system/linux
system/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux synced 2024-10-04 18:33:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

KVM x86 misc changes for 6.9:

Browse source 
- Explicitly initialize a variety of on-stack variables in the emulator that
    triggered KMSAN false positives (though in fairness in KMSAN, it's comically
    difficult to see that the uninitialized memory is never truly consumed).
 
  - Fix the deubgregs ABI for 32-bit KVM, and clean up code related to reading
    DR6 and DR7.
 
  - Rework the "force immediate exit" code so that vendor code ultimately
    decides how and when to force the exit.  This allows VMX to further optimize
    handling preemption timer exits, and allows SVM to avoid sending a duplicate
    IPI (SVM also has a need to force an exit).
 
  - Fix a long-standing bug where kvm_has_noapic_vcpu could be left elevated if
    vCPU creation ultimately failed, and add WARN to guard against similar bugs.
 
  - Provide a dedicated arch hook for checking if a different vCPU was in-kernel
    (for directed yield), and simplify the logic for checking if the currently
    loaded vCPU is in-kernel.
 
  - Misc cleanups and fixes.
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEEKTobbabEP7vbhhN9OlYIJqCjN/0FAmXrRjQACgkQOlYIJqCj
 N/2Dzw//b+ptSBAl1kGBRmk/DqsX7J9ZkQYCQOTeh1vXiUM+XRTSQoArN0Oo1roy
 3wcEnQ0beVw7jMuzZ8UUuTfU8WUMja/kwltnqXYNHwLnb6yH0I/BIengXWdUdAMc
 FmgPZ4qJR2IzKYzvDsc3eEQ515O8UHWakyVDnmLBtiakAeBcUTYceHpEEPpzE5y5
 ODASTQKM9o/h8R8JwKFTJ8/mrOLNcsu5SycwFdnmubLJCrNWtJWTijA6y1lh6shn
 hbEJex+ESoC2v8p7IP53u1SGJubVlPajt+RkYJtlEI3WVsevp024eYcF4nb1OjXi
 qS2Y3W7DQGWvyCBoSzoMY+9nRMgyOOpHYetdiz+9oZOmnjiYWY0ku59U7Gv+Aotj
 AUbCn4Ry/OpqsuZ7Oo7i3IT8R7uzsTeNNdxhYBn1OQquBEZ0KBYXlZkGfTk9K0t0
 Fhka/5Zu6fBlg5J+zCyaXUGmsGWBo/9HxsC5z1JuKo8fatro5qyqYE5KiM01dkqc
 6FET6gL+fFprC5c67JGRPdEtk6F9Emb+6oiTTA8/8q8JQQAKiJKk95Nlq7KzPfVS
 A5RQPTuTJ7acE/5CY4zB1DdxCjqgnonBEA2ULnA/J10Rk8orHJRnGJcEwKEyDrZh
 HpsxIIqt++i8KffORpCym6zSAVYuQjn1mu7MGth+zuCqhcEpBfc=
 =GX0O
 -----END PGP SIGNATURE-----

Merge tag 'kvm-x86-misc-6.9' of https://github.com/kvm-x86/linux into HEAD

KVM x86 misc changes for 6.9:

 - Explicitly initialize a variety of on-stack variables in the emulator that
   triggered KMSAN false positives (though in fairness in KMSAN, it's comically
   difficult to see that the uninitialized memory is never truly consumed).

 - Fix the deubgregs ABI for 32-bit KVM, and clean up code related to reading
   DR6 and DR7.

 - Rework the "force immediate exit" code so that vendor code ultimately
   decides how and when to force the exit.  This allows VMX to further optimize
   handling preemption timer exits, and allows SVM to avoid sending a duplicate
   IPI (SVM also has a need to force an exit).

 - Fix a long-standing bug where kvm_has_noapic_vcpu could be left elevated if
   vCPU creation ultimately failed, and add WARN to guard against similar bugs.

 - Provide a dedicated arch hook for checking if a different vCPU was in-kernel
   (for directed yield), and simplify the logic for checking if the currently
   loaded vCPU is in-kernel.

 - Misc cleanups and fixes.
This commit is contained in:
Paolo Bonzini 2024-03-11 10:24:56 -04:00
parent 507e72f899 78ccfce774
commit c9cd0beae9
15 changed files with 184 additions and 177 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
arch/x86/include/asm/kvm-x86-ops.h
View file

@ -103,7 +103,6 @@ KVM_X86_OP(write_tsc_multiplier)
KVM_X86_OP(get_exit_info)
KVM_X86_OP(check_intercept)
KVM_X86_OP(handle_exit_irqoff)
KVM_X86_OP(request_immediate_exit)
KVM_X86_OP(sched_in)
KVM_X86_OP_OPTIONAL(update_cpu_dirty_logging)
KVM_X86_OP_OPTIONAL(vcpu_blocking)

