QEMU-KVM: MCE: Add MCE simulation support to qemu/kvm
RFC 2/2 QEMU-KVM: MCE: Add MCE simulation support to qemu/kvm
KVM ioctls are used to initialize MCE simulation and inject MCE. The real MCE simulation is implemented in Linux kernel.
ChangeLog:
v3:
- Re-based on qemu/tcg MCE support patch
v2:
- Use new kernel MCE capability exportion interface.
Signed-off-by Huang Ying <[email protected]>
--- kvm/kernel/arch/x86/include/asm/kvm.h | 1 kvm/kernel/include/linux/kvm.h | 21 +++++++++++++++++ kvm/libkvm/libkvm-x86.c | 39 +++++++++++++++++++++++++++++++++ kvm/libkvm/libkvm.h | 4 +++ qemu-kvm-x86.c | 23 +++++++++++++++++++ qemu-kvm.c | 40 ++++++++++++++++++++++++++++++++++ qemu-kvm.h | 3 ++ target-i386/helper.c | 5 ++++ 8 files changed, 136 insertions(+)
kvm/libkvm/libkvm-x86.c +39 -0
@@ -378,6 +378,45 @@ int kvm_set_msrs(kvm_context_t kvm, intreturn r; } +int kvm_get_mce_cap_supported (kvm_context_t kvm,uint64_t *mce_cap, +int *max_banks) +{ +#ifdef KVM_CAP_MCE +int r; + + r =ioctl (kvm->fd,KVM_CHECK_EXTENSION ,KVM_CAP_MCE ); +if (r >0 ) { + *max_banks = r; +return ioctl (kvm->fd,KVM_X86_GET_MCE_CAP_SUPPORTED , mce_cap); + } +#endif +return -ENOSYS ; +} + +int kvm_setup_mce (kvm_context_t kvm,int vcpu,uint64_t *mcg_cap) +{ +#ifdef KVM_CAP_MCE +int r; + + r =ioctl (kvm->fd,KVM_CHECK_EXTENSION ,KVM_CAP_MCE ); +if (r >0 ) +return ioctl (kvm->vcpu_fd[vcpu],KVM_X86_SETUP_MCE , mcg_cap); +#endif +return -ENOSYS ; +} + +int kvm_set_mce (kvm_context_t kvm,int vcpu,struct kvm_x86_mce *m) +{ +#ifdef KVM_CAP_MCE +int r; + + r =ioctl (kvm->fd,KVM_CHECK_EXTENSION ,KVM_CAP_MCE ); +if (r >0 ) +return ioctl (kvm->vcpu_fd[vcpu],KVM_X86_SET_MCE , m); +#endif +return -ENOSYS ; +} +static void print_seg (FILE *file,const char *name,struct kvm_segment *seg) { fprintf(stderr,
kvm/libkvm/libkvm.h +4 -0
@@ -27,6 +27,10 @@ typedef struct kvm_context *kvm_context_struct kvm_msr_list *kvm_get_msr_list (kvm_context_t );int kvm_get_msrs (kvm_context_t ,int vcpu,struct kvm_msr_entry *msrs,int n);int kvm_set_msrs (kvm_context_t ,int vcpu,struct kvm_msr_entry *msrs,int n); +int kvm_get_mce_cap_supported (kvm_context_t ,uint64_t *mce_cap,int *max_banks); +int kvm_setup_mce (kvm_context_t ,int vcpu,uint64_t *mcg_cap); +struct kvm_x86_mce ; +int kvm_set_mce (kvm_context_t ,int vcpu,struct kvm_x86_mce *mce);#endif /*!
