From patchwork Wed Sep 9 02:28:02 2009 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Huang, Ying" X-Patchwork-Id: 46300 Received: from vger.kernel.org (vger.kernel.org [209.132.176.167]) by demeter.kernel.org (8.14.2/8.14.2) with ESMTP id n892S9o8015967 for ; Wed, 9 Sep 2009 02:28:09 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751434AbZIIC2D (ORCPT ); Tue, 8 Sep 2009 22:28:03 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751417AbZIIC2D (ORCPT ); Tue, 8 Sep 2009 22:28:03 -0400 Received: from mga02.intel.com ([134.134.136.20]:52453 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751284AbZIIC2B (ORCPT ); Tue, 8 Sep 2009 22:28:01 -0400 Received: from orsmga001.jf.intel.com ([10.7.209.18]) by orsmga101.jf.intel.com with ESMTP; 08 Sep 2009 19:14:24 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.44,355,1249282800"; d="scan'208";a="548912546" Received: from yhuang-dev.sh.intel.com (HELO [10.239.13.48]) ([10.239.13.48]) by orsmga001.jf.intel.com with ESMTP; 08 Sep 2009 19:27:22 -0700 Subject: [PATCH -v2] QEMU-KVM: MCE: Relay UCR MCE to guest From: Huang Ying To: Avi Kivity Cc: Andi Kleen , Anthony Liguori , "kvm@vger.kernel.org" Date: Wed, 09 Sep 2009 10:28:02 +0800 Message-Id: <1252463282.5212.44.camel@yhuang-dev.sh.intel.com> Mime-Version: 1.0 X-Mailer: Evolution 2.26.3 Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org UCR (uncorrected recovery) MCE is supported in recent Intel CPUs, where some hardware error such as some memory error can be reported without PCC (processor context corrupted). To recover from such MCE, the corresponding memory will be unmapped, and all processes accessing the memory will be killed via SIGBUS. For KVM, if QEMU/KVM is killed, all guest processes will be killed too. So we relay SIGBUS from host OS to guest system via a UCR MCE injection. Then guest OS can isolate corresponding memory and kill necessary guest processes only. SIGBUS sent to main thread (not VCPU threads) will be broadcast to all VCPU threads as UCR MCE. v2: - Use qemu_ram_addr_from_host instead of self made one to covert from host address to guest RAM address. Thanks Anthony Liguori. Signed-off-by: Huang Ying --- cpu-common.h | 1 exec.c | 20 +++++-- qemu-kvm.c | 154 ++++++++++++++++++++++++++++++++++++++++++++++++++---- target-i386/cpu.h | 20 ++++++- 4 files changed, 178 insertions(+), 17 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html --- a/qemu-kvm.c +++ b/qemu-kvm.c @@ -27,10 +27,23 @@ #include #include #include +#include +#include #define false 0 #define true 1 +#ifndef PR_MCE_KILL +#define PR_MCE_KILL 33 +#endif + +#ifndef BUS_MCEERR_AR +#define BUS_MCEERR_AR 4 +#endif +#ifndef BUS_MCEERR_AO +#define BUS_MCEERR_AO 5 +#endif + #define EXPECTED_KVM_API_VERSION 12 #if EXPECTED_KVM_API_VERSION != KVM_API_VERSION @@ -1507,6 +1520,37 @@ static void sig_ipi_handler(int n) { } +static void sigbus_handler(int n, struct signalfd_siginfo *siginfo, void *ctx) +{ + if (siginfo->ssi_code == BUS_MCEERR_AO) { + uint64_t status; + unsigned long paddr; + CPUState *cenv; + + /* Hope we are lucky for AO MCE */ + if (do_qemu_ram_addr_from_host((void *)siginfo->ssi_addr, &paddr)) { + fprintf(stderr, "Hardware memory error for memory used by " + "QEMU itself instead of guest system!: %llx\n", + (unsigned long long)siginfo->ssi_addr); + return; + } + status = MCI_STATUS_VAL | MCI_STATUS_UC | MCI_STATUS_EN + | MCI_STATUS_MISCV | MCI_STATUS_ADDRV | MCI_STATUS_S + | 0xc0; + kvm_inject_x86_mce(first_cpu, 9, status, + MCG_STATUS_MCIP | MCG_STATUS_RIPV, paddr, + (MCM_ADDR_PHYS << 6) | 0xc); + for (cenv = first_cpu->next_cpu; cenv != NULL; cenv = cenv->next_cpu) + kvm_inject_x86_mce(cenv, 1, MCI_STATUS_VAL | MCI_STATUS_UC, + MCG_STATUS_MCIP | MCG_STATUS_RIPV, 0, 0); + return; + } else if (siginfo->ssi_code == BUS_MCEERR_AR) + fprintf(stderr, "Hardware memory error!