8
arch/x86/include/asm/kvm_host.h
View file

@ -1665,7 +1665,8 @@ struct kvm_x86_ops {
void (*flush_tlb_guest)(struct kvm_vcpu *vcpu);
int (*vcpu_pre_run)(struct kvm_vcpu *vcpu);
enum exit_fastpath_completion (*vcpu_run)(struct kvm_vcpu *vcpu);
enum exit_fastpath_completion (*vcpu_run)(struct kvm_vcpu *vcpu,
bool force_immediate_exit);
int (*handle_exit)(struct kvm_vcpu *vcpu,
enum exit_fastpath_completion exit_fastpath);
int (*skip_emulated_instruction)(struct kvm_vcpu *vcpu);
@ -1733,8 +1734,6 @@ struct kvm_x86_ops {
struct x86_exception *exception);
void (*handle_exit_irqoff)(struct kvm_vcpu *vcpu);
void (*request_immediate_exit)(struct kvm_vcpu *vcpu);
void (*sched_in)(struct kvm_vcpu *vcpu, int cpu);
/*
@ -2047,7 +2046,7 @@ int kvm_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3);
int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4);
int kvm_set_cr8(struct kvm_vcpu *vcpu, unsigned long cr8);
int kvm_set_dr(struct kvm_vcpu *vcpu, int dr, unsigned long val);
void kvm_get_dr(struct kvm_vcpu *vcpu, int dr, unsigned long *val);
unsigned long kvm_get_dr(struct kvm_vcpu *vcpu, int dr);
unsigned long kvm_get_cr8(struct kvm_vcpu *vcpu);
void kvm_lmsw(struct kvm_vcpu *vcpu, unsigned long msw);
int kvm_emulate_xsetbv(struct kvm_vcpu *vcpu);
@ -2240,7 +2239,6 @@ extern bool kvm_find_async_pf_gfn(struct kvm_vcpu *vcpu, gfn_t gfn);
int kvm_skip_emulated_instruction(struct kvm_vcpu *vcpu);
int kvm_complete_insn_gp(struct kvm_vcpu *vcpu, int err);
void __kvm_request_immediate_exit(struct kvm_vcpu *vcpu);
void __user *__x86_set_memory_region(struct kvm *kvm, int id, gpa_t gpa,
u32 size);

8
arch/x86/include/asm/svm.h
View file

@ -358,10 +358,10 @@ struct sev_es_save_area {
struct vmcb_seg ldtr;
struct vmcb_seg idtr;
struct vmcb_seg tr;
u64 vmpl0_ssp;
u64 vmpl1_ssp;
u64 vmpl2_ssp;
u64 vmpl3_ssp;
u64 pl0_ssp;
u64 pl1_ssp;
u64 pl2_ssp;
u64 pl3_ssp;
u64 u_cet;
u8 reserved_0xc8[2];
u8 vmpl;

45
arch/x86/kvm/emulate.c
View file

@ -1820,22 +1820,22 @@ static int writeback(struct x86_emulate_ctxt *ctxt, struct operand *op)
return X86EMUL_CONTINUE;
}
static int push(struct x86_emulate_ctxt *ctxt, void *data, int bytes)
static int emulate_push(struct x86_emulate_ctxt *ctxt, const void *data, int len)
{
struct segmented_address addr;
rsp_increment(ctxt, -bytes);
rsp_increment(ctxt, -len);
addr.ea = reg_read(ctxt, VCPU_REGS_RSP) & stack_mask(ctxt);
addr.seg = VCPU_SREG_SS;
return segmented_write(ctxt, addr, data, bytes);
return segmented_write(ctxt, addr, data, len);
}
static int em_push(struct x86_emulate_ctxt *ctxt)
{
/* Disable writeback. */
ctxt->dst.type = OP_NONE;
return push(ctxt, &ctxt->src.val, ctxt->op_bytes);
return emulate_push(ctxt, &ctxt->src.val, ctxt->op_bytes);
}
static int emulate_pop(struct x86_emulate_ctxt *ctxt,
@ -1863,7 +1863,8 @@ static int emulate_popf(struct x86_emulate_ctxt *ctxt,
void *dest, int len)
{
int rc;
unsigned long val, change_mask;
unsigned long val = 0;
unsigned long change_mask;
int iopl = (ctxt->eflags & X86_EFLAGS_IOPL) >> X86_EFLAGS_IOPL_BIT;
int cpl = ctxt->ops->cpl(ctxt);
@ -1920,7 +1921,7 @@ static int em_enter(struct x86_emulate_ctxt *ctxt)
return X86EMUL_UNHANDLEABLE;
rbp = reg_read(ctxt, VCPU_REGS_RBP);
rc = push(ctxt, &rbp, stack_size(ctxt));
rc = emulate_push(ctxt, &rbp, stack_size(ctxt));
if (rc != X86EMUL_CONTINUE)
return rc;
assign_masked(reg_rmw(ctxt, VCPU_REGS_RBP), reg_read(ctxt, VCPU_REGS_RSP),
@ -1954,7 +1955,7 @@ static int em_push_sreg(struct x86_emulate_ctxt *ctxt)
static int em_pop_sreg(struct x86_emulate_ctxt *ctxt)
{
int seg = ctxt->src2.val;
unsigned long selector;
unsigned long selector = 0;
int rc;
rc = emulate_pop(ctxt, &selector, 2);
@ -2000,7 +2001,7 @@ static int em_popa(struct x86_emulate_ctxt *ctxt)
{
int rc = X86EMUL_CONTINUE;
int reg = VCPU_REGS_RDI;
u32 val;
u32 val = 0;
while (reg >= VCPU_REGS_RAX) {
if (reg == VCPU_REGS_RSP) {
@ -2229,7 +2230,7 @@ static int em_cmpxchg8b(struct x86_emulate_ctxt *ctxt)
static int em_ret(struct x86_emulate_ctxt *ctxt)
{
int rc;
unsigned long eip;
unsigned long eip = 0;
rc = emulate_pop(ctxt, &eip, ctxt->op_bytes);
if (rc != X86EMUL_CONTINUE)
@ -2241,7 +2242,8 @@ static int em_ret(struct x86_emulate_ctxt *ctxt)
static int em_ret_far(struct x86_emulate_ctxt *ctxt)
{
int rc;
unsigned long eip, cs;
unsigned long eip = 0;
unsigned long cs = 0;
int cpl = ctxt->ops->cpl(ctxt);
struct desc_struct new_desc;
@ -3011,7 +3013,7 @@ static int emulator_do_task_switch(struct x86_emulate_ctxt *ctxt,
ret = em_push(ctxt);
}
ops->get_dr(ctxt, 7, &dr7);
dr7 = ops->get_dr(ctxt, 7);
ops->set_dr(ctxt, 7, dr7 & ~(DR_LOCAL_ENABLE_MASK | DR_LOCAL_SLOWDOWN));
return ret;
@ -3184,7 +3186,7 @@ static int em_call_far(struct x86_emulate_ctxt *ctxt)
static int em_ret_near_imm(struct x86_emulate_ctxt *ctxt)
{
int rc;
unsigned long eip;
unsigned long eip = 0;
rc = emulate_pop(ctxt, &eip, ctxt->op_bytes);
if (rc != X86EMUL_CONTINUE)
@ -3866,15 +3868,6 @@ static int check_cr_access(struct x86_emulate_ctxt *ctxt)
return X86EMUL_CONTINUE;
}
static int check_dr7_gd(struct x86_emulate_ctxt *ctxt)
{
unsigned long dr7;
ctxt->ops->get_dr(ctxt, 7, &dr7);
return dr7 & DR7_GD;
}
static int check_dr_read(struct x86_emulate_ctxt *ctxt)
{
int dr = ctxt->modrm_reg;
@ -3887,10 +3880,10 @@ static int check_dr_read(struct x86_emulate_ctxt *ctxt)
if ((cr4 & X86_CR4_DE) && (dr == 4 || dr == 5))
return emulate_ud(ctxt);
if (check_dr7_gd(ctxt)) {
if (ctxt->ops->get_dr(ctxt, 7) & DR7_GD) {
ulong dr6;
ctxt->ops->get_dr(ctxt, 6, &dr6);
dr6 = ctxt->ops->get_dr(ctxt, 6);
dr6 &= ~DR_TRAP_BITS;
dr6 |= DR6_BD | DR6_ACTIVE_LOW;
ctxt->ops->set_dr(ctxt, 6, dr6);
@ -4505,11 +4498,11 @@ static const struct instr_dual instr_dual_0f_38_f1 = {
};
static const struct gprefix three_byte_0f_38_f0 = {
ID(0, &instr_dual_0f_38_f0), N, N, N
ID(0, &instr_dual_0f_38_f0), ID(0, &instr_dual_0f_38_f0), N, N
};
static const struct gprefix three_byte_0f_38_f1 = {
ID(0, &instr_dual_0f_38_f1), N, N, N
ID(0, &instr_dual_0f_38_f1), ID(0, &instr_dual_0f_38_f1), N, N
};
/*
@ -5449,7 +5442,7 @@ int x86_emulate_insn(struct x86_emulate_ctxt *ctxt)
ctxt->dst.val = ops->get_cr(ctxt, ctxt->modrm_reg);
break;
case 0x21: /* mov from dr to reg */
ops->get_dr(ctxt, ctxt->modrm_reg, &ctxt->dst.val);
ctxt->dst.val = ops->get_dr(ctxt, ctxt->modrm_reg);
break;
case 0x40 ... 0x4f: /* cmov */
if (test_cc(ctxt->b, ctxt->eflags))

