2017-06-06 20:21:31

by Bandan Das

[permalink] [raw]
Subject: Re: [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor

Jintack Lim <[email protected]> writes:

> Emulate taking an exception to the guest hypervisor running in the
> virtual EL2 as described in ARM ARM AArch64.TakeException().

ARM newbie here, I keep thinking of ARM ARM as a typo ;)
...
> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
> +{
> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
> + __func__);
> + return -EINVAL;
> +}
> +
> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
> +{
> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
> + __func__);
> + return -EINVAL;
> +}
> +

I see these function stubs for aarch32 in the patches. I don't see how they
can actually be called though. Is this because eventually, there will be
a virtual el2 mode for aarch32 ?

Bandan

> static inline void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu) { };
> static inline void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu) { };
> static inline void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt) { };
> diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
> index 8892c82..0987ee4 100644
> --- a/arch/arm64/include/asm/kvm_emulate.h
> +++ b/arch/arm64/include/asm/kvm_emulate.h
> @@ -42,6 +42,25 @@
> void kvm_inject_dabt(struct kvm_vcpu *vcpu, unsigned long addr);
> void kvm_inject_pabt(struct kvm_vcpu *vcpu, unsigned long addr);
>
> +#ifdef CONFIG_KVM_ARM_NESTED_HYP
> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2);
> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu);
> +#else
> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
> +{
> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
> + __func__);
> + return -EINVAL;
> +}
> +
> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
> +{
> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
> + __func__);
> + return -EINVAL;
> +}
> +#endif
> +
> void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu);
> void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu);
> void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt);
> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
> index 7811d27..b342bdd 100644
> --- a/arch/arm64/kvm/Makefile
> +++ b/arch/arm64/kvm/Makefile
> @@ -34,3 +34,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/irqchip.o
> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
> kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
> +
> +kvm-$(CONFIG_KVM_ARM_NESTED_HYP) += emulate-nested.o
> diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c
> new file mode 100644
> index 0000000..59d147f
> --- /dev/null
> +++ b/arch/arm64/kvm/emulate-nested.c
> @@ -0,0 +1,66 @@
> +/*
> + * Copyright (C) 2016 - Columbia University
> + * Author: Jintack Lim <[email protected]>
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/kvm.h>
> +#include <linux/kvm_host.h>
> +
> +#include <asm/kvm_emulate.h>
> +
> +#include "trace.h"
> +
> +#define EL2_EXCEPT_SYNC_OFFSET 0x400
> +#define EL2_EXCEPT_ASYNC_OFFSET 0x480
> +
> +
> +/*
> + * Emulate taking an exception. See ARM ARM J8.1.2 AArch64.TakeException()
> + */
> +static int kvm_inject_nested(struct kvm_vcpu *vcpu, u64 esr_el2,
> + int exception_offset)
> +{
> + int ret = 1;
> + kvm_cpu_context_t *ctxt = &vcpu->arch.ctxt;
> +
> + /* We don't inject an exception recursively to virtual EL2 */
> + if (vcpu_mode_el2(vcpu))
> + BUG();
> +
> + ctxt->el2_regs[SPSR_EL2] = *vcpu_cpsr(vcpu);
> + ctxt->el2_regs[ELR_EL2] = *vcpu_pc(vcpu);
> + ctxt->el2_regs[ESR_EL2] = esr_el2;
> +
> + /* On an exception, PSTATE.SP = 1 */
> + *vcpu_cpsr(vcpu) = PSR_MODE_EL2h;
> + *vcpu_cpsr(vcpu) |= (PSR_A_BIT | PSR_F_BIT | PSR_I_BIT | PSR_D_BIT);
> + *vcpu_pc(vcpu) = ctxt->el2_regs[VBAR_EL2] + exception_offset;
> +
> + trace_kvm_inject_nested_exception(vcpu, esr_el2, *vcpu_pc(vcpu));
> +
> + return ret;
> +}
> +
> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
> +{
> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_SYNC_OFFSET);
> +}
> +
> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
> +{
> + u64 esr_el2 = kvm_vcpu_get_hsr(vcpu);
> + /* We supports only IRQ and FIQ, so the esr_el2 is not updated. */
> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_ASYNC_OFFSET);
> +}
> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
> index 7fb0008..7c86cfb 100644
> --- a/arch/arm64/kvm/trace.h
> +++ b/arch/arm64/kvm/trace.h
> @@ -167,6 +167,26 @@
> );
>
>
> +TRACE_EVENT(kvm_inject_nested_exception,
> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long esr_el2,
> + unsigned long pc),
> + TP_ARGS(vcpu, esr_el2, pc),
> +
> + TP_STRUCT__entry(
> + __field(struct kvm_vcpu *, vcpu)
> + __field(unsigned long, esr_el2)
> + __field(unsigned long, pc)
> + ),
> +
> + TP_fast_assign(
> + __entry->vcpu = vcpu;
> + __entry->esr_el2 = esr_el2;
> + __entry->pc = pc;
> + ),
> +
> + TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
> + __entry->vcpu, __entry->esr_el2, __entry->pc)
> +);
> #endif /* _TRACE_ARM64_KVM_H */
>
> #undef TRACE_INCLUDE_PATH


