From patchwork Tue Jun 7 13:00:30 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Xiao Guangrong X-Patchwork-Id: 856082 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by demeter2.kernel.org (8.14.4/8.14.3) with ESMTP id p57Cwmwn014547 for ; Tue, 7 Jun 2011 12:58:48 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753714Ab1FGM61 (ORCPT ); Tue, 7 Jun 2011 08:58:27 -0400 Received: from cn.fujitsu.com ([222.73.24.84]:49618 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1753688Ab1FGM60 (ORCPT ); Tue, 7 Jun 2011 08:58:26 -0400 Received: from tang.cn.fujitsu.com (tang.cn.fujitsu.com [10.167.250.3]) by song.cn.fujitsu.com (Postfix) with ESMTP id AE18A1700BD; Tue, 7 Jun 2011 20:58:23 +0800 (CST) Received: from mailserver.fnst.cn.fujitsu.com (tang.cn.fujitsu.com [127.0.0.1]) by tang.cn.fujitsu.com (8.14.3/8.13.1) with ESMTP id p57CwNCd012495; Tue, 7 Jun 2011 20:58:23 +0800 Received: from eric.localdomain ([10.167.225.99]) by mailserver.fnst.cn.fujitsu.com (Lotus Domino Release 8.5.1FP4) with ESMTP id 2011060720581339-501811 ; Tue, 7 Jun 2011 20:58:13 +0800 Message-ID: <4DEE20EE.7060000@cn.fujitsu.com> Date: Tue, 07 Jun 2011 21:00:30 +0800 From: Xiao Guangrong User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.15) Gecko/20110307 Fedora/3.1.9-0.39.b3pre.fc14 Thunderbird/3.1.9 MIME-Version: 1.0 To: Avi Kivity CC: Marcelo Tosatti , LKML , KVM Subject: [PATCH 04/15] KVM: MMU: cache mmio info on page fault path References: <4DEE205E.8000601@cn.fujitsu.com> In-Reply-To: <4DEE205E.8000601@cn.fujitsu.com> X-MIMETrack: Itemize by SMTP Server on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-06-07 20:58:13, Serialize by Router on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-06-07 20:58:13, Serialize complete at 2011-06-07 20:58:13 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org X-Greylist: IP, sender and recipient auto-whitelisted, not delayed by milter-greylist-4.2.6 (demeter2.kernel.org [140.211.167.43]); Tue, 07 Jun 2011 12:58:48 +0000 (UTC) If the page fault is caused by mmio, we can cache the mmio info, later, we do not need to walk guest page table and quickly know it is a mmio fault while we emulate the mmio instruction Signed-off-by: Xiao Guangrong --- arch/x86/include/asm/kvm_host.h | 5 +++ arch/x86/kvm/mmu.c | 21 +++++---------- arch/x86/kvm/mmu.h | 23 +++++++++++++++++ arch/x86/kvm/paging_tmpl.h | 21 ++++++++++----- arch/x86/kvm/x86.c | 52 ++++++++++++++++++++++++++++++-------- arch/x86/kvm/x86.h | 36 +++++++++++++++++++++++++++ 6 files changed, 126 insertions(+), 32 deletions(-) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index d167039..326af42 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -414,6 +414,11 @@ struct kvm_vcpu_arch { u64 mcg_ctl; u64 *mce_banks; + /* Cache MMIO info */ + u64 mmio_gva; + unsigned access; + gfn_t mmio_gfn; + /* used for guest single stepping over the given code position */ unsigned long singlestep_rip; diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 125f78d..415030e 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -217,11 +217,6 @@ void kvm_mmu_set_mask_ptes(u64 user_mask, u64 accessed_mask, } EXPORT_SYMBOL_GPL(kvm_mmu_set_mask_ptes); -static bool is_write_protection(struct kvm_vcpu *vcpu) -{ - return kvm_read_cr0_bits(vcpu, X86_CR0_WP); -} - static int is_cpuid_PSE36(void) { return 1; @@ -243,11 +238,6 @@ static int is_large_pte(u64 pte) return pte & PT_PAGE_SIZE_MASK; } -static int is_writable_pte(unsigned long pte) -{ - return pte & PT_WRITABLE_MASK; -} - static int is_dirty_gpte(unsigned long pte) { return pte & PT_DIRTY_MASK; @@ -2238,15 +2228,17 @@ static void kvm_send_hwpoison_signal(unsigned long address, struct task_struct * send_sig_info(SIGBUS, &info, tsk); } -static int kvm_handle_bad_page(struct kvm *kvm, gfn_t gfn, pfn_t pfn) +static int kvm_handle_bad_page(struct kvm_vcpu *vcpu, gva_t gva, + unsigned access, gfn_t gfn, pfn_t pfn) { kvm_release_pfn_clean(pfn); if (is_hwpoison_pfn(pfn)) { - kvm_send_hwpoison_signal(gfn_to_hva(kvm, gfn), current); + kvm_send_hwpoison_signal(gfn_to_hva(vcpu->kvm, gfn), current); return 0; } else if (is_fault_pfn(pfn)) return -EFAULT; + vcpu_cache_mmio_info(vcpu, gva, gfn, access); return 1; } @@ -2328,7 +2320,7 @@ static int nonpaging_map(struct kvm_vcpu *vcpu, gva_t v, int write, gfn_t gfn, /* mmio */ if (is_error_pfn(pfn)) - return kvm_handle_bad_page(vcpu->kvm, gfn, pfn); + return kvm_handle_bad_page(vcpu, v, ACC_ALL, gfn, pfn); spin_lock(&vcpu->kvm->mmu_lock); if (mmu_notifier_retry(vcpu, mmu_seq)) @@ -2555,6 +2547,7 @@ static void mmu_sync_roots(struct kvm_vcpu *vcpu) if (!VALID_PAGE(vcpu->arch.mmu.root_hpa)) return; + vcpu_clear_mmio_info(vcpu, ~0ull); trace_kvm_mmu_audit(vcpu, AUDIT_PRE_SYNC); if (vcpu->arch.mmu.root_level == PT64_ROOT_LEVEL) { hpa_t root = vcpu->arch.mmu.root_hpa; @@ -2701,7 +2694,7 @@ static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa, u32 error_code, /* mmio */ if (is_error_pfn(pfn)) - return kvm_handle_bad_page(vcpu->kvm, gfn, pfn); + return kvm_handle_bad_page(vcpu, 0, 0, gfn, pfn); spin_lock(&vcpu->kvm->mmu_lock); if (mmu_notifier_retry(vcpu, mmu_seq)) goto out_unlock; diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h index 7086ca8..05310b1 100644 --- a/arch/x86/kvm/mmu.h +++ b/arch/x86/kvm/mmu.h @@ -76,4 +76,27 @@ static inline int is_present_gpte(unsigned long pte) return pte & PT_PRESENT_MASK; } +static inline int is_writable_pte(unsigned long pte) +{ + return pte & PT_WRITABLE_MASK; +} + +static inline bool is_write_protection(struct kvm_vcpu *vcpu) +{ + return kvm_read_cr0_bits(vcpu, X86_CR0_WP); +} + +static inline bool check_write_user_access(struct kvm_vcpu *vcpu, + bool write_fault, bool user_fault, + unsigned long pte) +{ + if (unlikely(write_fault && !is_writable_pte(pte) + && (user_fault || is_write_protection(vcpu)))) + return false; + + if (unlikely(user_fault && !(pte & PT_USER_MASK))) + return false; + + return true; +} #endif diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index 6c4dc01..b0c8184 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -201,11 +201,8 @@ walk: break; } - if (unlikely(write_fault && !is_writable_pte(pte) - && (user_fault || is_write_protection(vcpu)))) - eperm = true; - - if (unlikely(user_fault && !(pte & PT_USER_MASK))) + if (!check_write_user_access(vcpu, write_fault, user_fault, + pte)) eperm = true; #if PTTYPE == 64 @@ -624,8 +621,16 @@ static int FNAME(page_fault)(struct kvm_vcpu *vcpu, gva_t addr, u32 error_code, return 0; /* mmio */ - if (is_error_pfn(pfn)) - return kvm_handle_bad_page(vcpu->kvm, walker.gfn, pfn); + if (is_error_pfn(pfn)) { + unsigned access = walker.pte_access; + bool dirty = is_dirty_gpte(walker.ptes[walker.level - 1]); + + if (dirty) + access &= ~ACC_WRITE_MASK; + + return kvm_handle_bad_page(vcpu, mmu_is_nested(vcpu) ? 