Merge commit 'v2.6.27-rc1' into x86/core

Conflicts:

	include/asm-x86/dma-mapping.h
	include/asm-x86/namei.h
	include/asm-x86/uaccess.h

Signed-off-by: Ingo Molnar <mingo@elte.hu>

5 files changed, 71 insertions, 78 deletions

diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
index 8d45fabc5f3..ce3251ce550 100644
--- a/arch/x86/kvm/Kconfig
+++ b/arch/x86/kvm/Kconfig
@@ -21,6 +21,7 @@ config KVM
 	tristate "Kernel-based Virtual Machine (KVM) support"
 	depends on HAVE_KVM
 	select PREEMPT_NOTIFIERS
+	select MMU_NOTIFIER
 	select ANON_INODES
 	---help---
 	  Support hosting fully virtualized guest machines using hardware
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index b0e4ddca6c1..2fa231923cf 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -1814,6 +1814,7 @@ int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
 	spin_unlock(&vcpu->kvm->mmu_lock);
 	return r;
 }
+EXPORT_SYMBOL_GPL(kvm_mmu_unprotect_page_virt);
 
 void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu)
 {
@@ -1870,6 +1871,12 @@ void kvm_enable_tdp(void)
 }
 EXPORT_SYMBOL_GPL(kvm_enable_tdp);
 
+void kvm_disable_tdp(void)
+{
+	tdp_enabled = false;
+}
+EXPORT_SYMBOL_GPL(kvm_disable_tdp);
+
 static void free_mmu_pages(struct kvm_vcpu *vcpu)
 {
 	struct kvm_mmu_page *sp;
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index b756e876dce..e2ee264740c 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -453,7 +453,8 @@ static __init int svm_hardware_setup(void)
 	if (npt_enabled) {
 		printk(KERN_INFO "kvm: Nested Paging enabled\n");
 		kvm_enable_tdp();
-	}
+	} else
+		kvm_disable_tdp();
 
 	return 0;
 
@@ -1007,10 +1008,13 @@ static int pf_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 	struct kvm *kvm = svm->vcpu.kvm;
 	u64 fault_address;
 	u32 error_code;
+	bool event_injection = false;
 
 	if (!irqchip_in_kernel(kvm) &&
-		is_external_interrupt(exit_int_info))
+	    is_external_interrupt(exit_int_info)) {
+		event_injection = true;
 		push_irq(&svm->vcpu, exit_int_info & SVM_EVTINJ_VEC_MASK);
+	}
 
 	fault_address  = svm->vmcb->control.exit_info_2;
 	error_code = svm->vmcb->control.exit_info_1;
@@ -1024,6 +1028,8 @@ static int pf_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 			    (u32)fault_address, (u32)(fault_address >> 32),
 			    handler);
 
+	if (event_injection)
+		kvm_mmu_unprotect_page_virt(&svm->vcpu, fault_address);
 	return kvm_mmu_page_fault(&svm->vcpu, fault_address, error_code);
 }
 
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 0cac6370171..2a69773e3b2 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -2298,6 +2298,8 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		cr2 = vmcs_readl(EXIT_QUALIFICATION);
 		KVMTRACE_3D(PAGE_FAULT, vcpu, error_code, (u32)cr2,
 			    (u32)((u64)cr2 >> 32), handler);
+		if (vect_info & VECTORING_INFO_VALID_MASK)
+			kvm_mmu_unprotect_page_virt(vcpu, cr2);
 		return kvm_mmu_page_fault(vcpu, cr2, error_code);
 	}
 
@@ -3116,15 +3118,6 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 		return ERR_PTR(-ENOMEM);
 
 	allocate_vpid(vmx);
-	if (id == 0 && vm_need_ept()) {
-		kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK |
-			VMX_EPT_WRITABLE_MASK |
-			VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT);
-		kvm_mmu_set_mask_ptes(0ull, VMX_EPT_FAKE_ACCESSED_MASK,
-				VMX_EPT_FAKE_DIRTY_MASK, 0ull,
-				VMX_EPT_EXECUTABLE_MASK);
-		kvm_enable_tdp();
-	}
 
 	err = kvm_vcpu_init(&vmx->vcpu, kvm, id);
 	if (err)
@@ -3303,8 +3296,17 @@ static int __init vmx_init(void)
 	vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_ESP);
 	vmx_disable_intercept_for_msr(vmx_msr_bitmap, MSR_IA32_SYSENTER_EIP);
 
