From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-2.2 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,USER_AGENT_SANE_1 autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id EA187C5DF60 for ; Tue, 5 Nov 2019 22:22:10 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 9710D2087E for ; Tue, 5 Nov 2019 22:22:10 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 9710D2087E Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=intel.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 2F2E16B0003; Tue, 5 Nov 2019 17:22:10 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 27BF86B0005; Tue, 5 Nov 2019 17:22:10 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 11DC56B0006; Tue, 5 Nov 2019 17:22:10 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0096.hostedemail.com [216.40.44.96]) by kanga.kvack.org (Postfix) with ESMTP id EA3E76B0003 for ; Tue, 5 Nov 2019 17:22:09 -0500 (EST) Received: from smtpin28.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay03.hostedemail.com (Postfix) with SMTP id A2ADF824999B for ; Tue, 5 Nov 2019 22:22:09 +0000 (UTC) X-FDA: 76123647978.28.game70_4ce77bab6101b X-HE-Tag: game70_4ce77bab6101b X-Filterd-Recvd-Size: 9396 Received: from mga03.intel.com (mga03.intel.com [134.134.136.65]) by imf31.hostedemail.com (Postfix) with ESMTP for ; Tue, 5 Nov 2019 22:22:08 +0000 (UTC) X-Amp-Result: UNKNOWN X-Amp-Original-Verdict: FILE UNKNOWN X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by orsmga103.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 05 Nov 2019 14:22:06 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.68,271,1569308400"; d="scan'208";a="376846147" Received: from sjchrist-coffee.jf.intel.com (HELO linux.intel.com) ([10.54.74.41]) by orsmga005.jf.intel.com with ESMTP; 05 Nov 2019 14:22:05 -0800 Date: Tue, 5 Nov 2019 14:22:05 -0800 From: Sean Christopherson To: David Hildenbrand Cc: Dan Williams , Linux Kernel Mailing List , Linux MM , Michal Hocko , Andrew Morton , kvm-ppc@vger.kernel.org, linuxppc-dev , KVM list , linux-hyperv@vger.kernel.org, devel@driverdev.osuosl.org, xen-devel , X86 ML , Alexander Duyck , Alexander Duyck , Alex Williamson , Allison Randal , Andy Lutomirski , "Aneesh Kumar K.V" , Anshuman Khandual , Anthony Yznaga , Benjamin Herrenschmidt , Borislav Petkov , Boris Ostrovsky , Christophe Leroy , Cornelia Huck , Dave Hansen , Haiyang Zhang , "H. Peter Anvin" , Ingo Molnar , "Isaac J. Manjarres" , Jim Mattson , Joerg Roedel , Johannes Weiner , Juergen Gross , KarimAllah Ahmed , Kees Cook , "K. Y. Srinivasan" , "Matthew Wilcox (Oracle)" , Matt Sickler , Mel Gorman , Michael Ellerman , Michal Hocko , Mike Rapoport , Mike Rapoport , Nicholas Piggin , Oscar Salvador , Paolo Bonzini , Paul Mackerras , Paul Mackerras , Pavel Tatashin , Pavel Tatashin , Peter Zijlstra , Qian Cai , Radim =?utf-8?B?S3LEjW3DocWZ?= , Sasha Levin , Stefano Stabellini , Stephen Hemminger , Thomas Gleixner , Vitaly Kuznetsov , Vlastimil Babka , Wanpeng Li , YueHaibing , Adam Borowski Subject: Re: [PATCH v1 03/10] KVM: Prepare kvm_is_reserved_pfn() for PG_reserved changes Message-ID: <20191105222205.GA23297@linux.intel.com> References: <20191024120938.11237-1-david@redhat.com> <20191024120938.11237-4-david@redhat.com> <01adb4cb-6092-638c-0bab-e61322be7cf5@redhat.com> <613f3606-748b-0e56-a3ad-1efaffa1a67b@redhat.com> <20191105160000.GC8128@linux.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.24 (2015-08-30) X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On Tue, Nov 05, 2019 at 09:30:53PM +0100, David Hildenbrand wrote: > >>>I think I know what's going wrong: > >>> > >>>Pages that are pinned via gfn_to_pfn() and friends take a references, > >>>however are often released via > >>>kvm_release_pfn_clean()/kvm_release_pfn_dirty()/kvm_release_page_clean()... > >>> > >>> > >>>E.g., in arch/x86/kvm/x86.c:reexecute_instruction() > >>> > >>>... > >>>pfn = gfn_to_pfn(vcpu->kvm, gpa_to_gfn(gpa)); > >>>... > >>>kvm_release_pfn_clean(pfn); > >>> > >>> > >>> > >>>void kvm_release_pfn_clean(kvm_pfn_t pfn) > >>>{ > >>> if (!is_error_noslot_pfn(pfn) && !kvm_is_reserved_pfn(pfn)) > >>> put_page(pfn_to_page(pfn)); > >>>} > >>> > >>>This function makes perfect sense as the counterpart for kvm_get_pfn(): > >>> > >>>void