\n"); + else + fprintf(stderr, "Internal error in QEMU!\n"); + exit(1); +} + static void on_vcpu(CPUState *env, void (*func)(void *data), void *data) { struct qemu_work_item wi; @@ -1649,29 +1693,102 @@ static void flush_queued_work(CPUState * pthread_cond_broadcast(&qemu_work_cond); } +static void kvm_on_sigbus(CPUState *env, siginfo_t *siginfo) +{ +#if defined(KVM_CAP_MCE) && defined(TARGET_I386) + struct kvm_x86_mce mce = { + .bank = 9, + }; + unsigned long paddr; + int r; + + if (env->mcg_cap && siginfo->si_addr + && (siginfo->si_code == BUS_MCEERR_AR + || siginfo->si_code == BUS_MCEERR_AO)) { + if (siginfo->si_code == BUS_MCEERR_AR) { + /* Fake an Intel architectural Data Load SRAR UCR */ + mce.status = MCI_STATUS_VAL | MCI_STATUS_UC | MCI_STATUS_EN + | MCI_STATUS_MISCV | MCI_STATUS_ADDRV | MCI_STATUS_S + | MCI_STATUS_AR | 0x134; + mce.misc = (MCM_ADDR_PHYS << 6) | 0xc; + mce.mcg_status = MCG_STATUS_MCIP | MCG_STATUS_EIPV; + } else { + /* Fake an Intel architectural Memory scrubbing UCR */ + mce.status = MCI_STATUS_VAL | MCI_STATUS_UC | MCI_STATUS_EN + | MCI_STATUS_MISCV | MCI_STATUS_ADDRV | MCI_STATUS_S + | 0xc0; + mce.misc = (MCM_ADDR_PHYS << 6) | 0xc; + mce.mcg_status = MCG_STATUS_MCIP | MCG_STATUS_RIPV; + } + if (do_qemu_ram_addr_from_host((void *)siginfo->si_addr, &paddr)) { + fprintf(stderr, "Hardware memory error for memory used by " + "QEMU itself instaed of guest system!\n"); + /* Hope we are lucky for AO MCE */ + if (siginfo->si_code == BUS_MCEERR_AO) + return; + else + exit(1); + } + mce.addr = paddr; + r = kvm_set_mce(env->kvm_cpu_state.vcpu_ctx, &mce); + if (r < 0) { + fprintf(stderr, "kvm_set_mce: %s\n", strerror(errno)); + exit(1); + } + } else +#endif + { + if (siginfo->si_code == BUS_MCEERR_AO) + return; + if (siginfo->si_code == BUS_MCEERR_AR) + fprintf(stderr, "Hardware memory error!\n"); + else + fprintf(stderr, "Internal error in QEMU!\n"); + exit(1); + } +} + static void kvm_main_loop_wait(CPUState *env, int timeout) { struct timespec ts; int r, e; siginfo_t siginfo; sigset_t waitset; - - pthread_mutex_unlock(&qemu_mutex); + sigset_t chkset; ts.tv_sec = timeout / 1000; ts.tv_nsec = (timeout % 1000) * 1000000; sigemptyset(&waitset); sigaddset(&waitset, SIG_IPI); + sigaddset(&waitset, SIGBUS); - r = sigtimedwait(&waitset, &siginfo, &ts); - e = errno; + do { + pthread_mutex_unlock(&qemu_mutex); - pthread_mutex_lock(&qemu_mutex); + r = sigtimedwait(&waitset, &siginfo, &ts); + e = errno; - if (r == -1 && !(e == EAGAIN || e == EINTR)) { - printf("sigtimedwait: %s\n", strerror(e)); - exit(1); - } + pthread_mutex_lock(&qemu_mutex); + + if (r == -1 && !(e == EAGAIN || e == EINTR)) { + printf("sigtimedwait: %s\n", strerror(e)); + exit(1); + } + + switch (r) { + case SIGBUS: + kvm_on_sigbus(env, &siginfo); + break; + default: + break; + } + + r = sigpending(&chkset); + if (r == -1) { + printf("sigpending: %s\n", strerror(e)); + exit(1); + } + } while (sigismember(&chkset, SIG_IPI) || sigismember(&chkset, SIGBUS)); cpu_single_env = env; flush_queued_work(env); @@ -1752,6 +1869,7 @@ static void setup_kernel_sigmask(CPUStat sigprocmask(SIG_BLOCK, NULL, &set); sigdelset(&set, SIG_IPI); + sigdelset(&set, SIGBUS); kvm_set_signal_mask(env->kvm_cpu_state.vcpu_ctx, &set); } @@ -1877,12 +1995,20 @@ void kvm_hpet_enable_kpit(void) int kvm_init_ap(void) { + struct sigaction action; + #ifdef TARGET_I386 kvm_tpr_opt_setup(); #endif qemu_add_vm_change_state_handler(kvm_vm_state_change_handler, NULL); signal(SIG_IPI, sig_ipi_handler); + + memset(&action, 0, sizeof(action)); + action.sa_flags = SA_SIGINFO; + action.sa_sigaction = (void (*)(int, siginfo_t*, void*))sigbus_handler; + sigaction(SIGBUS, &action, NULL); + prctl(PR_MCE_KILL, 