2
arch/x86/kvm/kvm_emulate.h
View file

@ -203,7 +203,7 @@ struct x86_emulate_ops {
ulong (*get_cr)(struct x86_emulate_ctxt *ctxt, int cr);
int (*set_cr)(struct x86_emulate_ctxt *ctxt, int cr, ulong val);
int (*cpl)(struct x86_emulate_ctxt *ctxt);
void (*get_dr)(struct x86_emulate_ctxt *ctxt, int dr, ulong *dest);
ulong (*get_dr)(struct x86_emulate_ctxt *ctxt, int dr);
int (*set_dr)(struct x86_emulate_ctxt *ctxt, int dr, ulong value);
int (*set_msr_with_filter)(struct x86_emulate_ctxt *ctxt, u32 msr_index, u64 data);
int (*get_msr_with_filter)(struct x86_emulate_ctxt *ctxt, u32 msr_index, u64 *pdata);

27
arch/x86/kvm/lapic.c
View file

@ -124,6 +124,9 @@ static inline int __apic_test_and_clear_vector(int vec, void *bitmap)
return __test_and_clear_bit(VEC_POS(vec), (bitmap) + REG_POS(vec));
}
__read_mostly DEFINE_STATIC_KEY_FALSE(kvm_has_noapic_vcpu);
EXPORT_SYMBOL_GPL(kvm_has_noapic_vcpu);
__read_mostly DEFINE_STATIC_KEY_DEFERRED_FALSE(apic_hw_disabled, HZ);
__read_mostly DEFINE_STATIC_KEY_DEFERRED_FALSE(apic_sw_disabled, HZ);
@ -2466,8 +2469,10 @@ void kvm_free_lapic(struct kvm_vcpu *vcpu)
{
struct kvm_lapic *apic = vcpu->arch.apic;
if (!vcpu->arch.apic)
if (!vcpu->arch.apic) {
static_branch_dec(&kvm_has_noapic_vcpu);
return;
}
hrtimer_cancel(&apic->lapic_timer.timer);
@ -2809,6 +2814,11 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu, int timer_advance_ns)
ASSERT(vcpu != NULL);
if (!irqchip_in_kernel(vcpu->kvm)) {
static_branch_inc(&kvm_has_noapic_vcpu);
return 0;
}
apic = kzalloc(sizeof(*apic), GFP_KERNEL_ACCOUNT);
if (!apic)
goto nomem;
@ -2844,6 +2854,21 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu, int timer_advance_ns)
static_branch_inc(&apic_sw_disabled.key); /* sw disabled at reset */
kvm_iodevice_init(&apic->dev, &apic_mmio_ops);
/*
* Defer evaluating inhibits until the vCPU is first run, as this vCPU
* will not get notified of any changes until this vCPU is visible to
* other vCPUs (marked online and added to the set of vCPUs).
*
* Opportunistically mark APICv active as VMX in particularly is highly
* unlikely to have inhibits. Ignore the current per-VM APICv state so
* that vCPU creation is guaranteed to run with a deterministic value,
* the request will ensure the vCPU gets the correct state before VM-Entry.
*/
if (enable_apicv) {
apic->apicv_active = true;
kvm_make_request(KVM_REQ_APICV_UPDATE, vcpu);
}
return 0;
nomem_free_apic:
kfree(apic);

