Merge branch 'kvm-updates/2.6.28' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm

* 'kvm-updates/2.6.28' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm: (134 commits)
  KVM: ia64: Add intel iommu support for guests.
  KVM: ia64: add directed mmio range support for kvm guests
  KVM: ia64: Make pmt table be able to hold physical mmio entries.
  KVM: Move irqchip_in_kernel() from ioapic.h to irq.h
  KVM: Separate irq ack notification out of arch/x86/kvm/irq.c
  KVM: Change is_mmio_pfn to kvm_is_mmio_pfn, and make it common for all archs
  KVM: Move device assignment logic to common code
  KVM: Device Assignment: Move vtd.c from arch/x86/kvm/ to virt/kvm/
  KVM: VMX: enable invlpg exiting if EPT is disabled
  KVM: x86: Silence various LAPIC-related host kernel messages
  KVM: Device Assignment: Map mmio pages into VT-d page table
  KVM: PIC: enhance IPI avoidance
  KVM: MMU: add "oos_shadow" parameter to disable oos
  KVM: MMU: speed up mmu_unsync_walk
  KVM: MMU: out of sync shadow core
  KVM: MMU: mmu_convert_notrap helper
  KVM: MMU: awareness of new kvm_mmu_zap_page behaviour
  KVM: MMU: mmu_parent_walk
  KVM: x86: trap invlpg
  KVM: MMU: sync roots on mmu reload
  ...

10 files changed, 246 insertions, 57 deletions

diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h
index 2655e2a4831..34b52b7180c 100644
--- a/arch/powerpc/include/asm/kvm_host.h
+++ b/arch/powerpc/include/asm/kvm_host.h
@@ -81,11 +81,17 @@ struct kvm_vcpu_arch {
 	struct tlbe shadow_tlb[PPC44x_TLB_SIZE];
 	/* Pages which are referenced in the shadow TLB. */
 	struct page *shadow_pages[PPC44x_TLB_SIZE];
-	/* Copy of the host's TLB. */
-	struct tlbe host_tlb[PPC44x_TLB_SIZE];
+
+	/* Track which TLB entries we've modified in the current exit. */
+	u8 shadow_tlb_mod[PPC44x_TLB_SIZE];
 
 	u32 host_stack;
 	u32 host_pid;
+	u32 host_dbcr0;
+	u32 host_dbcr1;
+	u32 host_dbcr2;
+	u32 host_iac[4];
+	u32 host_msr;
 
 	u64 fpr[32];
 	u32 gpr[32];
@@ -123,7 +129,11 @@ struct kvm_vcpu_arch {
 	u32 ivor[16];
 	u32 ivpr;
 	u32 pir;
+
+	u32 shadow_pid;
 	u32 pid;
+	u32 swap_pid;
+
 	u32 pvr;
 	u32 ccr0;
 	u32 ccr1;
diff --git a/arch/powerpc/include/asm/kvm_ppc.h b/arch/powerpc/include/asm/kvm_ppc.h
index a8b06879226..8931ba729d2 100644
--- a/arch/powerpc/include/asm/kvm_ppc.h
+++ b/arch/powerpc/include/asm/kvm_ppc.h
@@ -64,6 +64,10 @@ extern void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn,
 extern void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, gva_t eaddr,
                                   gva_t eend, u32 asid);
 extern void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode);
+extern void kvmppc_mmu_switch_pid(struct kvm_vcpu *vcpu, u32 pid);
+
+/* XXX Book E specific */
+extern void kvmppc_tlbe_set_modified(struct kvm_vcpu *vcpu, unsigned int i);
 
 extern void kvmppc_check_and_deliver_interrupts(struct kvm_vcpu *vcpu);
 
@@ -92,4 +96,12 @@ static inline void kvmppc_set_msr(struct kvm_vcpu *vcpu, u32 new_msr)
 		kvm_vcpu_block(vcpu);
 }
 