1, 1); return 0; } @@ -1943,7 +2069,10 @@ static void sigfd_handler(void *opaque) } sigaction(info.ssi_signo, NULL, &action); - if (action.sa_handler) + if ((action.sa_flags & SA_SIGINFO) && action.sa_sigaction) + action.sa_sigaction(info.ssi_signo, + (siginfo_t *)&info, NULL); + else if (action.sa_handler) action.sa_handler(info.ssi_signo); } @@ -1993,6 +2122,7 @@ int kvm_main_loop(void) sigemptyset(&mask); sigaddset(&mask, SIGIO); sigaddset(&mask, SIGALRM); + sigaddset(&mask, SIGBUS); sigprocmask(SIG_BLOCK, &mask, NULL); sigfd = qemu_signalfd(&mask); @@ -2518,6 +2648,10 @@ void kvm_inject_x86_mce(CPUState *cenv, .mce = &mce, }; + if (!cenv->mcg_cap) { + fprintf(stderr, "MCE support is not enabled!\n"); + return; + } on_vcpu(cenv, kvm_do_inject_x86_mce, &data); #endif } --- a/target-i386/cpu.h +++ b/target-i386/cpu.h @@ -250,16 +250,32 @@ #define PG_ERROR_RSVD_MASK 0x08 #define PG_ERROR_I_D_MASK 0x10 -#define MCG_CTL_P (1UL<<8) /* MCG_CAP register available */ +#define MCG_CTL_P (1ULL<<8) /* MCG_CAP register available */ +#define MCG_SER_P (1ULL<<24) /* MCA recovery/new status bits */ -#define MCE_CAP_DEF MCG_CTL_P +#define MCE_CAP_DEF (MCG_CTL_P|MCG_SER_P) #define MCE_BANKS_DEF 10 +#define MCG_STATUS_RIPV (1ULL<<0) /* restart ip valid */ +#define MCG_STATUS_EIPV (1ULL<<1) /* ip points to correct instruction */ #define MCG_STATUS_MCIP (1ULL<<2) /* machine check in progress */ #define MCI_STATUS_VAL (1ULL<<63) /* valid error */ #define MCI_STATUS_OVER (1ULL<<62) /* previous errors lost */ #define MCI_STATUS_UC (1ULL<<61) /* uncorrected error */ +#define MCI_STATUS_EN (1ULL<<60) /* error enabled */ +#define MCI_STATUS_MISCV (1ULL<<59) /* misc error reg. valid */ +#define MCI_STATUS_ADDRV (1ULL<<58) /* addr reg. valid */ +#define MCI_STATUS_PCC (1ULL<<57) /* processor context corrupt */ +#define MCI_STATUS_S (1ULL<<56) /* Signaled machine check */ +#define MCI_STATUS_AR (1ULL<<55) /* Action required */ + +/* MISC register defines */ +#define MCM_ADDR_SEGOFF 0 /* segment offset */ +#define MCM_ADDR_LINEAR 1 /* linear address */ +#define MCM_ADDR_PHYS 2 /* physical address */ +#define MCM_ADDR_MEM 3 /* memory address */ +#define MCM_ADDR_GENERIC 7 /* generic */ #define MSR_IA32_TSC 0x10 #define MSR_IA32_APICBASE 0x1b --- a/cpu-common.h +++ b/cpu-common.h @@ -34,6 +34,7 @@ void qemu_ram_free(ram_addr_t addr); /* This should only be used for ram local to a device. */ void *qemu_get_ram_ptr(ram_addr_t addr); /* This should not be used by devices. */ +int do_qemu_ram_addr_from_host(void *ptr, ram_addr_t *ram_addr); ram_addr_t qemu_ram_addr_from_host(void *ptr); int cpu_register_io_memory(CPUReadMemoryFunc * const *mem_read, --- a/exec.c +++ b/exec.c @@ -2589,9 +2589,7 @@ void *qemu_get_ram_ptr(ram_addr_t addr) return block->host + (addr - block->offset); } -/* Some of the softmmu routines need to translate from a host pointer - (typically a TLB entry) back to a ram offset. */ -ram_addr_t qemu_ram_addr_from_host(void *ptr) +int do_qemu_ram_addr_from_host(void *ptr, ram_addr_t *ram_addr) { RAMBlock *prev; RAMBlock **prevp; @@ -2608,11 +2606,23 @@ ram_addr_t qemu_ram_addr_from_host(void prev = block; block = block->next; } - if (!block) { + if (!block) + return -1; + *ram_addr = block->offset + (host - block->host); + return 0; +} + +/* Some of the softmmu routines need to translate from a host pointer + (typically a TLB entry) back to a ram offset. */ +ram_addr_t qemu_ram_addr_from_host(void *ptr) +{ + ram_addr_t ram_addr; + + if (do_qemu_ram_addr_from_host(ptr, &ram_addr)) { fprintf(stderr, "Bad ram pointer %p\n", ptr); abort(); } - return block->offset + (host - block->host); + return ram_addr; } static uint32_t unassigned_mem_readb(void *opaque, target_phys_addr_t addr)