2017-06-06 20:39:01

by Jintack Lim

[permalink] [raw]
Subject: Re: [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor

Hi Bandan,

On Tue, Jun 6, 2017 at 4:21 PM, Bandan Das <[email protected]> wrote:
> Jintack Lim <[email protected]> writes:
>
>> Emulate taking an exception to the guest hypervisor running in the
>> virtual EL2 as described in ARM ARM AArch64.TakeException().
>
> ARM newbie here, I keep thinking of ARM ARM as a typo ;)

ARM ARM means ARM Architecture Reference Manual :)

> ...
>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>> +{
>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>> + __func__);
>> + return -EINVAL;
>> +}
>> +
>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>> +{
>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>> + __func__);
>> + return -EINVAL;
>> +}
>> +
>
> I see these function stubs for aarch32 in the patches. I don't see how they
> can actually be called though. Is this because eventually, there will be
> a virtual el2 mode for aarch32 ?

Current RFC doesn't support nested virtualization on 32bit arm
architecture and those functions will be never called. Those functions
are there for the compilation.

Thanks,
Jintack

>
> Bandan
>
>> static inline void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu) { };
>> static inline void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu) { };
>> static inline void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt) { };
>> diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
>> index 8892c82..0987ee4 100644
>> --- a/arch/arm64/include/asm/kvm_emulate.h
>> +++ b/arch/arm64/include/asm/kvm_emulate.h
>> @@ -42,6 +42,25 @@
>> void kvm_inject_dabt(struct kvm_vcpu *vcpu, unsigned long addr);
>> void kvm_inject_pabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>
>> +#ifdef CONFIG_KVM_ARM_NESTED_HYP
>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2);
>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu);
>> +#else
>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>> +{
>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>> + __func__);
>> + return -EINVAL;
>> +}
>> +
>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>> +{
>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>> + __func__);
>> + return -EINVAL;
>> +}
>> +#endif
>> +
>> void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu);
>> void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu);
>> void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt);
>> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
>> index 7811d27..b342bdd 100644
>> --- a/arch/arm64/kvm/Makefile
>> +++ b/arch/arm64/kvm/Makefile
>> @@ -34,3 +34,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/irqchip.o
>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
>> kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
>> +
>> +kvm-$(CONFIG_KVM_ARM_NESTED_HYP) += emulate-nested.o
>> diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c
>> new file mode 100644
>> index 0000000..59d147f
>> --- /dev/null
>> +++ b/arch/arm64/kvm/emulate-nested.c
>> @@ -0,0 +1,66 @@
>> +/*
>> + * Copyright (C) 2016 - Columbia University
>> + * Author: Jintack Lim <[email protected]>
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include <linux/kvm.h>
>> +#include <linux/kvm_host.h>
>> +
>> +#include <asm/kvm_emulate.h>
>> +
>> +#include "trace.h"
>> +
>> +#define EL2_EXCEPT_SYNC_OFFSET 0x400
>> +#define EL2_EXCEPT_ASYNC_OFFSET 0x480
>> +
>> +
>> +/*
