/*
* Common checks before entering the guest world. Call with interrupts
* disabled.
*
* returns:
*
* == 1 if we're ready to go into guest state
* <= 0 if we need to go back to the host with return value
*/
int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
{
int r;
WARN_ON(irqs_disabled());
hard_irq_disable();
while (true) {
if (need_resched()) {
local_irq_enable();
cond_resched();
hard_irq_disable();
continue;
}
if (signal_pending(current)) {
kvmppc_account_exit(vcpu, SIGNAL_EXITS);
vcpu->run->exit_reason = KVM_EXIT_INTR;
r = -EINTR;
break;
}
vcpu->mode = IN_GUEST_MODE;
/*
* Reading vcpu->requests must happen after setting vcpu->mode,
* so we don't miss a request because the requester sees
* OUTSIDE_GUEST_MODE and assumes we'll be checking requests
* before next entering the guest (and thus doesn't IPI).
* This also orders the write to mode from any reads
* to the page tables done while the VCPU is running.
* Please see the comment in kvm_flush_remote_tlbs.
*/
smp_mb();
if (vcpu->requests) {
/* Make sure we process requests preemptable */
local_irq_enable();
trace_kvm_check_requests(vcpu);
r = kvmppc_core_check_requests(vcpu);
hard_irq_disable();
if (r > 0)
continue;
break;
}
if (kvmppc_core_prepare_to_enter(vcpu)) {
/* interrupts got enabled in between, so we
are back at square 1 */
continue;
}
guest_enter_irqoff();
return 1;
}
/* return to host */
local_irq_enable();
return r;
}
static int emulate_mtdcr(struct kvm_vcpu *vcpu, int rs, int dcrn)
{
/* emulate some access in kernel */
switch (dcrn) {
case DCRN_CPR0_CONFIG_ADDR:
vcpu->arch.cpr0_cfgaddr = kvmppc_get_gpr(vcpu, rs);
return EMULATE_DONE;
default:
vcpu->run->dcr.dcrn = dcrn;
vcpu->run->dcr.data = kvmppc_get_gpr(vcpu, rs);
vcpu->run->dcr.is_write = 1;
vcpu->arch.dcr_is_write = 1;
vcpu->arch.dcr_needed = 1;
kvmppc_account_exit(vcpu, DCR_EXITS);
return EMULATE_DO_DCR;
}
}
static int emulate_mfdcr(struct kvm_vcpu *vcpu, int rt, int dcrn)
{
/* The guest may access CPR0 registers to determine the timebase
* frequency, and it must know the real host frequency because it
* can directly access the timebase registers.
*
* It would be possible to emulate those accesses in userspace,
* but userspace can really only figure out the end frequency.
* We could decompose that into the factors that compute it, but
* that's tricky math, and it's easier to just report the real
* CPR0 values.
*/
switch (dcrn) {
case DCRN_CPR0_CONFIG_ADDR:
kvmppc_set_gpr(vcpu, rt, vcpu->arch.cpr0_cfgaddr);
break;
case DCRN_CPR0_CONFIG_DATA:
local_irq_disable();
mtdcr(DCRN_CPR0_CONFIG_ADDR,
vcpu->arch.cpr0_cfgaddr);
kvmppc_set_gpr(vcpu, rt,
mfdcr(DCRN_CPR0_CONFIG_DATA));
local_irq_enable();
break;
default:
vcpu->run->dcr.dcrn = dcrn;
vcpu->run->dcr.data = 0;
vcpu->run->dcr.is_write = 0;
vcpu->arch.dcr_is_write = 0;
vcpu->arch.io_gpr = rt;
vcpu->arch.dcr_needed = 1;
kvmppc_account_exit(vcpu, DCR_EXITS);
return EMULATE_DO_DCR;
}
return EMULATE_DONE;
}
/*
* Common checks before entering the guest world. Call with interrupts
* disabled.