15
arch/x86/kvm/smm.c
View file

@ -184,7 +184,6 @@ static void enter_smm_save_state_32(struct kvm_vcpu *vcpu,
struct kvm_smram_state_32 *smram)
{
struct desc_ptr dt;
unsigned long val;
int i;
smram->cr0 = kvm_read_cr0(vcpu);
@ -195,10 +194,8 @@ static void enter_smm_save_state_32(struct kvm_vcpu *vcpu,
for (i = 0; i < 8; i++)
smram->gprs[i] = kvm_register_read_raw(vcpu, i);
kvm_get_dr(vcpu, 6, &val);
smram->dr6 = (u32)val;
kvm_get_dr(vcpu, 7, &val);
smram->dr7 = (u32)val;
smram->dr6 = (u32)vcpu->arch.dr6;
smram->dr7 = (u32)vcpu->arch.dr7;
enter_smm_save_seg_32(vcpu, &smram->tr, &smram->tr_sel, VCPU_SREG_TR);
enter_smm_save_seg_32(vcpu, &smram->ldtr, &smram->ldtr_sel, VCPU_SREG_LDTR);
@ -231,7 +228,6 @@ static void enter_smm_save_state_64(struct kvm_vcpu *vcpu,
struct kvm_smram_state_64 *smram)
{
struct desc_ptr dt;
unsigned long val;
int i;
for (i = 0; i < 16; i++)
@ -240,11 +236,8 @@ static void enter_smm_save_state_64(struct kvm_vcpu *vcpu,
smram->rip = kvm_rip_read(vcpu);
smram->rflags = kvm_get_rflags(vcpu);
kvm_get_dr(vcpu, 6, &val);
smram->dr6 = val;
kvm_get_dr(vcpu, 7, &val);
smram->dr7 = val;
smram->dr6 = vcpu->arch.dr6;
smram->dr7 = vcpu->arch.dr7;
smram->cr0 = kvm_read_cr0(vcpu);
smram->cr3 = kvm_read_cr3(vcpu);

25
arch/x86/kvm/svm/svm.c
View file

@ -2735,7 +2735,6 @@ static int dr_interception(struct kvm_vcpu *vcpu)
{
struct vcpu_svm *svm = to_svm(vcpu);
int reg, dr;
unsigned long val;
int err = 0;
/*
@ -2763,11 +2762,9 @@ static int dr_interception(struct kvm_vcpu *vcpu)
dr = svm->vmcb->control.exit_code - SVM_EXIT_READ_DR0;
if (dr >= 16) { /* mov to DRn */
dr -= 16;
val = kvm_register_read(vcpu, reg);
err = kvm_set_dr(vcpu, dr, val);
err = kvm_set_dr(vcpu, dr, kvm_register_read(vcpu, reg));
} else {
kvm_get_dr(vcpu, dr, &val);
kvm_register_write(vcpu, reg, val);
kvm_register_write(vcpu, reg, kvm_get_dr(vcpu, dr));
}
return kvm_complete_insn_gp(vcpu, err);
@ -4092,6 +4089,9 @@ static int svm_vcpu_pre_run(struct kvm_vcpu *vcpu)
static fastpath_t svm_exit_handlers_fastpath(struct kvm_vcpu *vcpu)
{
if (is_guest_mode(vcpu))
return EXIT_FASTPATH_NONE;
if (to_svm(vcpu)->vmcb->control.exit_code == SVM_EXIT_MSR &&
to_svm(vcpu)->vmcb->control.exit_info_1)
return handle_fastpath_set_msr_irqoff(vcpu);
@ -4115,12 +4115,13 @@ static noinstr void svm_vcpu_enter_exit(struct kvm_vcpu *vcpu, bool spec_ctrl_in
guest_state_exit_irqoff();
}
static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu)
static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu,
bool force_immediate_exit)
{
struct vcpu_svm *svm = to_svm(vcpu);
bool spec_ctrl_intercepted = msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL);
trace_kvm_entry(vcpu);
trace_kvm_entry(vcpu, force_immediate_exit);
svm->vmcb->save.rax = vcpu->arch.regs[VCPU_REGS_RAX];
svm->vmcb->save.rsp = vcpu->arch.regs[VCPU_REGS_RSP];
@ -4139,9 +4140,12 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu)
* is enough to force an immediate vmexit.
*/
disable_nmi_singlestep(svm);
smp_send_reschedule(vcpu->cpu);
force_immediate_exit = true;
}
if (force_immediate_exit)
smp_send_reschedule(vcpu->cpu);
pre_svm_run(vcpu);
sync_lapic_to_cr8(vcpu);
@ -4237,9 +4241,6 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu)
svm_complete_interrupts(vcpu);
if (is_guest_mode(vcpu))
return EXIT_FASTPATH_NONE;
return svm_exit_handlers_fastpath(vcpu);
}
@ -4997,8 +4998,6 @@ static struct kvm_x86_ops svm_x86_ops __initdata = {
.check_intercept = svm_check_intercept,
.handle_exit_irqoff = svm_handle_exit_irqoff,
.request_immediate_exit = __kvm_request_immediate_exit,
.sched_in = svm_sched_in,
.nested_ops = &svm_nested_ops,