>> + * Emulate taking an exception. See ARM ARM J8.1.2 AArch64.TakeException()
>> + */
>> +static int kvm_inject_nested(struct kvm_vcpu *vcpu, u64 esr_el2,
>> + int exception_offset)
>> +{
>> + int ret = 1;
>> + kvm_cpu_context_t *ctxt = &vcpu->arch.ctxt;
>> +
>> + /* We don't inject an exception recursively to virtual EL2 */
>> + if (vcpu_mode_el2(vcpu))
>> + BUG();
>> +
>> + ctxt->el2_regs[SPSR_EL2] = *vcpu_cpsr(vcpu);
>> + ctxt->el2_regs[ELR_EL2] = *vcpu_pc(vcpu);
>> + ctxt->el2_regs[ESR_EL2] = esr_el2;
>> +
>> + /* On an exception, PSTATE.SP = 1 */
>> + *vcpu_cpsr(vcpu) = PSR_MODE_EL2h;
>> + *vcpu_cpsr(vcpu) |= (PSR_A_BIT | PSR_F_BIT | PSR_I_BIT | PSR_D_BIT);
>> + *vcpu_pc(vcpu) = ctxt->el2_regs[VBAR_EL2] + exception_offset;
>> +
>> + trace_kvm_inject_nested_exception(vcpu, esr_el2, *vcpu_pc(vcpu));
>> +
>> + return ret;
>> +}
>> +
>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>> +{
>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_SYNC_OFFSET);
>> +}
>> +
>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>> +{
>> + u64 esr_el2 = kvm_vcpu_get_hsr(vcpu);
>> + /* We supports only IRQ and FIQ, so the esr_el2 is not updated. */
>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_ASYNC_OFFSET);
>> +}
>> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
>> index 7fb0008..7c86cfb 100644
>> --- a/arch/arm64/kvm/trace.h
>> +++ b/arch/arm64/kvm/trace.h
>> @@ -167,6 +167,26 @@
>> );
>>
>>
>> +TRACE_EVENT(kvm_inject_nested_exception,
>> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long esr_el2,
>> + unsigned long pc),
>> + TP_ARGS(vcpu, esr_el2, pc),
>> +
>> + TP_STRUCT__entry(
>> + __field(struct kvm_vcpu *, vcpu)
>> + __field(unsigned long, esr_el2)
>> + __field(unsigned long, pc)
>> + ),
>> +
>> + TP_fast_assign(
>> + __entry->vcpu = vcpu;
>> + __entry->esr_el2 = esr_el2;
>> + __entry->pc = pc;
>> + ),
>> +
>> + TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
>> + __entry->vcpu, __entry->esr_el2, __entry->pc)
>> +);
>> #endif /* _TRACE_ARM64_KVM_H */
>>
>> #undef TRACE_INCLUDE_PATH
> _______________________________________________
> kvmarm mailing list
> [email protected]
> https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

2017-06-06 22:07:12

by Bandan Das

[permalink] [raw]
Subject: Re: [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor

Hi Jintack,

Jintack Lim <[email protected]> writes:

> Hi Bandan,
>
> On Tue, Jun 6, 2017 at 4:21 PM, Bandan Das <[email protected]> wrote:
>> Jintack Lim <[email protected]> writes:
>>
>>> Emulate taking an exception to the guest hypervisor running in the
>>> virtual EL2 as described in ARM ARM AArch64.TakeException().
>>
>> ARM newbie here, I keep thinking of ARM ARM as a typo ;)
>
> ARM ARM means ARM Architecture Reference Manual :)
>
>> ...
>>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>> +{
>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>> + __func__);
>>> + return -EINVAL;
>>> +}
>>> +
>>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>> +{
>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>> + __func__);
>>> + return -EINVAL;
>>> +}
>>> +
>>
>> I see these function stubs for aarch32 in the patches. I don't see how they
>> can actually be called though. Is this because eventually, there will be
>> a virtual el2 mode for aarch32 ?
>
> Current RFC doesn't support nested virtualization on 32bit arm
> architecture and those functions will be never called. Those functions
> are there for the compilation.