kvm_get_pfn(kvm_pfn_t pfn) > >>>{ > >>> if (!kvm_is_reserved_pfn(pfn)) > >>> get_page(pfn_to_page(pfn)); > >>>} > >>> > >>> > >>>As all ZONE_DEVICE pages are currently reserved, pages pinned via > >>>gfn_to_pfn() and friends will often not see a put_page() AFAIKS. > > > >Assuming gup() takes a reference for ZONE_DEVICE pages, yes, this is a > >KVM bug. > > Yes, it does take a reference AFAIKs. E.g., > > mm/gup.c:gup_pte_range(): > ... > if (pte_devmap(pte)) { > if (unlikely(flags & FOLL_LONGTERM)) > goto pte_unmap; > > pgmap = get_dev_pagemap(pte_pfn(pte), pgmap); > if (unlikely(!pgmap)) { > undo_dev_pagemap(nr, nr_start, pages); > goto pte_unmap; > } > } else if (pte_special(pte)) > goto pte_unmap; > > VM_BUG_ON(!pfn_valid(pte_pfn(pte))); > page = pte_page(pte); > > head = try_get_compound_head(page, 1); > > try_get_compound_head() will increment the reference count. Doh, I looked right at that code and somehow didn't connect the dots. Thanks! > >>>Now, my patch does not change that, the result of > >>>kvm_is_reserved_pfn(pfn) will be unchanged. A proper fix for that would > >>>probably be > >>> > >>>a) To drop the reference to ZONE_DEVICE pages in gfn_to_pfn() and > >>>friends, after you successfully pinned the pages. (not sure if that's > >>>the right thing to do but you're the expert) > >>> > >>>b) To not use kvm_release_pfn_clean() and friends on pages that were > >>>definitely pinned. > > > >This is already KVM's intent, i.e. the purpose of the PageReserved() check > >is simply to avoid putting a non-existent reference. The problem is that > >KVM assumes pages with PG_reserved set are never pinned, which AFAICT was > >true when the code was first added. > > > >>(talking to myself, sorry) > >> > >>Thinking again, dropping this patch from this series could effectively also > >>fix that issue. E.g., kvm_release_pfn_clean() and friends would always do a > >>put_page() if "pfn_valid() and !PageReserved()", so after patch 9 also on > >>ZONDE_DEVICE pages. > > > >Yeah, this appears to be the correct fix. > > > >>But it would have side effects that might not be desired. E.g.,: > >> > >>1. kvm_pfn_to_page() would also return ZONE_DEVICE pages (might even be the > >>right thing to do). > > > >This should be ok, at least on x86. There are only three users of > >kvm_pfn_to_page(). Two of those are on allocations that are controlled by > >KVM and are guaranteed to be vanilla MAP_ANONYMOUS. The third is on guest > >memory when running a nested guest, and in that case supporting ZONE_DEVICE > >memory is desirable, i.e. KVM should play nice with a guest that is backed > >by ZONE_DEVICE memory. > > > >>2. kvm_set_pfn_dirty() would also set ZONE_DEVICE pages dirty (might be > >>okay) > > > >This is ok from a KVM perspective. > > What about > > void kvm_get_pfn(kvm_pfn_t pfn) > { > if (!kvm_is_reserved_pfn(pfn)) > get_page(pfn_to_page(pfn)); > } > > Is a pure get_page() sufficient in case of ZONE_DEVICE? > (asking because of the live references obtained via > get_dev_pagemap(pte_pfn(pte), pgmap) in mm/gup.c:gup_pte_range() somewhat > confuse me :) ) This ties into my concern with thp_adjust(). On x86, kvm_get_pfn() is only used in two flows, to manually get a ref for VM_IO/VM_PFNMAP pages and to switch the ref when mapping a non-hugetlbfs compound page, i.e. a THP. I assume VM_IO and PFNMAP can't apply to ZONE_DEVICE pages. In the thp_adjust() case, when a THP is encountered and the original PFN is for a non-PG_head page, KVM transfers the reference to the associated PG_head page[*] and maps the associated 2mb chunk/page. This is where KVM uses kvm_get_pfn() and could run afoul of the get_dev_pagemap() refcounts. [*] Technically I don't think it's guaranteed to be a PG_head, e.g. if the THP is a 1gb page, as KVM currently only maps THP as 2mb pages. But the idea is the same, transfer the refcount the PFN that's actually going into KVM's page tables. > > > >The scarier code (for me) is transparent_hugepage_adjust() and > >kvm_mmu_zap_collapsible_spte(), as I don't at all understand the > >interaction between THP and _PAGE_DEVMAP. > > The x86 KVM MMU code is one of the ugliest code I know (sorry, but it had to > be said :/ ). Luckily, this should be independent of the PG_reserved thingy > AFAIKs.