+static inline void kvmppc_set_pid(struct kvm_vcpu *vcpu, u32 new_pid)
+{
+	if (vcpu->arch.pid != new_pid) {
+		vcpu->arch.pid = new_pid;
+		vcpu->arch.swap_pid = 1;
+	}
+}
+
 #endif /* __POWERPC_KVM_PPC_H__ */
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 09febc58258..75c5dd0138f 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -359,8 +359,8 @@ int main(void)
 
 	DEFINE(VCPU_HOST_STACK, offsetof(struct kvm_vcpu, arch.host_stack));
 	DEFINE(VCPU_HOST_PID, offsetof(struct kvm_vcpu, arch.host_pid));
-	DEFINE(VCPU_HOST_TLB, offsetof(struct kvm_vcpu, arch.host_tlb));
 	DEFINE(VCPU_SHADOW_TLB, offsetof(struct kvm_vcpu, arch.shadow_tlb));
+	DEFINE(VCPU_SHADOW_MOD, offsetof(struct kvm_vcpu, arch.shadow_tlb_mod));
 	DEFINE(VCPU_GPRS, offsetof(struct kvm_vcpu, arch.gpr));
 	DEFINE(VCPU_LR, offsetof(struct kvm_vcpu, arch.lr));
 	DEFINE(VCPU_CR, offsetof(struct kvm_vcpu, arch.cr));
@@ -372,7 +372,7 @@ int main(void)
 	DEFINE(VCPU_SPRG5, offsetof(struct kvm_vcpu, arch.sprg5));
 	DEFINE(VCPU_SPRG6, offsetof(struct kvm_vcpu, arch.sprg6));
 	DEFINE(VCPU_SPRG7, offsetof(struct kvm_vcpu, arch.sprg7));
-	DEFINE(VCPU_PID, offsetof(struct kvm_vcpu, arch.pid));
+	DEFINE(VCPU_SHADOW_PID, offsetof(struct kvm_vcpu, arch.shadow_pid));
 
 	DEFINE(VCPU_LAST_INST, offsetof(struct kvm_vcpu, arch.last_inst));
 	DEFINE(VCPU_FAULT_DEAR, offsetof(struct kvm_vcpu, arch.fault_dear));
diff --git a/arch/powerpc/kvm/44x_tlb.c b/arch/powerpc/kvm/44x_tlb.c
index 5a5602da509..2e227a412bc 100644
--- a/arch/powerpc/kvm/44x_tlb.c
+++ b/arch/powerpc/kvm/44x_tlb.c
@@ -19,6 +19,7 @@
 
 #include <linux/types.h>
 #include <linux/string.h>
+#include <linux/kvm.h>
 #include <linux/kvm_host.h>
 #include <linux/highmem.h>
 #include <asm/mmu-44x.h>
@@ -109,7 +110,6 @@ static int kvmppc_44x_tlbe_is_writable(struct tlbe *tlbe)
 	return tlbe->word2 & (PPC44x_TLB_SW|PPC44x_TLB_UW);
 }
 
-/* Must be called with mmap_sem locked for writing. */
 static void kvmppc_44x_shadow_release(struct kvm_vcpu *vcpu,
                                       unsigned int index)
 {
@@ -124,6 +124,11 @@ static void kvmppc_44x_shadow_release(struct kvm_vcpu *vcpu,
 	}
 }
 
+void kvmppc_tlbe_set_modified(struct kvm_vcpu *vcpu, unsigned int i)
+{
+    vcpu->arch.shadow_tlb_mod[i] = 1;
+}
+
 /* Caller must ensure that the specified guest TLB entry is safe to insert into
  * the shadow TLB. */
 void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
@@ -142,19 +147,16 @@ void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
 	stlbe = &vcpu->arch.shadow_tlb[victim];
 
 	/* Get reference to new page. */
-	down_read(&current->mm->mmap_sem);
 	new_page = gfn_to_page(vcpu->kvm, gfn);
 	if (is_error_page(new_page)) {
 		printk(KERN_ERR "Couldn't get guest page for gfn %lx!\n", gfn);
 		kvm_release_page_clean(new_page);
-		up_read(&current->mm->mmap_sem);
 		return;
 	}
 	hpaddr = page_to_phys(new_page);
 