9
arch/x86/kvm/trace.h
View file

@ -15,20 +15,23 @@
* Tracepoint for guest mode entry.
*/
TRACE_EVENT(kvm_entry,
TP_PROTO(struct kvm_vcpu *vcpu),
TP_ARGS(vcpu),
TP_PROTO(struct kvm_vcpu *vcpu, bool force_immediate_exit),
TP_ARGS(vcpu, force_immediate_exit),
TP_STRUCT__entry(
__field( unsigned int, vcpu_id )
__field( unsigned long, rip )
__field( bool, immediate_exit )
),
TP_fast_assign(
__entry->vcpu_id = vcpu->vcpu_id;
__entry->rip = kvm_rip_read(vcpu);
__entry->immediate_exit = force_immediate_exit;
),
TP_printk("vcpu %u, rip 0x%lx", __entry->vcpu_id, __entry->rip)
TP_printk("vcpu %u, rip 0x%lx%s", __entry->vcpu_id, __entry->rip,
__entry->immediate_exit ? "[immediate exit]" : "")
);
/*

2
arch/x86/kvm/vmx/nested.c
View file

@ -4433,7 +4433,7 @@ static void sync_vmcs02_to_vmcs12(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
(vm_entry_controls_get(to_vmx(vcpu)) & VM_ENTRY_IA32E_MODE);
if (vmcs12->vm_exit_controls & VM_EXIT_SAVE_DEBUG_CONTROLS)
kvm_get_dr(vcpu, 7, (unsigned long *)&vmcs12->guest_dr7);
vmcs12->guest_dr7 = vcpu->arch.dr7;
if (vmcs12->vm_exit_controls & VM_EXIT_SAVE_IA32_EFER)
vmcs12->guest_ia32_efer = vcpu->arch.efer;