0 : + addr, access, walker.gfn, pfn); + } spin_lock(&vcpu->kvm->mmu_lock); if (mmu_notifier_retry(vcpu, mmu_seq)) @@ -665,6 +670,8 @@ static void FNAME(invlpg)(struct kvm_vcpu *vcpu, gva_t gva) u64 *sptep; int need_flush = 0; + vcpu_clear_mmio_info(vcpu, gva); + spin_lock(&vcpu->kvm->mmu_lock); for_each_shadow_entry(vcpu, gva, iterator) { diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 8be9ff6..a136181 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3903,6 +3903,38 @@ out: } EXPORT_SYMBOL_GPL(kvm_write_guest_virt_system); +static int vcpu_gva_to_gpa(struct kvm_vcpu *vcpu, unsigned long gva, + gpa_t *gpa, struct x86_exception *exception, + bool write) +{ + u32 access = (kvm_x86_ops->get_cpl(vcpu) == 3) ? PFERR_USER_MASK : 0; + + if (vcpu_match_mmio_gva(vcpu, gva) && + check_write_user_access(vcpu, write, access, + vcpu->arch.access)) { + *gpa = vcpu->arch.mmio_gfn << PAGE_SHIFT | + (gva & (PAGE_SIZE - 1)); + return 1; + } + + if (write) + access |= PFERR_WRITE_MASK; + + *gpa = vcpu->arch.walk_mmu->gva_to_gpa(vcpu, gva, access, exception); + + if (*gpa == UNMAPPED_GVA) + return -1; + + /* For APIC access vmexit */ + if ((*gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE) + return 1; + + if (vcpu_match_mmio_gpa(vcpu, *gpa)) + return 1; + + return 0; +} + static int emulator_read_emulated(struct x86_emulate_ctxt *ctxt, unsigned long addr, void *val, @@ -3911,7 +3943,7 @@ static int emulator_read_emulated(struct x86_emulate_ctxt *ctxt, { struct kvm_vcpu *vcpu = emul_to_vcpu(ctxt); gpa_t gpa; - int handled; + int handled, ret; if (vcpu->mmio_read_completed) { memcpy(val, vcpu->mmio_data, bytes); @@ -3921,13 +3953,12 @@ static int emulator_read_emulated(struct x86_emulate_ctxt *ctxt, return X86EMUL_CONTINUE; } - gpa = kvm_mmu_gva_to_gpa_read(vcpu, addr, exception); + ret = vcpu_gva_to_gpa(vcpu, addr, &gpa, exception, false); - if (gpa == UNMAPPED_GVA) + if (ret < 0) return X86EMUL_PROPAGATE_FAULT; - /* For APIC access vmexit */ - if ((gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE) + if (ret) goto mmio; if (!kvm_read_guest(vcpu->kvm, gpa, val, bytes)) @@ -3977,16 +4008,15 @@ static int emulator_write_emulated_onepage(unsigned long addr, struct x86_exception *exception, struct kvm_vcpu *vcpu) { - gpa_t gpa; - int handled; + gpa_t gpa; + int handled, ret; - gpa = kvm_mmu_gva_to_gpa_write(vcpu, addr, exception); + ret = vcpu_gva_to_gpa(vcpu, addr, &gpa, exception, true); - if (gpa == UNMAPPED_GVA) + if (ret < 0) return X86EMUL_PROPAGATE_FAULT; - /* For APIC access vmexit */ - if ((gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE) + if (ret) goto mmio; if (emulator_write_phys(vcpu, gpa, val, bytes)) diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h index 256da82..d36fe23 100644 --- a/arch/x86/kvm/x86.h +++ b/arch/x86/kvm/x86.h @@ -75,6 +75,42 @@ static inline u32 bit(int bitno) return 1 << (bitno & 31); } +static inline void vcpu_cache_mmio_info(struct kvm_vcpu *vcpu, + gva_t gva, gfn_t gfn, unsigned access) +{ + vcpu->arch.mmio_gva = gva & PAGE_MASK; + vcpu->arch.access = access; + vcpu->arch.mmio_gfn = gfn; +} + +/* + * Clear the mmio cache info for the given gva, + * specially, if gva is ~0ul, we clear all mmio cache info. + */ +static inline void vcpu_clear_mmio_info(struct kvm_vcpu *vcpu, gva_t gva) +{ + if (gva != (~0ul) && vcpu->arch.mmio_gva != (gva & PAGE_MASK)) + return; + + vcpu->arch.mmio_gva = 0; +} + +static inline bool vcpu_match_mmio_gva(struct kvm_vcpu *vcpu, unsigned long gva) +{ + if (vcpu->arch.mmio_gva && vcpu->arch.mmio_gva == (gva & PAGE_MASK)) + return true; + + return false; +} + +static inline bool vcpu_match_mmio_gpa(struct kvm_vcpu *vcpu, gpa_t gpa) +{ + if (vcpu->arch.mmio_gfn && vcpu->arch.mmio_gfn == gpa >> PAGE_SHIFT) + return true; + + return false; +} + void kvm_before_handle_nmi(struct kvm_vcpu *vcpu); void kvm_after_handle_nmi(struct kvm_vcpu *vcpu); int kvm_inject_realmode_interrupt(struct kvm_vcpu *vcpu, int irq, int inc_eip);