 	/* Drop reference to old page. */
 	kvmppc_44x_shadow_release(vcpu, victim);
-	up_read(&current->mm->mmap_sem);
 
 	vcpu->arch.shadow_pages[victim] = new_page;
 
@@ -164,27 +166,30 @@ void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
 
 	/* XXX what about AS? */
 
-	stlbe->tid = asid & 0xff;
+	stlbe->tid = !(asid & 0xff);
 
 	/* Force TS=1 for all guest mappings. */
 	/* For now we hardcode 4KB mappings, but it will be important to
 	 * use host large pages in the future. */
 	stlbe->word0 = (gvaddr & PAGE_MASK) | PPC44x_TLB_VALID | PPC44x_TLB_TS
 	               | PPC44x_TLB_4K;
-
 	stlbe->word1 = (hpaddr & 0xfffffc00) | ((hpaddr >> 32) & 0xf);
 	stlbe->word2 = kvmppc_44x_tlb_shadow_attrib(flags,
 	                                            vcpu->arch.msr & MSR_PR);
+	kvmppc_tlbe_set_modified(vcpu, victim);
+
+	KVMTRACE_5D(STLB_WRITE, vcpu, victim,
+			stlbe->tid, stlbe->word0, stlbe->word1, stlbe->word2,
+			handler);
 }
 
 void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, gva_t eaddr,
                            gva_t eend, u32 asid)
 {
-	unsigned int pid = asid & 0xff;
+	unsigned int pid = !(asid & 0xff);
 	int i;
 
 	/* XXX Replace loop with fancy data structures. */
-	down_write(&current->mm->mmap_sem);
 	for (i = 0; i <= tlb_44x_hwater; i++) {
 		struct tlbe *stlbe = &vcpu->arch.shadow_tlb[i];
 		unsigned int tid;
@@ -204,21 +209,35 @@ void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, gva_t eaddr,
 
 		kvmppc_44x_shadow_release(vcpu, i);
 		stlbe->word0 = 0;
+		kvmppc_tlbe_set_modified(vcpu, i);
+		KVMTRACE_5D(STLB_INVAL, vcpu, i,
+				stlbe->tid, stlbe->word0, stlbe->word1,
+				stlbe->word2, handler);
 	}
-	up_write(&current->mm->mmap_sem);
 }
 
-/* Invalidate all mappings, so that when they fault back in they will get the
- * proper permission bits. */
+/* Invalidate all mappings on the privilege switch after PID has been changed.
+ * The guest always runs with PID=1, so we must clear the entire TLB when
+ * switching address spaces. */
 void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode)
 {
 	int i;
 
-	/* XXX Replace loop with fancy data structures. */
-	down_write(&current->mm->mmap_sem);
-	for (i = 0; i <= tlb_44x_hwater; i++) {
-		kvmppc_44x_shadow_release(vcpu, i);
-		vcpu->arch.shadow_tlb[i].word0 = 0;
+	if (vcpu->arch.swap_pid) {
+		/* XXX Replace loop with fancy data structures. */
+		for (i = 0; i <= tlb_44x_hwater; i++) {
+			struct tlbe *stlbe = &vcpu->arch.shadow_tlb[i];
+
+			/* Future optimization: clear only userspace mappings. */
+			kvmppc_44x_shadow_release(vcpu, i);
+			stlbe->word0 = 0;
+			kvmppc_tlbe_set_modified(vcpu, i);
+			KVMTRACE_5D(STLB_INVAL, vcpu, i,
+			            stlbe->tid, stlbe->word0, stlbe->word1,
+			            stlbe->word2, handler);
+		}
+		vcpu->arch.swap_pid = 0;
 	}
-	up_write(&current->mm->mmap_sem);
+
+	vcpu->arch.shadow_pid = !usermode;
 }
diff --git a/arch/powerpc/kvm/Kconfig b/arch/powerpc/kvm/Kconfig
index 6b076010213..53aaa66b25e 100644
--- a/arch/powerpc/kvm/Kconfig
+++ b/arch/powerpc/kvm/Kconfig
@@ -37,6 +37,17 @@ config KVM_BOOKE_HOST
 	  Provides host support for KVM on Book E PowerPC processors. Currently
 	  this works on 440 processors only.
 
