int vmsi_deliver(struct domain *d, int pirq)
{
struct hvm_irq_dpci *hvm_irq_dpci = d->arch.hvm_domain.irq.dpci;
uint32_t flags = hvm_irq_dpci->mirq[pirq].gmsi.gflags;
int vector = hvm_irq_dpci->mirq[pirq].gmsi.gvec;
uint8_t dest = (uint8_t)flags;
uint8_t dest_mode = !!(flags & VMSI_DM_MASK);
uint8_t delivery_mode = (flags & VMSI_DELIV_MASK) >> GLFAGS_SHIFT_DELIV_MODE;
uint8_t trig_mode = (flags & VMSI_TRIG_MODE) >> GLFAGS_SHIFT_TRG_MODE;
struct vlapic *target;
struct vcpu *v;
HVM_DBG_LOG(DBG_LEVEL_IOAPIC,
"msi: dest=%x dest_mode=%x delivery_mode=%x "
"vector=%x trig_mode=%x\n",
dest, dest_mode, delivery_mode, vector, trig_mode);
if ( !( hvm_irq_dpci->mirq[pirq].flags & HVM_IRQ_DPCI_GUEST_MSI ) )
{
gdprintk(XENLOG_WARNING, "pirq %x not msi \n", pirq);
return 0;
}
switch ( delivery_mode )
{
case dest_LowestPrio:
{
target = vlapic_lowest_prio(d, NULL, 0, dest, dest_mode);
if ( target != NULL )
vmsi_inj_irq(d, target, vector, trig_mode, delivery_mode);
else
HVM_DBG_LOG(DBG_LEVEL_IOAPIC, "null round robin: "
"vector=%x delivery_mode=%x\n",
vector, dest_LowestPrio);
break;
}
case dest_Fixed:
case dest_ExtINT:
{
for_each_vcpu ( d, v )
if ( vlapic_match_dest(vcpu_vlapic(v), NULL,
0, dest, dest_mode) )
vmsi_inj_irq(d, vcpu_vlapic(v),
vector, trig_mode, delivery_mode);
break;
}
case dest_SMI:
case dest_NMI:
case dest_INIT:
case dest__reserved_2:
default:
gdprintk(XENLOG_WARNING, "Unsupported delivery mode %d\n",
delivery_mode);
break;
}
return 1;
}
int rdmsr_viridian_regs(uint32_t idx, uint64_t *val)
{
struct vcpu *v = current;
struct domain *d = v->domain;
if ( !is_viridian_domain(d) )
return 0;
switch ( idx )
{
case VIRIDIAN_MSR_GUEST_OS_ID:
perfc_incr(mshv_rdmsr_osid);
*val = d->arch.hvm_domain.viridian.guest_os_id.raw;
break;
case VIRIDIAN_MSR_HYPERCALL:
perfc_incr(mshv_rdmsr_hc_page);
*val = d->arch.hvm_domain.viridian.hypercall_gpa.raw;
break;
case VIRIDIAN_MSR_VP_INDEX:
perfc_incr(mshv_rdmsr_vp_index);
*val = v->vcpu_id;
break;
case VIRIDIAN_MSR_TSC_FREQUENCY:
perfc_incr(mshv_rdmsr_tsc_frequency);
*val = (uint64_t)d->arch.tsc_khz * 1000ull;
break;
case VIRIDIAN_MSR_APIC_FREQUENCY:
perfc_incr(mshv_rdmsr_apic_frequency);
*val = 1000000000ull / APIC_BUS_CYCLE_NS;
break;
case VIRIDIAN_MSR_ICR:
perfc_incr(mshv_rdmsr_icr);
*val = (((uint64_t)vlapic_get_reg(vcpu_vlapic(v), APIC_ICR2) << 32) |
vlapic_get_reg(vcpu_vlapic(v), APIC_ICR));
break;
case VIRIDIAN_MSR_TPR:
perfc_incr(mshv_rdmsr_tpr);
*val = vlapic_get_reg(vcpu_vlapic(v), APIC_TASKPRI);
break;
case VIRIDIAN_MSR_APIC_ASSIST:
perfc_incr(mshv_rdmsr_apic_msr);
*val = v->arch.hvm_vcpu.viridian.apic_assist.raw;
break;
default:
return 0;
}
return 1;
}
int32_t cr_access_vmexit_handler(struct acrn_vcpu *vcpu)
{
uint64_t reg;
uint32_t idx;
uint64_t exit_qual;
int32_t ret = 0;