qemu-kvm-x86.c +23 -0
@@ -566,6 +566,29 @@ int kvm_arch_qemu_init_env(CPUState *cendo_cpuid_ent (&cpuid_ent[cpuid_nent++], i,0 , ©);kvm_setup_cpuid2 (kvm_context, cenv->cpu_index, cpuid_nent, cpuid_ent); + +#ifdef KVM_CAP_MCE +if (((cenv->cpuid_version >>8 )&0xF ) >=6 + && (cenv->cpuid_features&(CPUID_MCE |CPUID_MCA )) == (CPUID_MCE |CPUID_MCA ) + &&kvm_check_extension (kvm_context,KVM_CAP_MCE ) >0 ) { +uint64_t mcg_cap; +int banks; + +if (kvm_get_mce_cap_supported (kvm_context, &mcg_cap, &banks)) +perror ("kvm_get_mce_cap_supported FAILED" ); +else { +if (banks >MCE_BANKS_DEF ) + banks =MCE_BANKS_DEF ; + mcg_cap &=MCE_CAP_DEF ; + mcg_cap |= banks; +if (kvm_setup_mce (kvm_context, cenv->cpu_index, &mcg_cap)) +perror ("kvm_setup_mce FAILED" ); +else + cenv->mcg_cap = mcg_cap; + } + } +#endif +return 0 ; }
qemu-kvm.h +3 -0
@@ -233,4 +233,7 @@ static inline void cpu_synchronize_state } } +void kvm_inject_x86_mce (CPUState *cenv,int bank,uint64_t status, +uint64_t mcg_status,uint64_t addr,uint64_t misc); +#endif
target-i386/helper.c +5 -0
@@ -1442,6 +1442,11 @@ void cpu_inject_x86_mce(CPUState *cenv,unsigned bank_num = mcg_cap &0xff ;uint64_t *banks = cenv->mce_banks; +if (kvm_enabled ()) { +kvm_inject_x86_mce (cenv, bank, status, mcg_status, addr, misc); +return ; + } +if (bank >= bank_num || !(status &MCI_STATUS_VAL ))return ;
kvm/kernel/include/linux/kvm.h +21 -0
@@ -455,6 +455,9 @@ struct kvm_trace_rec {#define KVM_CAP_ASSIGN_DEV_IRQ 29/* Another bug in KVM_SET_USER_MEMORY_REGION fixed: */ #define KVM_CAP_JOIN_MEMORY_REGIONS_WORKS 30 +#ifdef __KVM_HAVE_MCE +#define KVM_CAP_MCE 31 +#endif #ifdef KVM_CAP_IRQ_ROUTING
@@ -494,6 +497,20 @@ struct kvm_irq_routing {#endif +#ifdef KVM_CAP_MCE +/* x86 MCE */ +struct kvm_x86_mce { +__u64 status; +__u64 addr; +__u64 misc; +__u64 mcg_status; +__u8 bank; +__u8 pad1; +__u16 pad2; +__u32 pad3; +}; +#endif +/* * ioctls for VM fds */
@@ -581,6 +598,10 @@ struct kvm_irq_routing {#define KVM_NMI _IO(KVMIO, 0x9a)/* Available with KVM_CAP_SET_GUEST_DEBUG */ #define KVM_SET_GUEST_DEBUG _IOW(KVMIO, 0x9b, struct kvm_guest_debug) +/* MCE for x86 */ +#define KVM_X86_SETUP_MCE _IOW(KVMIO, 0x9c, __u64) +#define KVM_X86_GET_MCE_CAP_SUPPORTED _IOR(KVMIO, 0x9d, __u64) +#define KVM_X86_SET_MCE _IOW(KVMIO, 0x9e, struct kvm_x86_mce)/* * Deprecated interfaces
kvm/kernel/arch/x86/include/asm/kvm.h +1 -0
@@ -57,6 +57,7 @@#define __KVM_HAVE_USER_NMI#define __KVM_HAVE_GUEST_DEBUG#define __KVM_HAVE_MSIX +#define __KVM_HAVE_MCE/* Architectural interrupt line count. */ #define KVM_NR_INTERRUPTS 256
qemu-kvm.c +40 -0
@@ -1425,3 +1425,43 @@ void qemu_kvm_cpu_stop(CPUState *env)if (kvm_enabled ()) env->kvm_cpu_state.stopped =1 ; } + +#ifdef TARGET_I386 +#ifdef KVM_CAP_MCE +struct kvm_x86_mce_data +{ +CPUState *env; +struct kvm_x86_mce *mce; +}; + +static void kvm_do_inject_x86_mce (void *_data) +{ +struct kvm_x86_mce_data *data = _data; +int r; + + r =kvm_set_mce (kvm_context, data->env->cpu_index, data->mce); +if (r <0 ) +perror ("kvm_set_mce FAILED" ); +} +#endif + +void kvm_inject_x86_mce (CPUState *cenv,int bank,uint64_t status, +uint64_t mcg_status,uint64_t addr,uint64_t misc) +{ +#ifdef KVM_CAP_MCE +struct kvm_x86_mce mce = { + .bank = bank, + .status = status, + .mcg_status = mcg_status, + .addr = addr, + .misc = misc, + }; +struct kvm_x86_mce_data data = { + .env = cenv, + .mce = &mce, + }; + +on_vcpu (cenv, kvm_do_inject_x86_mce, &data); +#endif +} +#endif