+config KVM_TRACE
+	bool "KVM trace support"
+	depends on KVM && MARKERS && SYSFS
+	select RELAY
+	select DEBUG_FS
+	default n
+	---help---
+	  This option allows reading a trace of kvm-related events through
+	  relayfs.  Note the ABI is not considered stable and will be
+	  modified in future updates.
+
 source drivers/virtio/Kconfig
 
 endif # VIRTUALIZATION
diff --git a/arch/powerpc/kvm/Makefile b/arch/powerpc/kvm/Makefile
index 04e3449e1f4..2a5d4397ac4 100644
--- a/arch/powerpc/kvm/Makefile
+++ b/arch/powerpc/kvm/Makefile
@@ -4,9 +4,11 @@
 
 EXTRA_CFLAGS += -Ivirt/kvm -Iarch/powerpc/kvm
 
-common-objs = $(addprefix ../../../virt/kvm/, kvm_main.o coalesced_mmio.o)
+common-objs-y = $(addprefix ../../../virt/kvm/, kvm_main.o coalesced_mmio.o)
 
-kvm-objs := $(common-objs) powerpc.o emulate.o booke_guest.o
+common-objs-$(CONFIG_KVM_TRACE)  += $(addprefix ../../../virt/kvm/, kvm_trace.o)
+
+kvm-objs := $(common-objs-y) powerpc.o emulate.o booke_guest.o
 obj-$(CONFIG_KVM) += kvm.o
 
 AFLAGS_booke_interrupts.o := -I$(obj)
diff --git a/arch/powerpc/kvm/booke_guest.c b/arch/powerpc/kvm/booke_guest.c
index 9c8ad850c6e..7b2591e26ba 100644
--- a/arch/powerpc/kvm/booke_guest.c
+++ b/arch/powerpc/kvm/booke_guest.c
@@ -410,6 +410,21 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		break;
 	}
 
+	case BOOKE_INTERRUPT_DEBUG: {
+		u32 dbsr;
+
+		vcpu->arch.pc = mfspr(SPRN_CSRR0);
+
+		/* clear IAC events in DBSR register */
+		dbsr = mfspr(SPRN_DBSR);
+		dbsr &= DBSR_IAC1 | DBSR_IAC2 | DBSR_IAC3 | DBSR_IAC4;
+		mtspr(SPRN_DBSR, dbsr);
+
+		run->exit_reason = KVM_EXIT_DEBUG;
+		r = RESUME_HOST;
+		break;
+	}
+
 	default:
 		printk(KERN_EMERG "exit_nr %d\n", exit_nr);
 		BUG();
@@ -471,6 +486,8 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 	vcpu->arch.msr = 0;
 	vcpu->arch.gpr[1] = (16<<20) - 8; /* -8 for the callee-save LR slot */
 
+	vcpu->arch.shadow_pid = 1;
+
 	/* Eye-catching number so we know if the guest takes an interrupt
 	 * before it's programmed its own IVPR. */
 	vcpu->arch.ivpr = 0x55550000;
diff --git a/arch/powerpc/kvm/booke_interrupts.S b/arch/powerpc/kvm/booke_interrupts.S
index 3b653b5309b..95e165baf85 100644
--- a/arch/powerpc/kvm/booke_interrupts.S
+++ b/arch/powerpc/kvm/booke_interrupts.S
@@ -42,7 +42,8 @@
 #define HOST_STACK_LR   (HOST_STACK_SIZE + 4) /* In caller stack frame. */
 