exit_qual = vcpu->arch.exit_qualification;
idx = (uint32_t)vm_exit_cr_access_reg_idx(exit_qual);
ASSERT((idx <= 15U), "index out of range");
reg = vcpu_get_gpreg(vcpu, idx);
switch ((vm_exit_cr_access_type(exit_qual) << 4U) | vm_exit_cr_access_cr_num(exit_qual)) {
case 0x00UL:
/* mov to cr0 */
vcpu_set_cr0(vcpu, reg);
break;
case 0x04UL:
/* mov to cr4 */
vcpu_set_cr4(vcpu, reg);
break;
case 0x08UL:
/* mov to cr8 */
/* According to SDM 6.15 "Exception and interrupt Reference":
*
* set reserved bit in CR8 causes GP to guest
*/
if ((reg & ~0xFUL) != 0UL) {
pr_dbg("Invalid cr8 write operation from guest");
vcpu_inject_gp(vcpu, 0U);
break;
}
vlapic_set_cr8(vcpu_vlapic(vcpu), reg);
break;
case 0x18UL:
/* mov from cr8 */
reg = vlapic_get_cr8(vcpu_vlapic(vcpu));
vcpu_set_gpreg(vcpu, idx, reg);
break;
default:
ASSERT(false, "Unhandled CR access");
ret = -EINVAL;
break;
}
TRACE_2L(TRACE_VMEXIT_CR_ACCESS, vm_exit_cr_access_type(exit_qual),
vm_exit_cr_access_cr_num(exit_qual));
return ret;
}
static uint32_t ioapic_get_delivery_bitmask(
struct hvm_hw_vioapic *vioapic, uint16_t dest, uint8_t dest_mode)
{
uint32_t mask = 0;
struct vcpu *v;
HVM_DBG_LOG(DBG_LEVEL_IOAPIC, "dest %d dest_mode %d",
dest, dest_mode);
if ( dest_mode == 0 ) /* Physical mode. */
{
if ( dest == 0xFF ) /* Broadcast. */
{
for_each_vcpu ( vioapic_domain(vioapic), v )
mask |= 1 << v->vcpu_id;
goto out;
}
for_each_vcpu ( vioapic_domain(vioapic), v )
{
if ( VLAPIC_ID(vcpu_vlapic(v)) == dest )
{
mask = 1 << v->vcpu_id;
break;
}
}
}
else if ( dest != 0 ) /* Logical mode, MDA non-zero. */
void vpmu_do_interrupt(struct cpu_user_regs *regs)
{
struct vcpu *v = current;
struct vpmu_struct *vpmu = vcpu_vpmu(v);
if ( vpmu->arch_vpmu_ops )
{
struct vlapic *vlapic = vcpu_vlapic(v);
u32 vlapic_lvtpc;
if ( !vpmu->arch_vpmu_ops->do_interrupt(regs) ||
!is_vlapic_lvtpc_enabled(vlapic) )
return;
vlapic_lvtpc = vlapic_get_reg(vlapic, APIC_LVTPC);
switch ( GET_APIC_DELIVERY_MODE(vlapic_lvtpc) )
{
case APIC_MODE_FIXED:
vlapic_set_irq(vlapic, vlapic_lvtpc & APIC_VECTOR_MASK, 0);
break;
case APIC_MODE_NMI:
v->nmi_pending = 1;
break;
}
}
}
int rdmsr_viridian_regs(uint32_t idx, uint64_t *val)
{
struct vcpu *v = current;
if ( !is_viridian_domain(v->domain) )
return 0;
switch ( idx )
{
case VIRIDIAN_MSR_GUEST_OS_ID:
perfc_incr(mshv_rdmsr_osid);
*val = v->domain->arch.hvm_domain.viridian.guest_os_id.raw;
break;
case VIRIDIAN_MSR_HYPERCALL:
perfc_incr(mshv_rdmsr_hc_page);
*val = v->domain->arch.hvm_domain.viridian.hypercall_gpa.raw;
break;
case VIRIDIAN_MSR_VP_INDEX:
perfc_incr(mshv_rdmsr_vp_index);
*val = v->vcpu_id;
break;
case VIRIDIAN_MSR_ICR:
perfc_incr(mshv_rdmsr_icr);
*val = (((uint64_t)vlapic_get_reg(vcpu_vlapic(v), APIC_ICR2) << 32) |
vlapic_get_reg(vcpu_vlapic(v), APIC_ICR));
break;