Do you mean that compilation will fail ? It seems these functions are
defined separately in 32/64 bit specific header files. Or is it that
64 bit compilation also depends on the 32 bit header file ?

Bandan

> Thanks,
> Jintack
>
>>
>> Bandan
>>
>>> static inline void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu) { };
>>> static inline void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu) { };
>>> static inline void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt) { };
>>> diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
>>> index 8892c82..0987ee4 100644
>>> --- a/arch/arm64/include/asm/kvm_emulate.h
>>> +++ b/arch/arm64/include/asm/kvm_emulate.h
>>> @@ -42,6 +42,25 @@
>>> void kvm_inject_dabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>> void kvm_inject_pabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>>
>>> +#ifdef CONFIG_KVM_ARM_NESTED_HYP
>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2);
>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu);
>>> +#else
>>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>> +{
>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>> + __func__);
>>> + return -EINVAL;
>>> +}
>>> +
>>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>> +{
>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>> + __func__);
>>> + return -EINVAL;
>>> +}
>>> +#endif
>>> +
>>> void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu);
>>> void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu);
>>> void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt);
>>> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
>>> index 7811d27..b342bdd 100644
>>> --- a/arch/arm64/kvm/Makefile
>>> +++ b/arch/arm64/kvm/Makefile
>>> @@ -34,3 +34,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/irqchip.o
>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
>>> kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
>>> +
>>> +kvm-$(CONFIG_KVM_ARM_NESTED_HYP) += emulate-nested.o
>>> diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c
>>> new file mode 100644
>>> index 0000000..59d147f
>>> --- /dev/null
>>> +++ b/arch/arm64/kvm/emulate-nested.c
>>> @@ -0,0 +1,66 @@
>>> +/*
>>> + * Copyright (C) 2016 - Columbia University
>>> + * Author: Jintack Lim <[email protected]>
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License version 2 as
>>> + * published by the Free Software Foundation.
>>> + *
>>> + * This program is distributed in the hope that it will be useful,
>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>>> + * GNU General Public License for more details.
>>> + *
>>> + * You should have received a copy of the GNU General Public License
>>> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
>>> + */
>>> +
>>> +#include <linux/kvm.h>
>>> +#include <linux/kvm_host.h>
>>> +
>>> +#include <asm/kvm_emulate.h>
>>> +
>>> +#include "trace.h"
>>> +
>>> +#define EL2_EXCEPT_SYNC_OFFSET 0x400
>>> +#define EL2_EXCEPT_ASYNC_OFFSET 0x480
>>> +
>>> +
>>> +/*
>>> + * Emulate taking an exception. See ARM ARM J8.1.2 AArch64.TakeException()
>>> + */
>>> +static int kvm_inject_nested(struct kvm_vcpu *vcpu, u64 esr_el2,
>>> + int exception_offset)
>>> +{
>>> + int ret = 1;
>>> + kvm_cpu_context_t *ctxt = &vcpu->arch.ctxt;
>>> +
>>> + /* We don't inject an exception recursively to virtual EL2 */
>>> + if (vcpu_mode_el2(vcpu))
>>> + BUG();
>>> +
>>> + ctxt->el2_regs[SPSR_EL2] = *vcpu_cpsr(vcpu);
>>> + ctxt->el2_regs[ELR_EL2] = *vcpu_pc(vcpu);
>>> + ctxt->el2_regs[ESR_EL2] = esr_el2;
>>> +
>>> + /* On an exception, PSTATE.SP = 1 */
>>> + *vcpu_cpsr(vcpu) = PSR_MODE_EL2h;
>>> + *vcpu_cpsr(vcpu) |= (PSR_A_BIT | PSR_F_BIT | PSR_I_BIT | PSR_D_BIT);
>>> + *vcpu_pc(vcpu) = ctxt->el2_regs[VBAR_EL2] + exception_offset;
>>> +
>>> + trace_kvm_inject_nested_exception(vcpu, esr_el2, *vcpu_pc(vcpu));
>>> +
>>> + return ret;
>>> +}
>>> +
>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>> +{
>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_SYNC_OFFSET);
>>> +}
>>> +
>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>> +{
>>> + u64 esr_el2 = kvm_vcpu_get_hsr(vcpu);
>>> + /* We supports only IRQ and FIQ, so the esr_el2 is not updated. */
>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_ASYNC_OFFSET);
>>> +}
>>> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
>>> index 7fb0008..7c86cfb 100644
>>> --- a/arch/arm64/kvm/trace.h
>>> +++ b/arch/arm64/kvm/trace.h
>>> @@ -167,6 +167,26 @@
>>> );
>>>
>>>
>>> +TRACE_EVENT(kvm_inject_nested_exception,
>>> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long esr_el2,
>>> + unsigned long pc),
>>> + TP_ARGS(vcpu, esr_el2, pc),
>>> +
>>> + TP_STRUCT__entry(
>>> + __field(struct kvm_vcpu *, vcpu)
>>> + __field(unsigned long, esr_el2)
>>> + __field(unsigned long, pc)
>>> + ),
>>> +
>>> + TP_fast_assign(
>>> + __entry->vcpu = vcpu;
>>> + __entry->esr_el2 = esr_el2;
>>> + __entry->pc = pc;
>>> + ),
>>> +
>>> + TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
>>> + __entry->vcpu, __entry->esr_el2, __entry->pc)
>>> +);
>>> #endif /* _TRACE_ARM64_KVM_H */
>>>
>>> #undef TRACE_INCLUDE_PATH
>> _______________________________________________
>> kvmarm mailing list
>> [email protected]
>> https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

