From patchwork Wed Dec 23 11:25:49 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiao Guangrong X-Patchwork-Id: 7910791 Return-Path: X-Original-To: patchwork-kvm@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 7DBF4BEEE5 for ; Wed, 23 Dec 2015 11:34:40 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 7C2EF2037C for ; Wed, 23 Dec 2015 11:34:39 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 5C38620375 for ; Wed, 23 Dec 2015 11:34:38 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932947AbbLWLdI (ORCPT ); Wed, 23 Dec 2015 06:33:08 -0500 Received: from mga01.intel.com ([192.55.52.88]:33760 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932531AbbLWLdF (ORCPT ); Wed, 23 Dec 2015 06:33:05 -0500 Received: from orsmga002.jf.intel.com ([10.7.209.21]) by fmsmga101.fm.intel.com with ESMTP; 23 Dec 2015 03:33:05 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.20,468,1444719600"; d="scan'208";a="877401172" Received: from xiaoreal1.sh.intel.com (HELO xiaoreal1.sh.intel.com.sh.intel.com) ([10.239.48.79]) by orsmga002.jf.intel.com with ESMTP; 23 Dec 2015 03:33:03 -0800 From: Xiao Guangrong To: pbonzini@redhat.com Cc: gleb@kernel.org, mtosatti@redhat.com, kvm@vger.kernel.org, linux-kernel@vger.kernel.org, kai.huang@linux.intel.com, jike.song@intel.com, Xiao Guangrong Subject: [PATCH v2 06/11] KVM: MMU: let page fault handler be aware tracked page Date: Wed, 23 Dec 2015 19:25:49 +0800 Message-Id: <1450869954-30273-7-git-send-email-guangrong.xiao@linux.intel.com> X-Mailer: git-send-email 1.8.3.1 In-Reply-To: <1450869954-30273-1-git-send-email-guangrong.xiao@linux.intel.com> References: <1450869954-30273-1-git-send-email-guangrong.xiao@linux.intel.com> Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP The page fault caused by write access on the write tracked page can not be fixed, it always need to be emulated. page_fault_handle_page_track() is the fast path we introduce here to skip holding mmu-lock and shadow page table walking However, if the page table is not present, it is worth making the page table entry present and readonly to make the read access happy mmu_need_write_protect() need to be cooked to avoid page becoming writable when making page table present or sync/prefetch shadow page table entries Signed-off-by: Xiao Guangrong --- arch/x86/include/asm/kvm_page_track.h | 2 ++ arch/x86/kvm/mmu.c | 44 +++++++++++++++++++++++++++++------ arch/x86/kvm/page_track.c | 14 +++++++++++ arch/x86/kvm/paging_tmpl.h | 3 +++ 4 files changed, 56 insertions(+), 7 deletions(-) diff --git a/arch/x86/include/asm/kvm_page_track.h b/arch/x86/include/asm/kvm_page_track.h index c010124..97ac9c3 100644 --- a/arch/x86/include/asm/kvm_page_track.h +++ b/arch/x86/include/asm/kvm_page_track.h @@ -23,4 +23,6 @@ void kvm_slot_page_track_remove_page_nolock(struct kvm *kvm, enum kvm_page_track_mode mode); void kvm_page_track_remove_page(struct kvm *kvm, gfn_t gfn, enum kvm_page_track_mode mode); +bool kvm_page_track_check_mode(struct kvm_vcpu *vcpu, gfn_t gfn, + enum kvm_page_track_mode mode); #endif diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 39809b8..b23f9fc 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -41,6 +41,7 @@ #include #include #include +#include /* * When setting this variable to true it enables