85
arch/x86/kvm/vmx/vmx.c
View file

@ -49,6 +49,8 @@
#include <asm/spec-ctrl.h>
#include <asm/vmx.h>
#include <trace/events/ipi.h>
#include "capabilities.h"
#include "cpuid.h"
#include "hyperv.h"
@ -1290,8 +1292,6 @@ void vmx_prepare_switch_to_guest(struct kvm_vcpu *vcpu)
u16 fs_sel, gs_sel;
int i;
vmx->req_immediate_exit = false;
/*
* Note that guest MSRs to be saved/restored can also be changed
* when guest state is loaded. This happens when guest transitions
@ -5575,10 +5575,7 @@ static int handle_dr(struct kvm_vcpu *vcpu)
reg = DEBUG_REG_ACCESS_REG(exit_qualification);
if (exit_qualification & TYPE_MOV_FROM_DR) {
unsigned long val;
kvm_get_dr(vcpu, dr, &val);
kvm_register_write(vcpu, reg, val);
kvm_register_write(vcpu, reg, kvm_get_dr(vcpu, dr));
err = 0;
} else {
err = kvm_set_dr(vcpu, dr, kvm_register_read(vcpu, reg));
@ -6000,22 +5997,46 @@ static int handle_pml_full(struct kvm_vcpu *vcpu)
return 1;
}
static fastpath_t handle_fastpath_preemption_timer(struct kvm_vcpu *vcpu)
static fastpath_t handle_fastpath_preemption_timer(struct kvm_vcpu *vcpu,
bool force_immediate_exit)
{
struct vcpu_vmx *vmx = to_vmx(vcpu);
if (!vmx->req_immediate_exit &&
!unlikely(vmx->loaded_vmcs->hv_timer_soft_disabled)) {
kvm_lapic_expired_hv_timer(vcpu);
/*
* In the *extremely* unlikely scenario that this is a spurious VM-Exit
* due to the timer expiring while it was "soft" disabled, just eat the
* exit and re-enter the guest.
*/
if (unlikely(vmx->loaded_vmcs->hv_timer_soft_disabled))
return EXIT_FASTPATH_REENTER_GUEST;
}
return EXIT_FASTPATH_NONE;
/*
* If the timer expired because KVM used it to force an immediate exit,
* then mission accomplished.
*/
if (force_immediate_exit)
return EXIT_FASTPATH_EXIT_HANDLED;
/*
* If L2 is active, go down the slow path as emulating the guest timer
* expiration likely requires synthesizing a nested VM-Exit.
*/
if (is_guest_mode(vcpu))
return EXIT_FASTPATH_NONE;
kvm_lapic_expired_hv_timer(vcpu);
return EXIT_FASTPATH_REENTER_GUEST;
}
static int handle_preemption_timer(struct kvm_vcpu *vcpu)
{
handle_fastpath_preemption_timer(vcpu);
/*
* This non-fastpath handler is reached if and only if the preemption
* timer was being used to emulate a guest timer while L2 is active.
* All other scenarios are supposed to be handled in the fastpath.
*/
WARN_ON_ONCE(!is_guest_mode(vcpu));
kvm_lapic_expired_hv_timer(vcpu);
return 1;
}
@ -7155,13 +7176,13 @@ static void atomic_switch_perf_msrs(struct vcpu_vmx *vmx)
msrs[i].host, false);
}
static void vmx_update_hv_timer(struct kvm_vcpu *vcpu)
static void vmx_update_hv_timer(struct kvm_vcpu *vcpu, bool force_immediate_exit)
{
struct vcpu_vmx *vmx = to_vmx(vcpu);
u64 tscl;
u32 delta_tsc;
if (vmx->req_immediate_exit) {
if (force_immediate_exit) {
vmcs_write32(VMX_PREEMPTION_TIMER_VALUE, 0);
vmx->loaded_vmcs->hv_timer_soft_disabled = false;
} else if (vmx->hv_deadline_tsc != -1) {
@ -7214,13 +7235,22 @@ void noinstr vmx_spec_ctrl_restore_host(struct vcpu_vmx *vmx,
barrier_nospec();
}
static fastpath_t vmx_exit_handlers_fastpath(struct kvm_vcpu *vcpu)
static fastpath_t vmx_exit_handlers_fastpath(struct kvm_vcpu *vcpu,
bool force_immediate_exit)
{
/*
* If L2 is active, some VMX preemption timer exits can be handled in
* the fastpath even, all other exits must use the slow path.
*/
if (is_guest_mode(vcpu) &&
to_vmx(vcpu)->exit_reason.basic != EXIT_REASON_PREEMPTION_TIMER)
return EXIT_FASTPATH_NONE;
switch (to_vmx(vcpu)->exit_reason.basic) {
case EXIT_REASON_MSR_WRITE:
return handle_fastpath_set_msr_irqoff(vcpu);
case EXIT_REASON_PREEMPTION_TIMER:
return handle_fastpath_preemption_timer(vcpu);
return handle_fastpath_preemption_timer(vcpu, force_immediate_exit);
default:
return EXIT_FASTPATH_NONE;
}
@ -7280,7 +7310,7 @@ static noinstr void vmx_vcpu_enter_exit(struct kvm_vcpu *vcpu,
guest_state_exit_irqoff();
}
static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu)
static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu, bool force_immediate_exit)
{
struct vcpu_vmx *vmx = to_vmx(vcpu);
unsigned long cr3, cr4;
@ -7307,7 +7337,7 @@ static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu)
return EXIT_FASTPATH_NONE;
}
trace_kvm_entry(vcpu);
trace_kvm_entry(vcpu, force_immediate_exit);
if (vmx->ple_window_dirty) {
vmx->ple_window_dirty = false;
@ -7366,7 +7396,9 @@ static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu)
vmx_passthrough_lbr_msrs(vcpu);
if (enable_preemption_timer)
vmx_update_hv_timer(vcpu);
vmx_update_hv_timer(vcpu, force_immediate_exit);
else if (force_immediate_exit)
smp_send_reschedule(vcpu->cpu);
kvm_wait_lapic_expire(vcpu);
@ -7430,10 +7462,7 @@ static fastpath_t vmx_vcpu_run(struct kvm_vcpu *vcpu)
vmx_recover_nmi_blocking(vmx);
vmx_complete_interrupts(vmx);
if (is_guest_mode(vcpu))
return EXIT_FASTPATH_NONE;
return vmx_exit_handlers_fastpath(vcpu);
return vmx_exit_handlers_fastpath(vcpu, force_immediate_exit);
}
static void vmx_vcpu_free(struct kvm_vcpu *vcpu)
@ -7913,11 +7942,6 @@ static __init void vmx_set_cpu_caps(void)
kvm_cpu_cap_check_and_set(X86_FEATURE_WAITPKG);
}
static void vmx_request_immediate_exit(struct kvm_vcpu *vcpu)
{
to_vmx(vcpu)->req_immediate_exit = true;
}
static int vmx_check_intercept_io(struct kvm_vcpu *vcpu,
struct x86_instruction_info *info)
{
@ -8370,8 +8394,6 @@ static struct kvm_x86_ops vmx_x86_ops __initdata = {
.check_intercept = vmx_check_intercept,
.handle_exit_irqoff = vmx_handle_exit_irqoff,
.request_immediate_exit = vmx_request_immediate_exit,
.sched_in = vmx_sched_in,
.cpu_dirty_log_size = PML_ENTITY_NUM,
@ -8631,7 +8653,6 @@ static __init int hardware_setup(void)
if (!enable_preemption_timer) {
vmx_x86_ops.set_hv_timer = NULL;
vmx_x86_ops.cancel_hv_timer = NULL;
vmx_x86_ops.request_immediate_exit = __kvm_request_immediate_exit;
}
kvm_caps.supported_mce_cap |= MCG_LMCE_P;

2
arch/x86/kvm/vmx/vmx.h
View file

@ -332,8 +332,6 @@ struct vcpu_vmx {
unsigned int ple_window;
bool ple_window_dirty;
bool req_immediate_exit;
/* Support for PML */
#define PML_ENTITY_NUM 512
struct page *pml_pg;