 #define NEED_INST_MASK ((1<<BOOKE_INTERRUPT_PROGRAM) | \
-                        (1<<BOOKE_INTERRUPT_DTLB_MISS))
+                        (1<<BOOKE_INTERRUPT_DTLB_MISS) | \
+                        (1<<BOOKE_INTERRUPT_DEBUG))
 
 #define NEED_DEAR_MASK ((1<<BOOKE_INTERRUPT_DATA_STORAGE) | \
                         (1<<BOOKE_INTERRUPT_DTLB_MISS))
@@ -331,51 +332,57 @@ lightweight_exit:
 
 	mfspr	r3, SPRN_PID
 	stw	r3, VCPU_HOST_PID(r4)
-	lwz	r3, VCPU_PID(r4)
+	lwz	r3, VCPU_SHADOW_PID(r4)
 	mtspr	SPRN_PID, r3
 
-	/* Prevent all TLB updates. */
+	/* Prevent all asynchronous TLB updates. */
 	mfmsr	r5
 	lis	r6, (MSR_EE|MSR_CE|MSR_ME|MSR_DE)@h
 	ori	r6, r6, (MSR_EE|MSR_CE|MSR_ME|MSR_DE)@l
 	andc	r6, r5, r6
 	mtmsr	r6
 
-	/* Save the host's non-pinned TLB mappings, and load the guest mappings
-	 * over them. Leave the host's "pinned" kernel mappings in place. */
-	/* XXX optimization: use generation count to avoid swapping unmodified
-	 * entries. */
+	/* Load the guest mappings, leaving the host's "pinned" kernel mappings
+	 * in place. */
 	mfspr	r10, SPRN_MMUCR			/* Save host MMUCR. */
-	lis	r8, tlb_44x_hwater@ha
-	lwz	r8, tlb_44x_hwater@l(r8)
-	addi	r3, r4, VCPU_HOST_TLB - 4
-	addi	r9, r4, VCPU_SHADOW_TLB - 4
-	li	r6, 0
+	li	r5, PPC44x_TLB_SIZE
+	lis	r5, tlb_44x_hwater@ha
+	lwz	r5, tlb_44x_hwater@l(r5)
+	mtctr	r5
+	addi	r9, r4, VCPU_SHADOW_TLB
+	addi	r5, r4, VCPU_SHADOW_MOD
+	li	r3, 0
 1:
-	/* Save host entry. */
-	tlbre	r7, r6, PPC44x_TLB_PAGEID
-	mfspr	r5, SPRN_MMUCR
-	stwu	r5, 4(r3)
-	stwu	r7, 4(r3)
-	tlbre	r7, r6, PPC44x_TLB_XLAT
-	stwu	r7, 4(r3)
-	tlbre	r7, r6, PPC44x_TLB_ATTRIB
-	stwu	r7, 4(r3)
+	lbzx	r7, r3, r5
+	cmpwi	r7, 0
+	beq	3f
+
 	/* Load guest entry. */
-	lwzu	r7, 4(r9)
+	mulli	r11, r3, TLBE_BYTES
+	add	r11, r11, r9
+	lwz	r7, 0(r11)
 	mtspr	SPRN_MMUCR, r7
-	lwzu	r7, 4(r9)
-	tlbwe	r7, r6, PPC44x_TLB_PAGEID
-	lwzu	r7, 4(r9)
-	tlbwe	r7, r6, PPC44x_TLB_XLAT
-	lwzu	r7, 4(r9)
-	tlbwe	r7, r6, PPC44x_TLB_ATTRIB
-	/* Increment index. */
-	addi	r6, r6, 1
-	cmpw	r6, r8
-	blt	1b
+	lwz	r7, 4(r11)
+	tlbwe	r7, r3, PPC44x_TLB_PAGEID
+	lwz	r7, 8(r11)
+	tlbwe	r7, r3, PPC44x_TLB_XLAT
+	lwz	r7, 12(r11)
+	tlbwe	r7, r3, PPC44x_TLB_ATTRIB
+3:
+	addi	r3, r3, 1                       /* Increment index. */
+	bdnz	1b
+
 	mtspr	SPRN_MMUCR, r10			/* Restore host MMUCR. */
 