case VIRIDIAN_MSR_TPR:
perfc_incr(mshv_rdmsr_tpr);
*val = vlapic_get_reg(vcpu_vlapic(v), APIC_TASKPRI);
break;
default:
return 0;
}
return 1;
}
int vmsi_deliver(
struct domain *d, int vector,
uint8_t dest, uint8_t dest_mode,
uint8_t delivery_mode, uint8_t trig_mode)
{
struct vlapic *target;
struct vcpu *v;
switch ( delivery_mode )
{
case dest_LowestPrio:
target = vlapic_lowest_prio(d, NULL, 0, dest, dest_mode);
if ( target != NULL )
{
vmsi_inj_irq(target, vector, trig_mode, delivery_mode);
break;
}
HVM_DBG_LOG(DBG_LEVEL_VLAPIC, "null MSI round robin: vector=%02x\n",
vector);
return -ESRCH;
case dest_Fixed:
for_each_vcpu ( d, v )
if ( vlapic_match_dest(vcpu_vlapic(v), NULL,
0, dest, dest_mode) )
vmsi_inj_irq(vcpu_vlapic(v), vector,
trig_mode, delivery_mode);
break;
default:
printk(XENLOG_G_WARNING
"%pv: Unsupported MSI delivery mode %d for Dom%d\n",
current, delivery_mode, d->domain_id);
return -EINVAL;
}
return 0;
}
void vmx_intr_assist(void)
{
struct hvm_intack intack;
struct vcpu *v = current;
unsigned int tpr_threshold = 0;
enum hvm_intblk intblk;
/* Block event injection when single step with MTF. */
if ( unlikely(v->arch.hvm_vcpu.single_step) )
{
v->arch.hvm_vmx.exec_control |= CPU_BASED_MONITOR_TRAP_FLAG;
vmx_update_cpu_exec_control(v);
return;
}
/* Crank the handle on interrupt state. */
pt_update_irq(v);
do {
intack = hvm_vcpu_has_pending_irq(v);
if ( likely(intack.source == hvm_intsrc_none) )
goto out;
if ( unlikely(nvmx_intr_intercept(v, intack)) )
goto out;
intblk = hvm_interrupt_blocked(v, intack);
if ( intblk == hvm_intblk_tpr )
{
ASSERT(vlapic_enabled(vcpu_vlapic(v)));
ASSERT(intack.source == hvm_intsrc_lapic);
tpr_threshold = intack.vector >> 4;
goto out;
}
if ( (intblk != hvm_intblk_none) ||
(__vmread(VM_ENTRY_INTR_INFO) & INTR_INFO_VALID_MASK) )
{
enable_intr_window(v, intack);
goto out;
}
intack = hvm_vcpu_ack_pending_irq(v, intack);
} while ( intack.source == hvm_intsrc_none );
/* Return value, -1 : multi-dests, non-negative value: dest_vcpu_id */
int hvm_girq_dest_2_vcpu_id(struct domain *d, uint8_t dest, uint8_t dest_mode)
{
int dest_vcpu_id = -1, w = 0;
struct vcpu *v;
if ( d->max_vcpus == 1 )
return 0;
for_each_vcpu ( d, v )
{
if ( vlapic_match_dest(vcpu_vlapic(v), NULL, 0, dest, dest_mode) )
{
w++;
dest_vcpu_id = v->vcpu_id;
}
}
if ( w > 1 )
return -1;
return dest_vcpu_id;
}
static int construct_vmcs(struct vcpu *v)
{
uint16_t sysenter_cs;
unsigned long sysenter_eip;
vmx_vmcs_enter(v);
/* VMCS controls. */
__vmwrite(PIN_BASED_VM_EXEC_CONTROL, vmx_pin_based_exec_control);
__vmwrite(VM_EXIT_CONTROLS, vmx_vmexit_control);
__vmwrite(VM_ENTRY_CONTROLS, vmx_vmentry_control);
__vmwrite(CPU_BASED_VM_EXEC_CONTROL, vmx_cpu_based_exec_control);
v->arch.hvm_vmx.exec_control = vmx_cpu_based_exec_control;