2017-06-06 23:16:22

by Jintack Lim

[permalink] [raw]
Subject: Re: [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor

On Tue, Jun 6, 2017 at 6:07 PM, Bandan Das <[email protected]> wrote:
> Hi Jintack,
>
> Jintack Lim <[email protected]> writes:
>
>> Hi Bandan,
>>
>> On Tue, Jun 6, 2017 at 4:21 PM, Bandan Das <[email protected]> wrote:
>>> Jintack Lim <[email protected]> writes:
>>>
>>>> Emulate taking an exception to the guest hypervisor running in the
>>>> virtual EL2 as described in ARM ARM AArch64.TakeException().
>>>
>>> ARM newbie here, I keep thinking of ARM ARM as a typo ;)
>>
>> ARM ARM means ARM Architecture Reference Manual :)
>>
>>> ...
>>>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>>> +{
>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>> + __func__);
>>>> + return -EINVAL;
>>>> +}
>>>> +
>>>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>>> +{
>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>> + __func__);
>>>> + return -EINVAL;
>>>> +}
>>>> +
>>>
>>> I see these function stubs for aarch32 in the patches. I don't see how they
>>> can actually be called though. Is this because eventually, there will be
>>> a virtual el2 mode for aarch32 ?
>>
>> Current RFC doesn't support nested virtualization on 32bit arm
>> architecture and those functions will be never called. Those functions
>> are there for the compilation.
>
> Do you mean that compilation will fail ?

Compilation on 32bit arm architecture will fail without them.

> It seems these functions are
> defined separately in 32/64 bit specific header files. Or is it that
> 64 bit compilation also depends on the 32 bit header file ?

It's only for 32bit architecture. For example, kvm_inject_nested_irq()
is called in virt/kvm/arm/vgic/vgic.c which is shared between 32 and
64 bit.