+	/* Clear bitmap of modified TLB entries */
+	li	r5, PPC44x_TLB_SIZE>>2
+	mtctr	r5
+	addi	r5, r4, VCPU_SHADOW_MOD - 4
+	li	r6, 0
+1:
+	stwu	r6, 4(r5)
+	bdnz	1b
+
 	iccci	0, 0 /* XXX hack */
 
 	/* Load some guest volatiles. */
@@ -431,6 +438,14 @@ lightweight_exit:
 	oris	r3, r3, KVMPPC_MSR_MASK@h
 	ori	r3, r3, KVMPPC_MSR_MASK@l
 	mtsrr1	r3
+
+	/* Clear any debug events which occurred since we disabled MSR[DE].
+	 * XXX This gives us a 3-instruction window in which a breakpoint
+	 * intended for guest context could fire in the host instead. */
+	lis	r3, 0xffff
+	ori	r3, r3, 0xffff
+	mtspr	SPRN_DBSR, r3
+
 	lwz	r3, VCPU_GPR(r3)(r4)
 	lwz	r4, VCPU_GPR(r4)(r4)
 	rfi
diff --git a/arch/powerpc/kvm/emulate.c b/arch/powerpc/kvm/emulate.c
index 8c605d0a548..0fce4fbdc20 100644
--- a/arch/powerpc/kvm/emulate.c
+++ b/arch/powerpc/kvm/emulate.c
@@ -170,6 +170,10 @@ static int kvmppc_emul_tlbwe(struct kvm_vcpu *vcpu, u32 inst)
 		kvmppc_mmu_map(vcpu, eaddr, raddr >> PAGE_SHIFT, asid, flags);
 	}
 
+	KVMTRACE_5D(GTLB_WRITE, vcpu, index,
+			tlbe->tid, tlbe->word0, tlbe->word1, tlbe->word2,
+			handler);
+
 	return EMULATE_DONE;
 }
 
@@ -504,7 +508,7 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			case SPRN_MMUCR:
 				vcpu->arch.mmucr = vcpu->arch.gpr[rs]; break;
 			case SPRN_PID:
-				vcpu->arch.pid = vcpu->arch.gpr[rs]; break;
+				kvmppc_set_pid(vcpu, vcpu->arch.gpr[rs]); break;
 			case SPRN_CCR0:
 				vcpu->arch.ccr0 = vcpu->arch.gpr[rs]; break;
 			case SPRN_CCR1:
@@ -765,6 +769,8 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 		break;
 	}
 
+	KVMTRACE_3D(PPC_INSTR, vcpu, inst, vcpu->arch.pc, emulated, entryexit);
+
 	if (advance)
 		vcpu->arch.pc += 4; /* Advance past emulated instruction. */
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 53826a5f6c0..90a6fc422b2 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -27,6 +27,7 @@
 #include <asm/cputable.h>
 #include <asm/uaccess.h>
 #include <asm/kvm_ppc.h>
+#include <asm/tlbflush.h>
 
 
 gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
@@ -239,18 +240,114 @@ void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
 {
 }
 