if ( vmx_cpu_based_exec_control & CPU_BASED_ACTIVATE_SECONDARY_CONTROLS )
__vmwrite(SECONDARY_VM_EXEC_CONTROL, vmx_secondary_exec_control);
/* MSR access bitmap. */
if ( cpu_has_vmx_msr_bitmap )
{
char *msr_bitmap = alloc_xenheap_page();
if ( msr_bitmap == NULL )
return -ENOMEM;
memset(msr_bitmap, ~0, PAGE_SIZE);
v->arch.hvm_vmx.msr_bitmap = msr_bitmap;
__vmwrite(MSR_BITMAP, virt_to_maddr(msr_bitmap));
vmx_disable_intercept_for_msr(v, MSR_FS_BASE);
vmx_disable_intercept_for_msr(v, MSR_GS_BASE);
vmx_disable_intercept_for_msr(v, MSR_IA32_SYSENTER_CS);
vmx_disable_intercept_for_msr(v, MSR_IA32_SYSENTER_ESP);
vmx_disable_intercept_for_msr(v, MSR_IA32_SYSENTER_EIP);
}
/* I/O access bitmap. */
__vmwrite(IO_BITMAP_A, virt_to_maddr(hvm_io_bitmap));
__vmwrite(IO_BITMAP_B, virt_to_maddr(hvm_io_bitmap + PAGE_SIZE));
/* Host GDTR base. */
__vmwrite(HOST_GDTR_BASE, GDT_VIRT_START(v));
/* Host data selectors. */
__vmwrite(HOST_SS_SELECTOR, __HYPERVISOR_DS);
__vmwrite(HOST_DS_SELECTOR, __HYPERVISOR_DS);
__vmwrite(HOST_ES_SELECTOR, __HYPERVISOR_DS);
__vmwrite(HOST_FS_SELECTOR, 0);
__vmwrite(HOST_GS_SELECTOR, 0);
__vmwrite(HOST_FS_BASE, 0);
__vmwrite(HOST_GS_BASE, 0);
/* Host control registers. */
v->arch.hvm_vmx.host_cr0 = read_cr0() | X86_CR0_TS;
__vmwrite(HOST_CR0, v->arch.hvm_vmx.host_cr0);
__vmwrite(HOST_CR4, mmu_cr4_features);
/* Host CS:RIP. */
__vmwrite(HOST_CS_SELECTOR, __HYPERVISOR_CS);
__vmwrite(HOST_RIP, (unsigned long)vmx_asm_vmexit_handler);
/* Host SYSENTER CS:RIP. */
rdmsrl(MSR_IA32_SYSENTER_CS, sysenter_cs);
__vmwrite(HOST_SYSENTER_CS, sysenter_cs);
rdmsrl(MSR_IA32_SYSENTER_EIP, sysenter_eip);
__vmwrite(HOST_SYSENTER_EIP, sysenter_eip);
/* MSR intercepts. */
__vmwrite(VM_EXIT_MSR_LOAD_COUNT, 0);
__vmwrite(VM_EXIT_MSR_STORE_COUNT, 0);
__vmwrite(VM_ENTRY_MSR_LOAD_COUNT, 0);
__vmwrite(VM_ENTRY_INTR_INFO, 0);
__vmwrite(CR0_GUEST_HOST_MASK, ~0UL);
__vmwrite(CR4_GUEST_HOST_MASK, ~0UL);
__vmwrite(PAGE_FAULT_ERROR_CODE_MASK, 0);
__vmwrite(PAGE_FAULT_ERROR_CODE_MATCH, 0);
__vmwrite(CR3_TARGET_COUNT, 0);
__vmwrite(GUEST_ACTIVITY_STATE, 0);
/* Guest segment bases. */
__vmwrite(GUEST_ES_BASE, 0);
__vmwrite(GUEST_SS_BASE, 0);
__vmwrite(GUEST_DS_BASE, 0);
__vmwrite(GUEST_FS_BASE, 0);
__vmwrite(GUEST_GS_BASE, 0);
__vmwrite(GUEST_CS_BASE, 0);
/* Guest segment limits. */
__vmwrite(GUEST_ES_LIMIT, ~0u);
__vmwrite(GUEST_SS_LIMIT, ~0u);
__vmwrite(GUEST_DS_LIMIT, ~0u);
__vmwrite(GUEST_FS_LIMIT, ~0u);
__vmwrite(GUEST_GS_LIMIT, ~0u);
__vmwrite(GUEST_CS_LIMIT, ~0u);
/* Guest segment AR bytes. */
__vmwrite(GUEST_ES_AR_BYTES, 0xc093); /* read/write, accessed */
__vmwrite(GUEST_SS_AR_BYTES, 0xc093);
__vmwrite(GUEST_DS_AR_BYTES, 0xc093);
__vmwrite(GUEST_FS_AR_BYTES, 0xc093);
__vmwrite(GUEST_GS_AR_BYTES, 0xc093);
__vmwrite(GUEST_CS_AR_BYTES, 0xc09b); /* exec/read, accessed */
/* Guest IDT. */