Two-Dimensional-Paging @@ -2456,25 +2457,29 @@ static void kvm_unsync_pages(struct kvm_vcpu *vcpu, gfn_t gfn) } } -static int mmu_need_write_protect(struct kvm_vcpu *vcpu, gfn_t gfn, - bool can_unsync) +static bool mmu_need_write_protect(struct kvm_vcpu *vcpu, gfn_t gfn, + bool can_unsync) { struct kvm_mmu_page *s; bool need_unsync = false; + if (kvm_page_track_check_mode(vcpu, gfn, KVM_PAGE_TRACK_WRITE)) + return true; + for_each_gfn_indirect_valid_sp(vcpu->kvm, s, gfn) { if (!can_unsync) - return 1; + return true; if (s->role.level != PT_PAGE_TABLE_LEVEL) - return 1; + return true; if (!s->unsync) need_unsync = true; } if (need_unsync) kvm_unsync_pages(vcpu, gfn); - return 0; + + return false; } static bool kvm_is_mmio_pfn(pfn_t pfn) @@ -3388,10 +3393,30 @@ int handle_mmio_page_fault(struct kvm_vcpu *vcpu, u64 addr, bool direct) } EXPORT_SYMBOL_GPL(handle_mmio_page_fault); +static bool page_fault_handle_page_track(struct kvm_vcpu *vcpu, + u32 error_code, gfn_t gfn) +{ + if (unlikely(error_code & PFERR_RSVD_MASK)) + return false; + + if (!(error_code & PFERR_PRESENT_MASK) || + !(error_code & PFERR_WRITE_MASK)) + return false; + + /* + * guest is writing the page which is write tracked which can + * not be fixed by page fault handler. + */ + if (kvm_page_track_check_mode(vcpu, gfn, KVM_PAGE_TRACK_WRITE)) + return true; + + return false; +} + static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva, u32 error_code, bool prefault) { - gfn_t gfn; + gfn_t gfn = gva >> PAGE_SHIFT; int r; pgprintk("%s: gva %lx error %x\n", __func__, gva, error_code); @@ -3403,13 +3428,15 @@ static int nonpaging_page_fault(struct kvm_vcpu *vcpu, gva_t gva, return r; } + if (page_fault_handle_page_track(vcpu, error_code, gfn)) + return 1; + r = mmu_topup_memory_caches(vcpu); if (r) return r; MMU_WARN_ON(!VALID_PAGE(vcpu->arch.mmu.root_hpa)); - gfn = gva >> PAGE_SHIFT; return nonpaging_map(vcpu, gva & PAGE_MASK, error_code, gfn, prefault); @@ -3493,6 +3520,9 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code, return r; } + if (page_fault_handle_page_track(vcpu, error_code, gfn)) + return 1; + r = mmu_topup_memory_caches(vcpu); if (r) return r; diff --git a/arch/x86/kvm/page_track.c b/arch/x86/kvm/page_track.c index e17efe9..de9b32f 100644 --- a/arch/x86/kvm/page_track.c +++ b/arch/x86/kvm/page_track.c @@ -174,3 +174,17 @@ void kvm_page_track_remove_page(struct kvm *kvm, gfn_t gfn, spin_unlock(&kvm->mmu_lock); } } + +/* + * check if the corresponding access on the specified guest page is tracked. + */ +bool kvm_page_track_check_mode(struct kvm_vcpu *vcpu, gfn_t gfn, + enum kvm_page_track_mode mode) +{ + struct kvm_memory_slot *slot = kvm_vcpu_gfn_to_memslot(vcpu, gfn); + int index = gfn_to_index(gfn, slot->base_gfn, PT_PAGE_TABLE_LEVEL); + + WARN_ON(!check_mode(mode)); + + return !!ACCESS_ONCE(slot->arch.gfn_track[mode][index]); +} diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index 91e939b..ac85682 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -735,6 +735,9 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, u32 error_code, return 0; } + if (page_fault_handle_page_track(vcpu, error_code, walker.gfn)) + return 1; + vcpu->arch.write_fault_to_shadow_pgtable = false; is_self_change_mapping = FNAME(is_self_change_mapping)(vcpu,