qemu/target-arm/kvm32.c
Pavel Fedin eb5e1d3c85 target-arm: Use the kernel's idea of MPIDR if we're using KVM
When we're using KVM, the kernel's internal idea of the MPIDR
affinity fields must match the values we tell it for the guest
vcpu cluster configuration in the device tree. Since at the moment
the kernel doesn't support letting userspace tell it the correct
affinity fields to use, we must read the kernel's view and
reflect that back in the device tree.

Signed-off-by: Shlomo Pongratz <shlomo.pongratz@huawei.com>
Signed-off-by: Pavel Fedin <p.fedin@samsung.com>
Message-id: 02f601d0a1e6$90c7d630$b2578290$@samsung.com
[PMM: Use a local #define rather than a global variable for
 the TCG ARM_CPUS_PER_CLUSTER setting. Tweak a comment. Update the
 commit message.]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2015-06-15 18:06:09 +01:00

451 lines
14 KiB
C

/*
* ARM implementation of KVM hooks, 32 bit specific code.
*
* Copyright Christoffer Dall 2009-2010
*
* This work is licensed under the terms of the GNU GPL, version 2 or later.
* See the COPYING file in the top-level directory.
*
*/
#include <stdio.h>
#include <sys/types.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
#include <linux/kvm.h>
#include "qemu-common.h"
#include "qemu/timer.h"
#include "sysemu/sysemu.h"
#include "sysemu/kvm.h"
#include "kvm_arm.h"
#include "cpu.h"
#include "internals.h"
#include "hw/arm/arm.h"
static inline void set_feature(uint64_t *features, int feature)
{
*features |= 1ULL << feature;
}
bool kvm_arm_get_host_cpu_features(ARMHostCPUClass *ahcc)
{
/* Identify the feature bits corresponding to the host CPU, and
* fill out the ARMHostCPUClass fields accordingly. To do this
* we have to create a scratch VM, create a single CPU inside it,
* and then query that CPU for the relevant ID registers.
*/
int i, ret, fdarray[3];
uint32_t midr, id_pfr0, id_isar0, mvfr1;
uint64_t features = 0;
/* Old kernels may not know about the PREFERRED_TARGET ioctl: however
* we know these will only support creating one kind of guest CPU,
* which is its preferred CPU type.
*/
static const uint32_t cpus_to_try[] = {
QEMU_KVM_ARM_TARGET_CORTEX_A15,
QEMU_KVM_ARM_TARGET_NONE
};
struct kvm_vcpu_init init;
struct kvm_one_reg idregs[] = {
{
.id = KVM_REG_ARM | KVM_REG_SIZE_U32
| ENCODE_CP_REG(15, 0, 0, 0, 0, 0, 0),
.addr = (uintptr_t)&midr,
},
{
.id = KVM_REG_ARM | KVM_REG_SIZE_U32
| ENCODE_CP_REG(15, 0, 0, 0, 1, 0, 0),
.addr = (uintptr_t)&id_pfr0,
},
{
.id = KVM_REG_ARM | KVM_REG_SIZE_U32
| ENCODE_CP_REG(15, 0, 0, 0, 2, 0, 0),
.addr = (uintptr_t)&id_isar0,
},
{
.id = KVM_REG_ARM | KVM_REG_SIZE_U32
| KVM_REG_ARM_VFP | KVM_REG_ARM_VFP_MVFR1,
.addr = (uintptr_t)&mvfr1,
},
};
if (!kvm_arm_create_scratch_host_vcpu(cpus_to_try, fdarray, &init)) {
return false;
}
ahcc->target = init.target;
/* This is not strictly blessed by the device tree binding docs yet,
* but in practice the kernel does not care about this string so
* there is no point maintaining an KVM_ARM_TARGET_* -> string table.
*/
ahcc->dtb_compatible = "arm,arm-v7";
for (i = 0; i < ARRAY_SIZE(idregs); i++) {
ret = ioctl(fdarray[2], KVM_GET_ONE_REG, &idregs[i]);
if (ret) {
break;
}
}
kvm_arm_destroy_scratch_host_vcpu(fdarray);
if (ret) {
return false;
}
/* Now we've retrieved all the register information we can
* set the feature bits based on the ID register fields.
* We can assume any KVM supporting CPU is at least a v7
* with VFPv3, LPAE and the generic timers; this in turn implies
* most of the other feature bits, but a few must be tested.