__vmwrite(GUEST_IDTR_BASE, 0);
__vmwrite(GUEST_IDTR_LIMIT, 0);
/* Guest GDT. */
__vmwrite(GUEST_GDTR_BASE, 0);
__vmwrite(GUEST_GDTR_LIMIT, 0);
/* Guest LDT. */
__vmwrite(GUEST_LDTR_AR_BYTES, 0x0082); /* LDT */
__vmwrite(GUEST_LDTR_SELECTOR, 0);
__vmwrite(GUEST_LDTR_BASE, 0);
__vmwrite(GUEST_LDTR_LIMIT, 0);
/* Guest TSS. */
__vmwrite(GUEST_TR_AR_BYTES, 0x008b); /* 32-bit TSS (busy) */
__vmwrite(GUEST_TR_BASE, 0);
__vmwrite(GUEST_TR_LIMIT, 0xff);
__vmwrite(GUEST_INTERRUPTIBILITY_INFO, 0);
__vmwrite(GUEST_DR7, 0);
__vmwrite(VMCS_LINK_POINTER, ~0UL);
#if defined(__i386__)
__vmwrite(VMCS_LINK_POINTER_HIGH, ~0UL);
#endif
__vmwrite(EXCEPTION_BITMAP, (HVM_TRAP_MASK |
(1U << TRAP_page_fault) |
(1U << TRAP_no_device)));
v->arch.hvm_vcpu.guest_cr[0] = X86_CR0_PE | X86_CR0_ET;
hvm_update_guest_cr(v, 0);
v->arch.hvm_vcpu.guest_cr[4] = 0;
hvm_update_guest_cr(v, 4);
if ( cpu_has_vmx_tpr_shadow )
{
__vmwrite(VIRTUAL_APIC_PAGE_ADDR,
page_to_maddr(vcpu_vlapic(v)->regs_page));
__vmwrite(TPR_THRESHOLD, 0);
}
vmx_vmcs_exit(v);
paging_update_paging_modes(v); /* will update HOST & GUEST_CR3 as reqd */
vmx_vlapic_msr_changed(v);
return 0;
}
void vpmu_do_interrupt(struct cpu_user_regs *regs)
{
struct vcpu *sampled = current, *sampling;
struct vpmu_struct *vpmu;
struct vlapic *vlapic;
u32 vlapic_lvtpc;
/*
* dom0 will handle interrupt for special domains (e.g. idle domain) or,
* in XENPMU_MODE_ALL, for everyone.
*/
if ( (vpmu_mode & XENPMU_MODE_ALL) ||
(sampled->domain->domain_id >= DOMID_FIRST_RESERVED) )
{
sampling = choose_hwdom_vcpu();
if ( !sampling )
return;
}
else
sampling = sampled;
vpmu = vcpu_vpmu(sampling);
if ( !vpmu->arch_vpmu_ops )
return;
/* PV(H) guest */
if ( !is_hvm_vcpu(sampling) || (vpmu_mode & XENPMU_MODE_ALL) )
{
const struct cpu_user_regs *cur_regs;
uint64_t *flags = &vpmu->xenpmu_data->pmu.pmu_flags;
domid_t domid;
if ( !vpmu->xenpmu_data )
return;
if ( is_pvh_vcpu(sampling) &&
!(vpmu_mode & XENPMU_MODE_ALL) &&
!vpmu->arch_vpmu_ops->do_interrupt(regs) )
return;
if ( vpmu_is_set(vpmu, VPMU_CACHED) )
return;
/* PV guest will be reading PMU MSRs from xenpmu_data */
vpmu_set(vpmu, VPMU_CONTEXT_SAVE | VPMU_CONTEXT_LOADED);
vpmu->arch_vpmu_ops->arch_vpmu_save(sampling, 1);
vpmu_reset(vpmu, VPMU_CONTEXT_SAVE | VPMU_CONTEXT_LOADED);
if ( has_hvm_container_vcpu(sampled) )
*flags = 0;
else
*flags = PMU_SAMPLE_PV;
if ( sampled == sampling )
domid = DOMID_SELF;
else
domid = sampled->domain->domain_id;
/* Store appropriate registers in xenpmu_data */
/* FIXME: 32-bit PVH should go here as well */
if ( is_pv_32bit_vcpu(sampling) )
{
/*
* 32-bit dom0 cannot process Xen's addresses (which are 64 bit)
* and therefore we treat it the same way as a non-privileged
* PV 32-bit domain.