110
arch/x86/kvm/x86.c
View file

@ -1399,22 +1399,19 @@ int kvm_set_dr(struct kvm_vcpu *vcpu, int dr, unsigned long val)
}
EXPORT_SYMBOL_GPL(kvm_set_dr);
void kvm_get_dr(struct kvm_vcpu *vcpu, int dr, unsigned long *val)
unsigned long kvm_get_dr(struct kvm_vcpu *vcpu, int dr)
{
size_t size = ARRAY_SIZE(vcpu->arch.db);
switch (dr) {
case 0 ... 3:
*val = vcpu->arch.db[array_index_nospec(dr, size)];
break;
return vcpu->arch.db[array_index_nospec(dr, size)];
case 4:
case 6:
*val = vcpu->arch.dr6;
break;
return vcpu->arch.dr6;
case 5:
default: /* 7 */
*val = vcpu->arch.dr7;
break;
return vcpu->arch.dr7;
}
}
EXPORT_SYMBOL_GPL(kvm_get_dr);
@ -5061,8 +5058,7 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
int idx;
if (vcpu->preempted) {
if (!vcpu->arch.guest_state_protected)
vcpu->arch.preempted_in_kernel = !static_call(kvm_x86_get_cpl)(vcpu);
vcpu->arch.preempted_in_kernel = kvm_arch_vcpu_in_kernel(vcpu);
/*
* Take the srcu lock as memslots will be accessed to check the gfn
@ -5509,18 +5505,23 @@ static int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu,
static void kvm_vcpu_ioctl_x86_get_debugregs(struct kvm_vcpu *vcpu,
struct kvm_debugregs *dbgregs)
{
unsigned long val;
unsigned int i;
memset(dbgregs, 0, sizeof(*dbgregs));
memcpy(dbgregs->db, vcpu->arch.db, sizeof(vcpu->arch.db));
kvm_get_dr(vcpu, 6, &val);
dbgregs->dr6 = val;
BUILD_BUG_ON(ARRAY_SIZE(vcpu->arch.db) != ARRAY_SIZE(dbgregs->db));
for (i = 0; i < ARRAY_SIZE(vcpu->arch.db); i++)
dbgregs->db[i] = vcpu->arch.db[i];
dbgregs->dr6 = vcpu->arch.dr6;
dbgregs->dr7 = vcpu->arch.dr7;
}
static int kvm_vcpu_ioctl_x86_set_debugregs(struct kvm_vcpu *vcpu,
struct kvm_debugregs *dbgregs)
{
unsigned int i;
if (dbgregs->flags)
return -EINVAL;
@ -5529,7 +5530,9 @@ static int kvm_vcpu_ioctl_x86_set_debugregs(struct kvm_vcpu *vcpu,
if (!kvm_dr7_valid(dbgregs->dr7))
return -EINVAL;
memcpy(vcpu->arch.db, dbgregs->db, sizeof(vcpu->arch.db));
for (i = 0; i < ARRAY_SIZE(vcpu->arch.db); i++)
vcpu->arch.db[i] = dbgregs->db[i];
kvm_update_dr0123(vcpu);
vcpu->arch.dr6 = dbgregs->dr6;
vcpu->arch.dr7 = dbgregs->dr7;
@ -8167,10 +8170,9 @@ static void emulator_wbinvd(struct x86_emulate_ctxt *ctxt)
kvm_emulate_wbinvd_noskip(emul_to_vcpu(ctxt));
}
static void emulator_get_dr(struct x86_emulate_ctxt *ctxt, int dr,
unsigned long *dest)
static unsigned long emulator_get_dr(struct x86_emulate_ctxt *ctxt, int dr)
{
kvm_get_dr(emul_to_vcpu(ctxt), dr, dest);
return kvm_get_dr(emul_to_vcpu(ctxt), dr);
}
static int emulator_set_dr(struct x86_emulate_ctxt *ctxt, int dr,
@ -9635,11 +9637,13 @@ static void kvm_x86_check_cpu_compat(void *ret)
*(int *)ret = kvm_x86_check_processor_compatibility();
}
static int __kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
int kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
{
u64 host_pat;
int r, cpu;
guard(mutex)(&vendor_module_lock);
if (kvm_x86_ops.hardware_enable) {
pr_err("already loaded vendor module '%s'\n", kvm_x86_ops.name);
return -EEXIST;
@ -9769,17 +9773,6 @@ static int __kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
kmem_cache_destroy(x86_emulator_cache);
return r;
}
int kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
{
int r;
mutex_lock(&vendor_module_lock);
r = __kvm_x86_vendor_init(ops);
mutex_unlock(&vendor_module_lock);
return r;
}
EXPORT_SYMBOL_GPL(kvm_x86_vendor_init);
void kvm_x86_vendor_exit(void)
@ -10676,12 +10669,6 @@ static void kvm_vcpu_reload_apic_access_page(struct kvm_vcpu *vcpu)
static_call_cond(kvm_x86_set_apic_access_page_addr)(vcpu);
}
void __kvm_request_immediate_exit(struct kvm_vcpu *vcpu)
{
smp_send_reschedule(vcpu->cpu);
}
EXPORT_SYMBOL_GPL(__kvm_request_immediate_exit);
/*
* Called within kvm->srcu read side.
* Returns 1 to let vcpu_run() continue the guest execution loop without
@ -10931,10 +10918,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
goto cancel_injection;
}
if (req_immediate_exit) {
if (req_immediate_exit)
kvm_make_request(KVM_REQ_EVENT, vcpu);
static_call(kvm_x86_request_immediate_exit)(vcpu);
}
fpregs_assert_state_consistent();
if (test_thread_flag(TIF_NEED_FPU_LOAD))
@ -10965,7 +10950,7 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
WARN_ON_ONCE((kvm_vcpu_apicv_activated(vcpu) != kvm_vcpu_apicv_active(vcpu)) &&
(kvm_get_apic_mode(vcpu) != LAPIC_MODE_DISABLED));
exit_fastpath = static_call(kvm_x86_vcpu_run)(vcpu);
exit_fastpath = static_call(kvm_x86_vcpu_run)(vcpu, req_immediate_exit);
if (likely(exit_fastpath != EXIT_FASTPATH_REENTER_GUEST))
break;
@ -12063,27 +12048,9 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu)
if (r < 0)
return r;
if (irqchip_in_kernel(vcpu->kvm)) {
r = kvm_create_lapic(vcpu, lapic_timer_advance_ns);
if (r < 0)
goto fail_mmu_destroy;
/*
* Defer evaluating inhibits until the vCPU is first run, as
* this vCPU will not get notified of any changes until this
* vCPU is visible to other vCPUs (marked online and added to
* the set of vCPUs). Opportunistically mark APICv active as
* VMX in particularly is highly unlikely to have inhibits.
* Ignore the current per-VM APICv state so that vCPU creation
* is guaranteed to run with a deterministic value, the request
* will ensure the vCPU gets the correct state before VM-Entry.
*/
if (enable_apicv) {
vcpu->arch.apic->apicv_active = true;
kvm_make_request(KVM_REQ_APICV_UPDATE, vcpu);
}
} else
static_branch_inc(&kvm_has_noapic_vcpu);
r = kvm_create_lapic(vcpu, lapic_timer_advance_ns);
if (r < 0)
goto fail_mmu_destroy;
r = -ENOMEM;
@ -12204,8 +12171,6 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
srcu_read_unlock(&vcpu->kvm->srcu, idx);
free_page((unsigned long)vcpu->arch.pio_data);
kvfree(vcpu->arch.cpuid_entries);
if (!lapic_in_kernel(vcpu))
static_branch_dec(&kvm_has_noapic_vcpu);
}
void kvm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
@ -12482,9 +12447,6 @@ bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu)
return (vcpu->arch.apic_base & MSR_IA32_APICBASE_BSP) != 0;
}
__read_mostly DEFINE_STATIC_KEY_FALSE(kvm_has_noapic_vcpu);
EXPORT_SYMBOL_GPL(kvm_has_noapic_vcpu);
void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu)
{
struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
@ -13087,11 +13049,13 @@ int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
bool kvm_arch_dy_has_pending_interrupt(struct kvm_vcpu *vcpu)
{
if (kvm_vcpu_apicv_active(vcpu) &&
static_call(kvm_x86_dy_apicv_has_pending_interrupt)(vcpu))
return true;
return kvm_vcpu_apicv_active(vcpu) &&
static_call(kvm_x86_dy_apicv_has_pending_interrupt)(vcpu);
}
return false;
bool kvm_arch_vcpu_preempted_in_kernel(struct kvm_vcpu *vcpu)
{
return vcpu->arch.preempted_in_kernel;
}
bool kvm_arch_dy_runnable(struct kvm_vcpu *vcpu)
@ -13114,9 +13078,6 @@ bool kvm_arch_vcpu_in_kernel(struct kvm_vcpu *vcpu)
if (vcpu->arch.guest_state_protected)
return true;
if (vcpu != kvm_get_running_vcpu())
return vcpu->arch.preempted_in_kernel;
return static_call(kvm_x86_get_cpl)(vcpu) == 0;
}
@ -13911,9 +13872,6 @@ module_init(kvm_x86_init);
static void __exit kvm_x86_exit(void)
{
/*
* If module_init() is implemented, module_exit() must also be
* implemented to allow module unload.
*/
WARN_ON_ONCE(static_branch_unlikely(&kvm_has_noapic_vcpu));
}
module_exit(kvm_x86_exit);