-	if (cpu_has_vmx_ept())
+	if (vm_need_ept()) {
 		bypass_guest_pf = 0;
+		kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK |
+			VMX_EPT_WRITABLE_MASK |
+			VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT);
+		kvm_mmu_set_mask_ptes(0ull, VMX_EPT_FAKE_ACCESSED_MASK,
+				VMX_EPT_FAKE_DIRTY_MASK, 0ull,
+				VMX_EPT_EXECUTABLE_MASK);
+		kvm_enable_tdp();
+	} else
+		kvm_disable_tdp();
 
 	if (bypass_guest_pf)
 		kvm_mmu_set_nonpresent_ptes(~0xffeull, 0ull);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9f1cdb011cf..5916191420c 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3184,6 +3184,10 @@ static void seg_desct_to_kvm_desct(struct desc_struct *seg_desc, u16 selector,
 	kvm_desct->base |= seg_desc->base2 << 24;
 	kvm_desct->limit = seg_desc->limit0;
 	kvm_desct->limit |= seg_desc->limit << 16;
+	if (seg_desc->g) {
+		kvm_desct->limit <<= 12;
+		kvm_desct->limit |= 0xfff;
+	}
 	kvm_desct->selector = selector;
 	kvm_desct->type = seg_desc->type;
 	kvm_desct->present = seg_desc->p;
@@ -3223,6 +3227,7 @@ static void get_segment_descritptor_dtable(struct kvm_vcpu *vcpu,
 static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 					 struct desc_struct *seg_desc)
 {
+	gpa_t gpa;
 	struct descriptor_table dtable;
 	u16 index = selector >> 3;
 
@@ -3232,13 +3237,16 @@ static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 		kvm_queue_exception_e(vcpu, GP_VECTOR, selector & 0xfffc);
 		return 1;
 	}
-	return kvm_read_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8);
+	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, dtable.base);
+	gpa += index * 8;
+	return kvm_read_guest(vcpu->kvm, gpa, seg_desc, 8);
 }
 
 /* allowed just for 8 bytes segments */
 static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 					 struct desc_struct *seg_desc)
 {
+	gpa_t gpa;
 	struct descriptor_table dtable;
 	u16 index = selector >> 3;
 
@@ -3246,7 +3254,9 @@ static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 
 	if (dtable.limit < index * 8 + 7)
 		return 1;
-	return kvm_write_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8);
+	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, dtable.base);
+	gpa += index * 8;
+	return kvm_write_guest(vcpu->kvm, gpa, seg_desc, 8);
 }
 
 static u32 get_tss_base_addr(struct kvm_vcpu *vcpu,
@@ -3258,55 +3268,7 @@ static u32 get_tss_base_addr(struct kvm_vcpu *vcpu,
 	base_addr |= (seg_desc->base1 << 16);
 	base_addr |= (seg_desc->base2 << 24);
 
-	return base_addr;
-}
-
-static int load_tss_segment32(struct kvm_vcpu *vcpu,
-			      struct desc_struct *seg_desc,
-			      struct tss_segment_32 *tss)
-{
-	u32 base_addr;
-
-	base_addr = get_tss_base_addr(vcpu, seg_desc);
-
-	return kvm_read_guest(vcpu->kvm, base_addr, tss,
-			      sizeof(struct tss_segment_32));
-}
-
-static int save_tss_segment32(struct kvm_vcpu *vcpu,
-			      struct desc_struct *seg_desc,
-			      struct tss_segment_32 *tss)
-{
-	u32 base_addr;
-
-	base_addr = get_tss_base_addr(vcpu, seg_desc);
-
-	return kvm_write_guest(vcpu->kvm, base_addr, tss,
-			       sizeof(struct tss_segment_32));
-}
-
-static int load_tss_segment16(struct kvm_vcpu *vcpu,
-			      struct desc_struct *seg_desc,
-			      struct tss_segment_16 *tss)
-{
-	u32 base_addr;
-
-	base_addr = get_tss_base_addr(vcpu, seg_desc);
-
-	return kvm_read_guest(vcpu->kvm, base_addr, tss,
-			      sizeof(struct tss_segment_16));
-}
-
-static int save_tss_segment16(struct kvm_vcpu *vcpu,
-			      struct desc_struct *seg_desc,
-			      struct tss_segment_16 *tss)
-{
-	u32 base_addr;
-
-	base_addr = get_tss_base_addr(vcpu, seg_desc);
-
-	return kvm_write_guest(vcpu->kvm, base_addr, tss,
-			       sizeof(struct tss_segment_16));
+	return vcpu->arch.mmu.gva_to_gpa(vcpu, base_addr);
 }
 
 static u16 get_segment_selector(struct kvm_vcpu *vcpu, int seg)
@@ -3466,20 +3428,26 @@ static int load_state_from_tss16(struct kvm_vcpu *vcpu,
 }
 
 static int kvm_task_switch_16(struct kvm_vcpu *vcpu, u16 tss_selector,
-		       struct desc_struct *cseg_desc,
+		       u32 old_tss_base,
 		       struct desc_struct *nseg_desc)
 {
 	struct tss_segment_16 tss_segment_16;
 	int ret = 0;
 