*/
struct compat_pmu_regs *cmp;
cur_regs = guest_cpu_user_regs();
cmp = (void *)&vpmu->xenpmu_data->pmu.r.regs;
cmp->ip = cur_regs->rip;
cmp->sp = cur_regs->rsp;
cmp->flags = cur_regs->eflags;
cmp->ss = cur_regs->ss;
cmp->cs = cur_regs->cs;
if ( (cmp->cs & 3) > 1 )
*flags |= PMU_SAMPLE_USER;
}
else
{
struct xen_pmu_regs *r = &vpmu->xenpmu_data->pmu.r.regs;
if ( (vpmu_mode & XENPMU_MODE_SELF) )
cur_regs = guest_cpu_user_regs();
else if ( !guest_mode(regs) &&
is_hardware_domain(sampling->domain) )
{
cur_regs = regs;
domid = DOMID_XEN;
}
else
cur_regs = guest_cpu_user_regs();
r->ip = cur_regs->rip;
r->sp = cur_regs->rsp;
r->flags = cur_regs->eflags;
if ( !has_hvm_container_vcpu(sampled) )
{
r->ss = cur_regs->ss;
r->cs = cur_regs->cs;
if ( !(sampled->arch.flags & TF_kernel_mode) )
*flags |= PMU_SAMPLE_USER;
}
else
{
struct segment_register seg;
hvm_get_segment_register(sampled, x86_seg_cs, &seg);
r->cs = seg.sel;
hvm_get_segment_register(sampled, x86_seg_ss, &seg);
r->ss = seg.sel;
r->cpl = seg.attr.fields.dpl;
if ( !(sampled->arch.hvm_vcpu.guest_cr[0] & X86_CR0_PE) )
*flags |= PMU_SAMPLE_REAL;
}
}
vpmu->xenpmu_data->domain_id = domid;
vpmu->xenpmu_data->vcpu_id = sampled->vcpu_id;
if ( is_hardware_domain(sampling->domain) )
vpmu->xenpmu_data->pcpu_id = smp_processor_id();
else
vpmu->xenpmu_data->pcpu_id = sampled->vcpu_id;
vpmu->hw_lapic_lvtpc |= APIC_LVT_MASKED;
apic_write(APIC_LVTPC, vpmu->hw_lapic_lvtpc);
*flags |= PMU_CACHED;
vpmu_set(vpmu, VPMU_CACHED);
send_guest_vcpu_virq(sampling, VIRQ_XENPMU);
return;
}
/* HVM guests */
vlapic = vcpu_vlapic(sampling);
/* We don't support (yet) HVM dom0 */
ASSERT(sampling == sampled);
if ( !vpmu->arch_vpmu_ops->do_interrupt(regs) ||
!is_vlapic_lvtpc_enabled(vlapic) )
return;
vlapic_lvtpc = vlapic_get_reg(vlapic, APIC_LVTPC);
switch ( GET_APIC_DELIVERY_MODE(vlapic_lvtpc) )
{
case APIC_MODE_FIXED:
vlapic_set_irq(vlapic, vlapic_lvtpc & APIC_VECTOR_MASK, 0);
break;
case APIC_MODE_NMI:
sampling->nmi_pending = 1;
break;
}
}
void vmx_intr_assist(void)
{
struct hvm_intack intack;
struct vcpu *v = current;
unsigned int tpr_threshold = 0;
enum hvm_intblk intblk;
int pt_vector = -1;
/* Block event injection when single step with MTF. */
if ( unlikely(v->arch.hvm_vcpu.single_step) )
{
v->arch.hvm_vmx.exec_control |= CPU_BASED_MONITOR_TRAP_FLAG;
vmx_update_cpu_exec_control(v);
return;
}
/* Crank the handle on interrupt state. */
if ( is_hvm_vcpu(v) )
pt_vector = pt_update_irq(v);
do {
unsigned long intr_info;
intack = hvm_vcpu_has_pending_irq(v);
if ( likely(intack.source == hvm_intsrc_none) )
goto out;
if ( unlikely(nvmx_intr_intercept(v, intack)) )
goto out;
intblk = hvm_interrupt_blocked(v, intack);
if ( cpu_has_vmx_virtual_intr_delivery )
{
/* Set "Interrupt-window exiting" for ExtINT and NMI. */
if ( (intblk != hvm_intblk_none) &&
(intack.source == hvm_intsrc_pic ||
intack.source == hvm_intsrc_vector ||
intack.source == hvm_intsrc_nmi) )
{
vmx_enable_intr_window(v, intack);
goto out;
}
__vmread(VM_ENTRY_INTR_INFO, &intr_info);
if ( intr_info & INTR_INFO_VALID_MASK )
{
if ( (intack.source == hvm_intsrc_pic) ||
(intack.source == hvm_intsrc_nmi) ||
(intack.source == hvm_intsrc_mce) )
vmx_enable_intr_window(v, intack);
goto out;
}
} else if ( intblk == hvm_intblk_tpr )
{
ASSERT(vlapic_enabled(vcpu_vlapic(v)));
ASSERT(intack.source == hvm_intsrc_lapic);
tpr_threshold = intack.vector >> 4;
goto out;
}
else if ( intblk != hvm_intblk_none )
{