1
include/linux/kvm_host.h
View file

@ -1504,6 +1504,7 @@ bool kvm_arch_vcpu_in_kernel(struct kvm_vcpu *vcpu);
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
bool kvm_arch_dy_runnable(struct kvm_vcpu *vcpu);
bool kvm_arch_dy_has_pending_interrupt(struct kvm_vcpu *vcpu);
bool kvm_arch_vcpu_preempted_in_kernel(struct kvm_vcpu *vcpu);
int kvm_arch_post_init_vm(struct kvm *kvm);
void kvm_arch_pre_destroy_vm(struct kvm *kvm);
void kvm_arch_create_vm_debugfs(struct kvm *kvm);

21
virt/kvm/kvm_main.c
View file

@ -4046,6 +4046,18 @@ static bool vcpu_dy_runnable(struct kvm_vcpu *vcpu)
return false;
}
/*
* By default, simply query the target vCPU's current mode when checking if a
* vCPU was preempted in kernel mode. All architectures except x86 (or more
* specifical, except VMX) allow querying whether or not a vCPU is in kernel
* mode even if the vCPU is NOT loaded, i.e. using kvm_arch_vcpu_in_kernel()
* directly for cross-vCPU checks is functionally correct and accurate.
*/
bool __weak kvm_arch_vcpu_preempted_in_kernel(struct kvm_vcpu *vcpu)
{
return kvm_arch_vcpu_in_kernel(vcpu);
}
bool __weak kvm_arch_dy_has_pending_interrupt(struct kvm_vcpu *vcpu)
{
return false;
@ -4082,9 +4094,16 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *me, bool yield_to_kernel_mode)
continue;
if (kvm_vcpu_is_blocking(vcpu) && !vcpu_dy_runnable(vcpu))
continue;
/*
* Treat the target vCPU as being in-kernel if it has a
* pending interrupt, as the vCPU trying to yield may
* be spinning waiting on IPI delivery, i.e. the target
* vCPU is in-kernel for the purposes of directed yield.
*/
if (READ_ONCE(vcpu->preempted) && yield_to_kernel_mode &&
!kvm_arch_dy_has_pending_interrupt(vcpu) &&
!kvm_arch_vcpu_in_kernel(vcpu))
!kvm_arch_vcpu_preempted_in_kernel(vcpu))
continue;
if (!kvm_vcpu_eligible_for_directed_yield(vcpu))
continue;