*/
set_feature(&features, ARM_FEATURE_V7);
set_feature(&features, ARM_FEATURE_VFP3);
set_feature(&features, ARM_FEATURE_LPAE);
set_feature(&features, ARM_FEATURE_GENERIC_TIMER);
switch (extract32(id_isar0, 24, 4)) {
case 1:
set_feature(&features, ARM_FEATURE_THUMB_DIV);
break;
case 2:
set_feature(&features, ARM_FEATURE_ARM_DIV);
set_feature(&features, ARM_FEATURE_THUMB_DIV);
break;
default:
break;
}
if (extract32(id_pfr0, 12, 4) == 1) {
set_feature(&features, ARM_FEATURE_THUMB2EE);
}
if (extract32(mvfr1, 20, 4) == 1) {
set_feature(&features, ARM_FEATURE_VFP_FP16);
}
if (extract32(mvfr1, 12, 4) == 1) {
set_feature(&features, ARM_FEATURE_NEON);
}
if (extract32(mvfr1, 28, 4) == 1) {
/* FMAC support implies VFPv4 */
set_feature(&features, ARM_FEATURE_VFP4);
}
ahcc->features = features;
return true;
}
bool kvm_arm_reg_syncs_via_cpreg_list(uint64_t regidx)
{
/* Return true if the regidx is a register we should synchronize
* via the cpreg_tuples array (ie is not a core reg we sync by
* hand in kvm_arch_get/put_registers())
*/
switch (regidx & KVM_REG_ARM_COPROC_MASK) {
case KVM_REG_ARM_CORE:
case KVM_REG_ARM_VFP:
return false;
default:
return true;
}
}
#define ARM_MPIDR_HWID_BITMASK 0xFFFFFF
#define ARM_CPU_ID_MPIDR 0, 0, 0, 5
int kvm_arch_init_vcpu(CPUState *cs)
{
int ret;
uint64_t v;
uint32_t mpidr;
struct kvm_one_reg r;
ARMCPU *cpu = ARM_CPU(cs);
if (cpu->kvm_target == QEMU_KVM_ARM_TARGET_NONE) {
fprintf(stderr, "KVM is not supported for this guest CPU type\n");
return -EINVAL;
}
/* Determine init features for this CPU */
memset(cpu->kvm_init_features, 0, sizeof(cpu->kvm_init_features));
if (cpu->start_powered_off) {
cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_POWER_OFF;
}
if (kvm_check_extension(cs->kvm_state, KVM_CAP_ARM_PSCI_0_2)) {
cpu->psci_version = 2;
cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_PSCI_0_2;
}
/* Do KVM_ARM_VCPU_INIT ioctl */
ret = kvm_arm_vcpu_init(cs);
if (ret) {
return ret;
}
/* Query the kernel to make sure it supports 32 VFP
* registers: QEMU's "cortex-a15" CPU is always a
* VFP-D32 core. The simplest way to do this is just
* to attempt to read register d31.
*/
r.id = KVM_REG_ARM | KVM_REG_SIZE_U64 | KVM_REG_ARM_VFP | 31;
r.addr = (uintptr_t)(&v);
ret = kvm_vcpu_ioctl(cs, KVM_GET_ONE_REG, &r);
if (ret == -ENOENT) {
return -EINVAL;
}
/*
* When KVM is in use, PSCI is emulated in-kernel and not by qemu.
* Currently KVM has its own idea about MPIDR assignment, so we
* override our defaults with what we get from KVM.