vmx_enable_intr_window(v, intack);
goto out;
}
else
{
__vmread(VM_ENTRY_INTR_INFO, &intr_info);
if ( intr_info & INTR_INFO_VALID_MASK )
{
vmx_enable_intr_window(v, intack);
goto out;
}
}
intack = hvm_vcpu_ack_pending_irq(v, intack);
} while ( intack.source == hvm_intsrc_none );
int rdmsr_viridian_regs(uint32_t idx, uint64_t *val)
{
struct vcpu *v = current;
struct domain *d = v->domain;
if ( !is_viridian_domain(d) )
return 0;
switch ( idx )
{
case VIRIDIAN_MSR_GUEST_OS_ID:
perfc_incr(mshv_rdmsr_osid);
*val = d->arch.hvm_domain.viridian.guest_os_id.raw;
break;
case VIRIDIAN_MSR_HYPERCALL:
perfc_incr(mshv_rdmsr_hc_page);
*val = d->arch.hvm_domain.viridian.hypercall_gpa.raw;
break;
case VIRIDIAN_MSR_VP_INDEX:
perfc_incr(mshv_rdmsr_vp_index);
*val = v->vcpu_id;
break;
case VIRIDIAN_MSR_TSC_FREQUENCY:
if ( viridian_feature_mask(d) & HVMPV_no_freq )
return 0;
perfc_incr(mshv_rdmsr_tsc_frequency);
*val = (uint64_t)d->arch.tsc_khz * 1000ull;
break;
case VIRIDIAN_MSR_APIC_FREQUENCY:
if ( viridian_feature_mask(d) & HVMPV_no_freq )
return 0;
perfc_incr(mshv_rdmsr_apic_frequency);
*val = 1000000000ull / APIC_BUS_CYCLE_NS;
break;
case VIRIDIAN_MSR_ICR:
perfc_incr(mshv_rdmsr_icr);
*val = (((uint64_t)vlapic_get_reg(vcpu_vlapic(v), APIC_ICR2) << 32) |
vlapic_get_reg(vcpu_vlapic(v), APIC_ICR));
break;
case VIRIDIAN_MSR_TPR:
perfc_incr(mshv_rdmsr_tpr);
*val = vlapic_get_reg(vcpu_vlapic(v), APIC_TASKPRI);
break;
case VIRIDIAN_MSR_APIC_ASSIST:
perfc_incr(mshv_rdmsr_apic_msr);
*val = v->arch.hvm_vcpu.viridian.apic_assist.msr.raw;
break;
case VIRIDIAN_MSR_REFERENCE_TSC:
if ( !(viridian_feature_mask(d) & HVMPV_reference_tsc) )
return 0;
perfc_incr(mshv_rdmsr_tsc_msr);
*val = d->arch.hvm_domain.viridian.reference_tsc.raw;
break;
case VIRIDIAN_MSR_TIME_REF_COUNT:
{
struct viridian_time_ref_count *trc;
trc = &d->arch.hvm_domain.viridian.time_ref_count;
if ( !(viridian_feature_mask(d) & HVMPV_time_ref_count) )
return 0;
if ( !test_and_set_bit(_TRC_accessed, &trc->flags) )
printk(XENLOG_G_INFO "d%d: VIRIDIAN MSR_TIME_REF_COUNT: accessed\n",
d->domain_id);
perfc_incr(mshv_rdmsr_time_ref_count);
*val = raw_trc_val(d) + trc->off;
break;
}
default:
return 0;
}
return 1;
}
int wrmsr_viridian_regs(uint32_t idx, uint64_t val)
{
struct vcpu *v = current;
struct domain *d = v->domain;
if ( !is_viridian_domain(d) )
return 0;
switch ( idx )
{
case VIRIDIAN_MSR_GUEST_OS_ID:
perfc_incr(mshv_wrmsr_osid);
d->arch.hvm_domain.viridian.guest_os_id.raw = val;
dump_guest_os_id(d);
break;
case VIRIDIAN_MSR_HYPERCALL:
perfc_incr(mshv_wrmsr_hc_page);
d->arch.hvm_domain.viridian.hypercall_gpa.raw = val;
dump_hypercall(d);
if ( d->arch.hvm_domain.viridian.hypercall_gpa.fields.enabled )
enable_hypercall_page(d);
break;
case VIRIDIAN_MSR_VP_INDEX:
perfc_incr(mshv_wrmsr_vp_index);
break;
case VIRIDIAN_MSR_EOI:
perfc_incr(mshv_wrmsr_eoi);
vlapic_EOI_set(vcpu_vlapic(v));
break;
case VIRIDIAN_MSR_ICR: {
u32 eax = (u32)val, edx = (u32)(val >> 32);
struct vlapic *vlapic = vcpu_vlapic(v);
perfc_incr(mshv_wrmsr_icr);
eax &= ~(1 << 12);
edx &= 0xff000000;
vlapic_set_reg(vlapic, APIC_ICR2, edx);
vlapic_ipi(vlapic, eax, edx);
vlapic_set_reg(vlapic, APIC_ICR, eax);
break;
}
case VIRIDIAN_MSR_TPR:
perfc_incr(mshv_wrmsr_tpr);
vlapic_set_reg(vcpu_vlapic(v), APIC_TASKPRI, (uint8_t)val);
break;
case VIRIDIAN_MSR_APIC_ASSIST:
perfc_incr(mshv_wrmsr_apic_msr);
teardown_apic_assist(v); /* release any previous mapping */
v->arch.hvm_vcpu.viridian.apic_assist.msr.raw = val;
