Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751815AbdHCIL5 (ORCPT ); Thu, 3 Aug 2017 04:11:57 -0400 Received: from mx2.suse.de ([195.135.220.15]:59831 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751083AbdHCILy (ORCPT ); Thu, 3 Aug 2017 04:11:54 -0400 Date: Thu, 3 Aug 2017 10:11:52 +0200 From: Michal Hocko To: Paul Moore Cc: Jeff Vander Stoep , linux-mm@kvack.org, LKML , selinux@tycho.nsa.gov, Mel Gorman Subject: Re: suspicious __GFP_NOMEMALLOC in selinux Message-ID: <20170803081152.GC12521@dhcp22.suse.cz> References: <20170802105018.GA2529@dhcp22.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4652 Lines: 125 [CC Mel] On Wed 02-08-17 17:45:56, Paul Moore wrote: > On Wed, Aug 2, 2017 at 6:50 AM, Michal Hocko wrote: > > Hi, > > while doing something completely unrelated to selinux I've noticed a > > really strange __GFP_NOMEMALLOC usage pattern in selinux, especially > > GFP_ATOMIC | __GFP_NOMEMALLOC doesn't make much sense to me. GFP_ATOMIC > > on its own allows to access memory reserves while the later flag tells > > we cannot use memory reserves at all. The primary usecase for > > __GFP_NOMEMALLOC is to override a global PF_MEMALLOC should there be a > > need. > > > > It all leads to fa1aa143ac4a ("selinux: extended permissions for > > ioctls") which doesn't explain this aspect so let me ask. Why is the > > flag used at all? Moreover shouldn't GFP_ATOMIC be actually GFP_NOWAIT. > > What makes this path important to access memory reserves? > > [NOTE: added the SELinux list to the CC line, please include that list > when asking SELinux questions] Sorry about that. Will keep it in mind for next posts > The GFP_ATOMIC|__GFP_NOMEMALLOC use in SELinux appears to be limited > to security/selinux/avc.c, and digging a bit, I'm guessing commit > fa1aa143ac4a copied the combination from 6290c2c43973 ("selinux: tag > avc cache alloc as non-critical") and the avc_alloc_node() function. Thanks for the pointer. That makes much more sense now. Back in 2012 we really didn't have a good way to distinguish non sleeping and atomic with reserves allocations. > I can't say that I'm an expert at the vm subsystem and the variety of > different GFP_* flags, but your suggestion of moving to GFP_NOWAIT in > security/selinux/avc.c seems reasonable and in keeping with the idea > behind commit 6290c2c43973. What do you think about the following? I haven't tested it but it should be rather straightforward. --- >From 6d506a75da83c0724ed399966971711f37d67411 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Thu, 3 Aug 2017 10:04:20 +0200 Subject: [PATCH] selinux: replace GFP_ATOMIC | __GFP_NOMEMALLOC with GFP_NOWAIT selinux avc code uses a weird combination of gfp flags. It asks for GFP_ATOMIC which allows access to memory reserves while it requests to not consume memory reserves by __GFP_NOMEMALLOC. This seems to be copying the pattern from 6290c2c43973 ("selinux: tag avc cache alloc as non-critical"). Back then (before d0164adc89f6 ("mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd")) we didn't have a good way to distinguish nowait and atomic allocations so the construct made some sense. Now we do not have to play tricks though and GFP_NOWAIT will provide the required semantic (aka do not sleep and do not consume any memory reserves). Signed-off-by: Michal Hocko --- security/selinux/avc.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/security/selinux/avc.c b/security/selinux/avc.c index 4b4293194aee..b2c159e832b6 100644 --- a/security/selinux/avc.c +++ b/security/selinux/avc.c @@ -346,27 +346,26 @@ static struct avc_xperms_decision_node struct avc_xperms_decision_node *xpd_node; struct extended_perms_decision *xpd; - xpd_node = kmem_cache_zalloc(avc_xperms_decision_cachep, - GFP_ATOMIC | __GFP_NOMEMALLOC); + xpd_node = kmem_cache_zalloc(avc_xperms_decision_cachep, GFP_NOWAIT); if (!xpd_node) return NULL; xpd = &xpd_node->xpd; if (which & XPERMS_ALLOWED) { xpd->allowed = kmem_cache_zalloc(avc_xperms_data_cachep, - GFP_ATOMIC | __GFP_NOMEMALLOC); + GFP_NOWAIT); if (!xpd->allowed) goto error; } if (which & XPERMS_AUDITALLOW) { xpd->auditallow = kmem_cache_zalloc(avc_xperms_data_cachep, - GFP_ATOMIC | __GFP_NOMEMALLOC); + GFP_NOWAIT); if (!xpd->auditallow) goto error; } if (which & XPERMS_DONTAUDIT) { xpd->dontaudit = kmem_cache_zalloc(avc_xperms_data_cachep, - GFP_ATOMIC | __GFP_NOMEMALLOC); + GFP_NOWAIT); if (!xpd->dontaudit) goto error; } @@ -394,8 +393,7 @@ static struct avc_xperms_node *avc_xperms_alloc(void) { struct avc_xperms_node *xp_node; - xp_node = kmem_cache_zalloc(avc_xperms_cachep, - GFP_ATOMIC|__GFP_NOMEMALLOC); + xp_node = kmem_cache_zalloc(avc_xperms_cachep, GFP_NOWAIT); if (!xp_node) return xp_node; INIT_LIST_HEAD(&xp_node->xpd_head); @@ -548,7 +546,7 @@ static struct avc_node *avc_alloc_node(void) { struct avc_node *node; - node = kmem_cache_zalloc(avc_node_cachep, GFP_ATOMIC|__GFP_NOMEMALLOC); + node = kmem_cache_zalloc(avc_node_cachep, GFP_NOWAIT); if (!node) goto out; -- 2.13.2 -- Michal Hocko SUSE Labs