*/
ret = kvm_get_one_reg(cs, ARM_CP15_REG32(ARM_CPU_ID_MPIDR), &mpidr);
if (ret) {
return ret;
}
cpu->mp_affinity = mpidr & ARM_MPIDR_HWID_BITMASK;
return kvm_arm_init_cpreg_list(cpu);
}
typedef struct Reg {
uint64_t id;
int offset;
} Reg;
#define COREREG(KERNELNAME, QEMUFIELD) \
{ \
KVM_REG_ARM | KVM_REG_SIZE_U32 | \
KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(KERNELNAME), \
offsetof(CPUARMState, QEMUFIELD) \
}
#define VFPSYSREG(R) \
{ \
KVM_REG_ARM | KVM_REG_SIZE_U32 | KVM_REG_ARM_VFP | \
KVM_REG_ARM_VFP_##R, \
offsetof(CPUARMState, vfp.xregs[ARM_VFP_##R]) \
}
/* Like COREREG, but handle fields which are in a uint64_t in CPUARMState. */
#define COREREG64(KERNELNAME, QEMUFIELD) \
{ \
KVM_REG_ARM | KVM_REG_SIZE_U32 | \
KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(KERNELNAME), \
offsetoflow32(CPUARMState, QEMUFIELD) \
}
static const Reg regs[] = {
/* R0_usr .. R14_usr */
COREREG(usr_regs.uregs[0], regs[0]),
COREREG(usr_regs.uregs[1], regs[1]),
COREREG(usr_regs.uregs[2], regs[2]),
COREREG(usr_regs.uregs[3], regs[3]),
COREREG(usr_regs.uregs[4], regs[4]),
COREREG(usr_regs.uregs[5], regs[5]),
COREREG(usr_regs.uregs[6], regs[6]),
COREREG(usr_regs.uregs[7], regs[7]),
COREREG(usr_regs.uregs[8], usr_regs[0]),
COREREG(usr_regs.uregs[9], usr_regs[1]),
COREREG(usr_regs.uregs[10], usr_regs[2]),
COREREG(usr_regs.uregs[11], usr_regs[3]),
COREREG(usr_regs.uregs[12], usr_regs[4]),
COREREG(usr_regs.uregs[13], banked_r13[0]),
COREREG(usr_regs.uregs[14], banked_r14[0]),
/* R13, R14, SPSR for SVC, ABT, UND, IRQ banks */
COREREG(svc_regs[0], banked_r13[1]),
COREREG(svc_regs[1], banked_r14[1]),
COREREG64(svc_regs[2], banked_spsr[1]),
COREREG(abt_regs[0], banked_r13[2]),
COREREG(abt_regs[1], banked_r14[2]),
COREREG64(abt_regs[2], banked_spsr[2]),
COREREG(und_regs[0], banked_r13[3]),
COREREG(und_regs[1], banked_r14[3]),
COREREG64(und_regs[2], banked_spsr[3]),
COREREG(irq_regs[0], banked_r13[4]),
COREREG(irq_regs[1], banked_r14[4]),
COREREG64(irq_regs[2], banked_spsr[4]),
/* R8_fiq .. R14_fiq and SPSR_fiq */
COREREG(fiq_regs[0], fiq_regs[0]),
COREREG(fiq_regs[1], fiq_regs[1]),
COREREG(fiq_regs[2], fiq_regs[2]),
COREREG(fiq_regs[3], fiq_regs[3]),
COREREG(fiq_regs[4], fiq_regs[4]),
COREREG(fiq_regs[5], banked_r13[5]),
COREREG(fiq_regs[6], banked_r14[5]),
COREREG64(fiq_regs[7], banked_spsr[5]),
/* R15 */
COREREG(usr_regs.uregs[15], regs[15]),
/* VFP system registers */
VFPSYSREG(FPSID),
VFPSYSREG(MVFR1),
VFPSYSREG(MVFR0),
VFPSYSREG(FPEXC),
VFPSYSREG(FPINST),
VFPSYSREG(FPINST2),
};
int kvm_arch_put_registers(CPUState *cs, int level)
{
ARMCPU *cpu = ARM_CPU(cs);
CPUARMState *env = &cpu->env;
struct kvm_one_reg r;
int mode, bn;
int ret, i;
uint32_t cpsr, fpscr;
/* Make sure the banked regs are properly set */
mode = env->uncached_cpsr & CPSR_M;
bn = bank_number(mode);
if (mode == ARM_CPU_MODE_FIQ) {
memcpy(env->fiq_regs, env->regs + 8, 5 * sizeof(uint32_t));
} else {
memcpy(env->usr_regs, env->regs + 8, 5 * sizeof(uint32_t));
}
env->banked_r13[bn] = env->regs[13];
env->banked_r14[bn] = env->regs[14];
env->banked_spsr[bn] = env->spsr;
/* Now we can safely copy stuff down to the kernel */
for (i = 0; i < ARRAY_SIZE(regs); i++) {
r.id = regs[i].id;
r.addr = (uintptr_t)(env) + regs[i].offset;
ret = kvm_vcpu_ioctl(cs, KVM_SET_ONE_REG, &r);
if (ret) {
return ret;
}
}
/* Special cases which aren't a single CPUARMState field */
cpsr = cpsr_read(env);
r.id = KVM_REG_ARM | KVM_REG_SIZE_U32 |
KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(usr_regs.ARM_cpsr);
r.addr = (uintptr_t)(&cpsr);
ret = kvm_vcpu_ioctl(cs, KVM_SET_ONE_REG, &r);
if (ret) {
return ret;
}
/* VFP registers */
r.id = KVM_REG_ARM | KVM_REG_SIZE_U64 | KVM_REG_ARM_VFP;
for (i = 0; i < 32; i++) {
r.addr = (uintptr_t)(&env->vfp.regs[i]);
ret = kvm_vcpu_ioctl(cs, KVM_SET_ONE_REG, &r);
if (ret) {
return ret;
}
r.id++;
}
r.id = KVM_REG_ARM | KVM_REG_SIZE_U32 | KVM_REG_ARM_VFP |
KVM_REG_ARM_VFP_FPSCR;
fpscr = vfp_get_fpscr(env);
r.addr = (uintptr_t)&fpscr;
ret = kvm_vcpu_ioctl(cs, KVM_SET_ONE_REG, &r);
if (ret) {
return ret;
}
/* Note that we do not call write_cpustate_to_list()
* here, so we are only writing the tuple list back to
* KVM. This is safe because nothing can change the
* CPUARMState cp15 fields (in particular gdb accesses cannot)
* and so there are no changes to sync. In fact syncing would
* be wrong at this point: for a constant register where TCG and
* KVM disagree about its value, the preceding write_list_to_cpustate()
* would not have had any effect on the CPUARMState value (since the
* register is read-only), and a write_cpustate_to_list() here would
* then try to write the TCG value back into KVM -- this would either
* fail or incorrectly change the value the guest sees.
*
* If we ever want to allow the user to modify cp15 registers via
* the gdb stub, we would need to be more clever here (for instance
* tracking the set of registers kvm_arch_get_registers() successfully
* managed to update the CPUARMState with, and only allowing those
* to be written back up into the kernel).
*/
if (!write_list_to_kvmstate(cpu)) {
return EINVAL;
}
kvm_arm_sync_mpstate_to_kvm(cpu);
return ret;
}
int kvm_arch_get_registers(CPUState *cs)
{
ARMCPU *cpu = ARM_CPU(cs);
CPUARMState *env = &cpu->env;
struct kvm_one_reg r;
int mode, bn;
int ret, i;
uint32_t cpsr, fpscr;
for (i = 0; i < ARRAY_SIZE(regs); i++) {
r.id = regs[i].id;
r.addr = (uintptr_t)(env) + regs[i].offset;
ret = kvm_vcpu_ioctl(cs, KVM_GET_ONE_REG, &r);
if (ret) {
return ret;
}
}
/* Special cases which aren't a single CPUARMState field */
r.id = KVM_REG_ARM | KVM_REG_SIZE_U32 |
KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(usr_regs.ARM_cpsr);
r.addr = (uintptr_t)(&cpsr);
ret = kvm_vcpu_ioctl(cs, KVM_GET_ONE_REG, &r);
if (ret) {
return ret;
}
cpsr_write(env, cpsr, 0xffffffff);
/* Make sure the current mode regs are properly set */
mode = env->uncached_cpsr & CPSR_M;
bn = bank_number(mode);
if (mode == ARM_CPU_MODE_FIQ) {
memcpy(env->regs + 8, env->fiq_regs, 5 * sizeof(uint32_t));
} else {
memcpy(env->regs + 8, env->usr_regs, 5 * sizeof(uint32_t));
}
env->regs[13] = env->banked_r13[bn];
env->regs[14] = env->banked_r14[bn];
env->spsr = env->banked_spsr[bn];
/* VFP registers */
r.id = KVM_REG_ARM | KVM_REG_SIZE_U64 | KVM_REG_ARM_VFP;
for (i = 0; i < 32; i++) {
r.addr = (uintptr_t)(&env->vfp.regs[i]);
ret = kvm_vcpu_ioctl(cs, KVM_GET_ONE_REG, &r);
if (ret) {
return ret;
}
r.id++;
}
r.id = KVM_REG_ARM | KVM_REG_SIZE_U32 | KVM_REG_ARM_VFP |
KVM_REG_ARM_VFP_FPSCR;
r.addr = (uintptr_t)&fpscr;
ret = kvm_vcpu_ioctl(cs, KVM_GET_ONE_REG, &r);
if (ret) {
return ret;
}
vfp_set_fpscr(env, fpscr);
if (!write_kvmstate_to_list(cpu)) {
return EINVAL;
}
/* Note that it's OK to have registers which aren't in CPUState,
* so we can ignore a failure return here.
*/
write_list_to_cpustate(cpu);
kvm_arm_sync_mpstate_to_qemu(cpu);
return 0;
}