*
* returns:
*
* == 1 if we're ready to go into guest state
* <= 0 if we need to go back to the host with return value
*/
int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
{
int r = 1;
WARN_ON_ONCE(!irqs_disabled());
while (true) {
if (need_resched()) {
local_irq_enable();
cond_resched();
local_irq_disable();
continue;
}
if (signal_pending(current)) {
kvmppc_account_exit(vcpu, SIGNAL_EXITS);
vcpu->run->exit_reason = KVM_EXIT_INTR;
r = -EINTR;
break;
}
vcpu->mode = IN_GUEST_MODE;
/*
* Reading vcpu->requests must happen after setting vcpu->mode,
* so we don't miss a request because the requester sees
* OUTSIDE_GUEST_MODE and assumes we'll be checking requests
* before next entering the guest (and thus doesn't IPI).
*/
smp_mb();
if (vcpu->requests) {
/* Make sure we process requests preemptable */
local_irq_enable();
trace_kvm_check_requests(vcpu);
r = kvmppc_core_check_requests(vcpu);
local_irq_disable();
if (r > 0)
continue;
break;
}
if (kvmppc_core_prepare_to_enter(vcpu)) {
/* interrupts got enabled in between, so we
are back at square 1 */
continue;
}
#ifdef CONFIG_PPC64
/* lazy EE magic */
hard_irq_disable();
if (lazy_irq_pending()) {
/* Got an interrupt in between, try again */
local_irq_enable();
local_irq_disable();
kvm_guest_exit();
continue;
}
#endif
kvm_guest_enter();
break;
}
return r;
}
int kvmppc_core_emulate_op(struct kvm_run *run, struct kvm_vcpu *vcpu,
unsigned int inst, int *advance)
{
int emulated = EMULATE_DONE;
int dcrn;
int ra;
int rb;
int rc;
int rs;
int rt;
int ws;
switch (get_op(inst)) {
case 31:
switch (get_xop(inst)) {
case XOP_MFDCR:
dcrn = get_dcrn(inst);
rt = get_rt(inst);
/* The guest may access CPR0 registers to determine the timebase
* frequency, and it must know the real host frequency because it
* can directly access the timebase registers.
*
* It would be possible to emulate those accesses in userspace,
* but userspace can really only figure out the end frequency.
* We could decompose that into the factors that compute it, but
* that's tricky math, and it's easier to just report the real
* CPR0 values.
*/
switch (dcrn) {
case DCRN_CPR0_CONFIG_ADDR:
kvmppc_set_gpr(vcpu, rt, vcpu->arch.cpr0_cfgaddr);
break;
case DCRN_CPR0_CONFIG_DATA:
local_irq_disable();
mtdcr(DCRN_CPR0_CONFIG_ADDR,
vcpu->arch.cpr0_cfgaddr);
kvmppc_set_gpr(vcpu, rt,
mfdcr(DCRN_CPR0_CONFIG_DATA));
local_irq_enable();
break;
default:
run->dcr.dcrn = dcrn;
run->dcr.data = 0;
run->dcr.is_write = 0;
vcpu->arch.io_gpr = rt;
vcpu->arch.dcr_needed = 1;
kvmppc_account_exit(vcpu, DCR_EXITS);
emulated = EMULATE_DO_DCR;
}
break;
case XOP_MTDCR:
dcrn = get_dcrn(inst);
rs = get_rs(inst);
/* emulate some access in kernel */
switch (dcrn) {
case DCRN_CPR0_CONFIG_ADDR:
vcpu->arch.cpr0_cfgaddr = kvmppc_get_gpr(vcpu, rs);
break;
default:
run->dcr.dcrn = dcrn;
run->dcr.data = kvmppc_get_gpr(vcpu, rs);
run->dcr.is_write = 1;
vcpu->arch.dcr_needed = 1;
kvmppc_account_exit(vcpu, DCR_EXITS);
emulated = EMULATE_DO_DCR;
}
break;
case XOP_TLBWE:
ra = get_ra(inst);
rs = get_rs(inst);
ws = get_ws(inst);
emulated = kvmppc_44x_emul_tlbwe(vcpu, ra, rs, ws);
break;
case XOP_TLBSX:
rt = get_rt(inst);
ra = get_ra(inst);
rb = get_rb(inst);
rc = get_rc(inst);
emulated = kvmppc_44x_emul_tlbsx(vcpu, rt, ra, rb, rc);
break;
case XOP_ICCCI:
break;
default:
emulated = EMULATE_FAIL;
}
break;
default:
emulated = EMULATE_FAIL;
}
if (emulated == EMULATE_FAIL)
emulated = kvmppc_booke_emulate_op(run, vcpu, inst, advance);
return emulated;
}
int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
unsigned int exit_nr)
{
enum emulation_result er;
int r = RESUME_HOST;
/* update before a new last_exit_type is rewritten */
kvmppc_update_timing_stats(vcpu);
local_irq_enable();
run->exit_reason = KVM_EXIT_UNKNOWN;
run->ready_for_interrupt_injection = 1;
switch (exit_nr) {
case BOOKE_INTERRUPT_MACHINE_CHECK:
printk("MACHINE CHECK: %lx\n", mfspr(SPRN_MCSR));
kvmppc_dump_vcpu(vcpu);
r = RESUME_HOST;
break;
case BOOKE_INTERRUPT_EXTERNAL:
kvmppc_account_exit(vcpu, EXT_INTR_EXITS);
if (need_resched())
cond_resched();
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_DECREMENTER:
/* Since we switched IVPR back to the host's value, the host
* handled this interrupt the moment we enabled interrupts.
* Now we just offer it a chance to reschedule the guest. */
kvmppc_account_exit(vcpu, DEC_EXITS);
if (need_resched())
cond_resched();
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_PROGRAM:
if (vcpu->arch.msr & MSR_PR) {
/* Program traps generated by user-level software must be handled
* by the guest kernel. */
kvmppc_core_queue_program(vcpu, vcpu->arch.fault_esr);
r = RESUME_GUEST;
kvmppc_account_exit(vcpu, USR_PR_INST);
break;
}
er = kvmppc_emulate_instruction(run, vcpu);
switch (er) {
case EMULATE_DONE:
/* don't overwrite subtypes, just account kvm_stats */
kvmppc_account_exit_stat(vcpu, EMULATED_INST_EXITS);
/* Future optimization: only reload non-volatiles if
* they were actually modified by emulation. */
r = RESUME_GUEST_NV;
break;
case EMULATE_DO_DCR:
run->exit_reason = KVM_EXIT_DCR;
r = RESUME_HOST;
break;
case EMULATE_FAIL:
/* XXX Deliver Program interrupt to guest. */
printk(KERN_CRIT "%s: emulation at %lx failed (%08x)\n",
__func__, vcpu->arch.pc, vcpu->arch.last_inst);
/* For debugging, encode the failing instruction and
* report it to userspace. */
run->hw.hardware_exit_reason = ~0ULL << 32;
run->hw.hardware_exit_reason |= vcpu->arch.last_inst;
r = RESUME_HOST;
break;
default:
BUG();
}
break;
case BOOKE_INTERRUPT_FP_UNAVAIL:
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_FP_UNAVAIL);
kvmppc_account_exit(vcpu, FP_UNAVAIL);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_SPE_UNAVAIL:
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SPE_UNAVAIL);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_SPE_FP_DATA:
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SPE_FP_DATA);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_SPE_FP_ROUND:
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SPE_FP_ROUND);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_DATA_STORAGE:
kvmppc_core_queue_data_storage(vcpu, vcpu->arch.fault_dear,
vcpu->arch.fault_esr);
kvmppc_account_exit(vcpu, DSI_EXITS);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_INST_STORAGE:
