handled in pmd_modify(). > > I disagree. It means we will have to enable hzp per arch. Bad idea. > pmd_modify() only exists for those architectures with thp support already, so you've already implicitly enabled for all the necessary architectures with your patchset. > What's wrong with the check? > Anybody using pmd_modify() to set new protections in the future perhaps without knowledge of huge zero page can incorrectly make the huge zero page writable, which can never be allowed to happen. It's better to make sure it can never happen with the usual interface to modify protections. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/