+/* Note: clearing MSR[DE] just means that the debug interrupt will not be
+ * delivered *immediately*. Instead, it simply sets the appropriate DBSR bits.
+ * If those DBSR bits are still set when MSR[DE] is re-enabled, the interrupt
+ * will be delivered as an "imprecise debug event" (which is indicated by
+ * DBSR[IDE].
+ */
+static void kvmppc_disable_debug_interrupts(void)
+{
+	mtmsr(mfmsr() & ~MSR_DE);
+}
+
+static void kvmppc_restore_host_debug_state(struct kvm_vcpu *vcpu)
+{
+	kvmppc_disable_debug_interrupts();
+
+	mtspr(SPRN_IAC1, vcpu->arch.host_iac[0]);
+	mtspr(SPRN_IAC2, vcpu->arch.host_iac[1]);
+	mtspr(SPRN_IAC3, vcpu->arch.host_iac[2]);
+	mtspr(SPRN_IAC4, vcpu->arch.host_iac[3]);
+	mtspr(SPRN_DBCR1, vcpu->arch.host_dbcr1);
+	mtspr(SPRN_DBCR2, vcpu->arch.host_dbcr2);
+	mtspr(SPRN_DBCR0, vcpu->arch.host_dbcr0);
+	mtmsr(vcpu->arch.host_msr);
+}
+
+static void kvmppc_load_guest_debug_registers(struct kvm_vcpu *vcpu)
+{
+	struct kvm_guest_debug *dbg = &vcpu->guest_debug;
+	u32 dbcr0 = 0;
+
+	vcpu->arch.host_msr = mfmsr();
+	kvmppc_disable_debug_interrupts();
+
+	/* Save host debug register state. */
+	vcpu->arch.host_iac[0] = mfspr(SPRN_IAC1);
+	vcpu->arch.host_iac[1] = mfspr(SPRN_IAC2);
+	vcpu->arch.host_iac[2] = mfspr(SPRN_IAC3);
+	vcpu->arch.host_iac[3] = mfspr(SPRN_IAC4);
+	vcpu->arch.host_dbcr0 = mfspr(SPRN_DBCR0);
+	vcpu->arch.host_dbcr1 = mfspr(SPRN_DBCR1);
+	vcpu->arch.host_dbcr2 = mfspr(SPRN_DBCR2);
+
+	/* set registers up for guest */
+
+	if (dbg->bp[0]) {
+		mtspr(SPRN_IAC1, dbg->bp[0]);
+		dbcr0 |= DBCR0_IAC1 | DBCR0_IDM;
+	}
+	if (dbg->bp[1]) {
+		mtspr(SPRN_IAC2, dbg->bp[1]);
+		dbcr0 |= DBCR0_IAC2 | DBCR0_IDM;
+	}
+	if (dbg->bp[2]) {
+		mtspr(SPRN_IAC3, dbg->bp[2]);
+		dbcr0 |= DBCR0_IAC3 | DBCR0_IDM;
+	}
+	if (dbg->bp[3]) {
+		mtspr(SPRN_IAC4, dbg->bp[3]);
+		dbcr0 |= DBCR0_IAC4 | DBCR0_IDM;
+	}
+
+	mtspr(SPRN_DBCR0, dbcr0);
+	mtspr(SPRN_DBCR1, 0);
+	mtspr(SPRN_DBCR2, 0);
+}
+
 void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 {
+	int i;
+
+	if (vcpu->guest_debug.enabled)
+		kvmppc_load_guest_debug_registers(vcpu);
+
+	/* Mark every guest entry in the shadow TLB entry modified, so that they
+	 * will all be reloaded on the next vcpu run (instead of being
+	 * demand-faulted). */
+	for (i = 0; i <= tlb_44x_hwater; i++)
+		kvmppc_tlbe_set_modified(vcpu, i);
 }
 
 void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
 {
+	if (vcpu->guest_debug.enabled)
+		kvmppc_restore_host_debug_state(vcpu);
+
+	/* Don't leave guest TLB entries resident when being de-scheduled. */
+	/* XXX It would be nice to differentiate between heavyweight exit and
+	 * sched_out here, since we could avoid the TLB flush for heavyweight
+	 * exits. */
+	_tlbia();
 }
 
 int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
                                     struct kvm_debug_guest *dbg)
 {
-	return -ENOTSUPP;
+	int i;
+
+	vcpu->guest_debug.enabled = dbg->enabled;
+	if (vcpu->guest_debug.enabled) {
+		for (i=0; i < ARRAY_SIZE(vcpu->guest_debug.bp); i++) {
+			if (dbg->breakpoints[i].enabled)
+				vcpu->guest_debug.bp[i] = dbg->breakpoints[i].address;
+			else
+				vcpu->guest_debug.bp[i] = 0;
+		}
+	}
+
+	return 0;
 }
 
 static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,