-	if (load_tss_segment16(vcpu, cseg_desc, &tss_segment_16))
+	if (kvm_read_guest(vcpu->kvm, old_tss_base, &tss_segment_16,
+			   sizeof tss_segment_16))
 		goto out;
 
 	save_state_to_tss16(vcpu, &tss_segment_16);
-	save_tss_segment16(vcpu, cseg_desc, &tss_segment_16);
 
-	if (load_tss_segment16(vcpu, nseg_desc, &tss_segment_16))
+	if (kvm_write_guest(vcpu->kvm, old_tss_base, &tss_segment_16,
+			    sizeof tss_segment_16))
 		goto out;
+
+	if (kvm_read_guest(vcpu->kvm, get_tss_base_addr(vcpu, nseg_desc),
+			   &tss_segment_16, sizeof tss_segment_16))
+		goto out;
+
 	if (load_state_from_tss16(vcpu, &tss_segment_16))
 		goto out;
 
@@ -3489,20 +3457,26 @@ out:
 }
 
 static int kvm_task_switch_32(struct kvm_vcpu *vcpu, u16 tss_selector,
-		       struct desc_struct *cseg_desc,
+		       u32 old_tss_base,
 		       struct desc_struct *nseg_desc)
 {
 	struct tss_segment_32 tss_segment_32;
 	int ret = 0;
 
-	if (load_tss_segment32(vcpu, cseg_desc, &tss_segment_32))
+	if (kvm_read_guest(vcpu->kvm, old_tss_base, &tss_segment_32,
+			   sizeof tss_segment_32))
 		goto out;
 
 	save_state_to_tss32(vcpu, &tss_segment_32);
-	save_tss_segment32(vcpu, cseg_desc, &tss_segment_32);
 
-	if (load_tss_segment32(vcpu, nseg_desc, &tss_segment_32))
+	if (kvm_write_guest(vcpu->kvm, old_tss_base, &tss_segment_32,
+			    sizeof tss_segment_32))
+		goto out;
+
+	if (kvm_read_guest(vcpu->kvm, get_tss_base_addr(vcpu, nseg_desc),
+			   &tss_segment_32, sizeof tss_segment_32))
 		goto out;
+
 	if (load_state_from_tss32(vcpu, &tss_segment_32))
 		goto out;
 
@@ -3517,16 +3491,20 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason)
 	struct desc_struct cseg_desc;
 	struct desc_struct nseg_desc;
 	int ret = 0;
+	u32 old_tss_base = get_segment_base(vcpu, VCPU_SREG_TR);
+	u16 old_tss_sel = get_segment_selector(vcpu, VCPU_SREG_TR);
 
-	kvm_get_segment(vcpu, &tr_seg, VCPU_SREG_TR);
+	old_tss_base = vcpu->arch.mmu.gva_to_gpa(vcpu, old_tss_base);
 
+	/* FIXME: Handle errors. Failure to read either TSS or their
+	 * descriptors should generate a pagefault.
+	 */
 	if (load_guest_segment_descriptor(vcpu, tss_selector, &nseg_desc))
 		goto out;
 
-	if (load_guest_segment_descriptor(vcpu, tr_seg.selector, &cseg_desc))
+	if (load_guest_segment_descriptor(vcpu, old_tss_sel, &cseg_desc))
 		goto out;
 
-
 	if (reason != TASK_SWITCH_IRET) {
 		int cpl;
 
@@ -3544,8 +3522,7 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason)
 
 	if (reason == TASK_SWITCH_IRET || reason == TASK_SWITCH_JMP) {
 		cseg_desc.type &= ~(1 << 1); //clear the B flag
-		save_guest_segment_descriptor(vcpu, tr_seg.selector,
-					      &cseg_desc);
+		save_guest_segment_descriptor(vcpu, old_tss_sel, &cseg_desc);
 	}
 
 	if (reason == TASK_SWITCH_IRET) {
@@ -3557,10 +3534,10 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason)
 	kvm_x86_ops->cache_regs(vcpu);
 
 	if (nseg_desc.type & 8)
-		ret = kvm_task_switch_32(vcpu, tss_selector, &cseg_desc,
+		ret = kvm_task_switch_32(vcpu, tss_selector, old_tss_base,
 					 &nseg_desc);
 	else
-		ret = kvm_task_switch_16(vcpu, tss_selector, &cseg_desc,
+		ret = kvm_task_switch_16(vcpu, tss_selector, old_tss_base,
 					 &nseg_desc);
 
 	if (reason == TASK_SWITCH_CALL || reason == TASK_SWITCH_GATE) {