dump_apic_assist(v);
if ( v->arch.hvm_vcpu.viridian.apic_assist.msr.fields.enabled )
initialize_apic_assist(v);
break;
case VIRIDIAN_MSR_REFERENCE_TSC:
if ( !(viridian_feature_mask(d) & HVMPV_reference_tsc) )
return 0;
perfc_incr(mshv_wrmsr_tsc_msr);
d->arch.hvm_domain.viridian.reference_tsc.raw = val;
dump_reference_tsc(d);
if ( d->arch.hvm_domain.viridian.reference_tsc.fields.enabled )
update_reference_tsc(d, 1);
break;
default:
return 0;
}
return 1;
}
int wrmsr_viridian_regs(uint32_t idx, uint64_t val)
{
struct domain *d = current->domain;
if ( !is_viridian_domain(d) )
return 0;
switch ( idx )
{
case VIRIDIAN_MSR_GUEST_OS_ID:
perfc_incr(mshv_wrmsr_osid);
d->arch.hvm_domain.viridian.guest_os_id.raw = val;
gdprintk(XENLOG_INFO, "Guest os:\n");
gdprintk(XENLOG_INFO, "\tvendor: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.vendor);
gdprintk(XENLOG_INFO, "\tos: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.os);
gdprintk(XENLOG_INFO, "\tmajor: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.major);
gdprintk(XENLOG_INFO, "\tminor: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.minor);
gdprintk(XENLOG_INFO, "\tsp: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.service_pack);
gdprintk(XENLOG_INFO, "\tbuild: %x\n",
d->arch.hvm_domain.viridian.guest_os_id.fields.build_number);
break;
case VIRIDIAN_MSR_HYPERCALL:
perfc_incr(mshv_wrmsr_hc_page);
gdprintk(XENLOG_INFO, "Set hypercall page %"PRIx64".\n", val);
if ( d->arch.hvm_domain.viridian.guest_os_id.raw == 0 )
break;
d->arch.hvm_domain.viridian.hypercall_gpa.raw = val;
if ( d->arch.hvm_domain.viridian.hypercall_gpa.fields.enabled )
enable_hypercall_page();
break;
case VIRIDIAN_MSR_VP_INDEX:
perfc_incr(mshv_wrmsr_vp_index);
gdprintk(XENLOG_INFO, "Set VP index %"PRIu64".\n", val);
break;
case VIRIDIAN_MSR_EOI:
perfc_incr(mshv_wrmsr_eoi);
vlapic_EOI_set(vcpu_vlapic(current));
break;
case VIRIDIAN_MSR_ICR: {
u32 eax = (u32)val, edx = (u32)(val >> 32);
struct vlapic *vlapic = vcpu_vlapic(current);
perfc_incr(mshv_wrmsr_icr);
eax &= ~(1 << 12);
edx &= 0xff000000;
vlapic_set_reg(vlapic, APIC_ICR2, edx);
if ( vlapic_ipi(vlapic, eax, edx) == X86EMUL_OKAY )
vlapic_set_reg(vlapic, APIC_ICR, eax);
break;
}
case VIRIDIAN_MSR_TPR:
perfc_incr(mshv_wrmsr_tpr);
vlapic_set_reg(vcpu_vlapic(current), APIC_TASKPRI, (uint8_t)val);
break;
case VIRIDIAN_MSR_APIC_ASSIST:
/*
* We don't support the APIC assist page, and that fact is reflected in
* our CPUID flags. However, Windows 7 build 7000 has a bug which means
* that it doesn't recognise that, and tries to use the page anyway. We
* therefore have to fake up just enough to keep win7 happy.
* Fortunately, that's really easy: just setting the first four bytes
* in the page to zero effectively disables the page again, so that's
* what we do. Semantically, the first four bytes are supposed to be a
* flag saying whether the guest really needs to issue an EOI. Setting
* that flag to zero means that it must always issue one, which is what
* we want. Once a page has been repurposed as an APIC assist page the
* guest isn't allowed to set anything in it, so the flag remains zero
* and all is fine. The guest is allowed to clear flags in the page,
* but that doesn't cause us any problems.
*/
if ( val & 1 ) /* APIC assist page enabled? */
{
uint32_t word = 0;
paddr_t page_start = val & ~1ul;
(void)hvm_copy_to_guest_phys(page_start, &word, sizeof(word));
}
break;
default:
return 0;
}
return 1;
}