kvmppc_core_queue_inst_storage(vcpu, vcpu->arch.fault_esr);
kvmppc_account_exit(vcpu, ISI_EXITS);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_SYSCALL:
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SYSCALL);
kvmppc_account_exit(vcpu, SYSCALL_EXITS);
r = RESUME_GUEST;
break;
case BOOKE_INTERRUPT_DTLB_MISS: {
unsigned long eaddr = vcpu->arch.fault_dear;
int gtlb_index;
gpa_t gpaddr;
gfn_t gfn;
/* Check the guest TLB. */
gtlb_index = kvmppc_mmu_dtlb_index(vcpu, eaddr);
if (gtlb_index < 0) {
/* The guest didn't have a mapping for it. */
kvmppc_core_queue_dtlb_miss(vcpu,
vcpu->arch.fault_dear,
vcpu->arch.fault_esr);
kvmppc_mmu_dtlb_miss(vcpu);
kvmppc_account_exit(vcpu, DTLB_REAL_MISS_EXITS);
r = RESUME_GUEST;
break;
}
gpaddr = kvmppc_mmu_xlate(vcpu, gtlb_index, eaddr);
gfn = gpaddr >> PAGE_SHIFT;
if (kvm_is_visible_gfn(vcpu->kvm, gfn)) {
/* The guest TLB had a mapping, but the shadow TLB
* didn't, and it is RAM. This could be because:
* a) the entry is mapping the host kernel, or
* b) the guest used a large mapping which we're faking
* Either way, we need to satisfy the fault without
* invoking the guest. */
kvmppc_mmu_map(vcpu, eaddr, gpaddr, gtlb_index);
kvmppc_account_exit(vcpu, DTLB_VIRT_MISS_EXITS);
r = RESUME_GUEST;
} else {
/* Guest has mapped and accessed a page which is not
* actually RAM. */
vcpu->arch.paddr_accessed = gpaddr;
r = kvmppc_emulate_mmio(run, vcpu);
kvmppc_account_exit(vcpu, MMIO_EXITS);
}
break;
}
case BOOKE_INTERRUPT_ITLB_MISS: {
unsigned long eaddr = vcpu->arch.pc;
gpa_t gpaddr;
gfn_t gfn;
int gtlb_index;
r = RESUME_GUEST;
/* Check the guest TLB. */
gtlb_index = kvmppc_mmu_itlb_index(vcpu, eaddr);
if (gtlb_index < 0) {
/* The guest didn't have a mapping for it. */
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_ITLB_MISS);
kvmppc_mmu_itlb_miss(vcpu);
kvmppc_account_exit(vcpu, ITLB_REAL_MISS_EXITS);
break;
}
kvmppc_account_exit(vcpu, ITLB_VIRT_MISS_EXITS);
gpaddr = kvmppc_mmu_xlate(vcpu, gtlb_index, eaddr);
gfn = gpaddr >> PAGE_SHIFT;
if (kvm_is_visible_gfn(vcpu->kvm, gfn)) {
/* The guest TLB had a mapping, but the shadow TLB
* didn't. This could be because:
* a) the entry is mapping the host kernel, or
* b) the guest used a large mapping which we're faking
* Either way, we need to satisfy the fault without
* invoking the guest. */
kvmppc_mmu_map(vcpu, eaddr, gpaddr, gtlb_index);
} else {
/* Guest mapped and leaped at non-RAM! */
kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_MACHINE_CHECK);
}
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;
kvmppc_account_exit(vcpu, DEBUG_EXITS);
r = RESUME_HOST;
break;
}
default:
printk(KERN_EMERG "exit_nr %d\n", exit_nr);
BUG();
}
local_irq_disable();
kvmppc_core_deliver_interrupts(vcpu);
if (!(r & RESUME_HOST)) {
/* To avoid clobbering exit_reason, only check for signals if
* we aren't already exiting to userspace for some other
* reason. */
if (signal_pending(current)) {
run->exit_reason = KVM_EXIT_INTR;
r = (-EINTR << 2) | RESUME_HOST | (r & RESUME_FLAG_NV);
kvmppc_account_exit(vcpu, SIGNAL_EXITS);
}
}
return r;
}