>
> Bandan
>
>> Thanks,
>> Jintack
>>
>>>
>>> Bandan
>>>
>>>> static inline void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu) { };
>>>> static inline void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu) { };
>>>> static inline void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt) { };
>>>> diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
>>>> index 8892c82..0987ee4 100644
>>>> --- a/arch/arm64/include/asm/kvm_emulate.h
>>>> +++ b/arch/arm64/include/asm/kvm_emulate.h
>>>> @@ -42,6 +42,25 @@
>>>> void kvm_inject_dabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>>> void kvm_inject_pabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>>>
>>>> +#ifdef CONFIG_KVM_ARM_NESTED_HYP
>>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2);
>>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu);
>>>> +#else
>>>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>>> +{
>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>> + __func__);
>>>> + return -EINVAL;
>>>> +}
>>>> +
>>>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>>> +{
>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>> + __func__);
>>>> + return -EINVAL;
>>>> +}
>>>> +#endif
>>>> +
>>>> void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu);
>>>> void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu);
>>>> void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt);
>>>> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
>>>> index 7811d27..b342bdd 100644
>>>> --- a/arch/arm64/kvm/Makefile
>>>> +++ b/arch/arm64/kvm/Makefile
>>>> @@ -34,3 +34,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
>>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/irqchip.o
>>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
>>>> kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
>>>> +
>>>> +kvm-$(CONFIG_KVM_ARM_NESTED_HYP) += emulate-nested.o
>>>> diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c
>>>> new file mode 100644
>>>> index 0000000..59d147f
>>>> --- /dev/null
>>>> +++ b/arch/arm64/kvm/emulate-nested.c
>>>> @@ -0,0 +1,66 @@
>>>> +/*
>>>> + * Copyright (C) 2016 - Columbia University
>>>> + * Author: Jintack Lim <[email protected]>
>>>> + *
>>>> + * This program is free software; you can redistribute it and/or modify
>>>> + * it under the terms of the GNU General Public License version 2 as
>>>> + * published by the Free Software Foundation.
>>>> + *
>>>> + * This program is distributed in the hope that it will be useful,
>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>>>> + * GNU General Public License for more details.
>>>> + *
>>>> + * You should have received a copy of the GNU General Public License
>>>> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
>>>> + */
>>>> +
>>>> +#include <linux/kvm.h>
>>>> +#include <linux/kvm_host.h>
>>>> +
>>>> +#include <asm/kvm_emulate.h>
>>>> +
>>>> +#include "trace.h"
>>>> +
>>>> +#define EL2_EXCEPT_SYNC_OFFSET 0x400
>>>> +#define EL2_EXCEPT_ASYNC_OFFSET 0x480
>>>> +
>>>> +
>>>> +/*
>>>> + * Emulate taking an exception. See ARM ARM J8.1.2 AArch64.TakeException()
>>>> + */
>>>> +static int kvm_inject_nested(struct kvm_vcpu *vcpu, u64 esr_el2,
>>>> + int exception_offset)
>>>> +{
>>>> + int ret = 1;
>>>> + kvm_cpu_context_t *ctxt = &vcpu->arch.ctxt;
>>>> +
>>>> + /* We don't inject an exception recursively to virtual EL2 */
>>>> + if (vcpu_mode_el2(vcpu))
>>>> + BUG();
>>>> +
>>>> + ctxt->el2_regs[SPSR_EL2] = *vcpu_cpsr(vcpu);
>>>> + ctxt->el2_regs[ELR_EL2] = *vcpu_pc(vcpu);
>>>> + ctxt->el2_regs[ESR_EL2] = esr_el2;
>>>> +
>>>> + /* On an exception, PSTATE.SP = 1 */
>>>> + *vcpu_cpsr(vcpu) = PSR_MODE_EL2h;
>>>> + *vcpu_cpsr(vcpu) |= (PSR_A_BIT | PSR_F_BIT | PSR_I_BIT | PSR_D_BIT);
>>>> + *vcpu_pc(vcpu) = ctxt->el2_regs[VBAR_EL2] + exception_offset;
>>>> +
>>>> + trace_kvm_inject_nested_exception(vcpu, esr_el2, *vcpu_pc(vcpu));
>>>> +
>>>> + return ret;
>>>> +}
>>>> +
>>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>>> +{
>>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_SYNC_OFFSET);
>>>> +}
>>>> +
>>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>>> +{
>>>> + u64 esr_el2 = kvm_vcpu_get_hsr(vcpu);
>>>> + /* We supports only IRQ and FIQ, so the esr_el2 is not updated. */
>>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_ASYNC_OFFSET);
>>>> +}
>>>> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
>>>> index 7fb0008..7c86cfb 100644
>>>> --- a/arch/arm64/kvm/trace.h
>>>> +++ b/arch/arm64/kvm/trace.h
>>>> @@ -167,6 +167,26 @@
>>>> );
>>>>
>>>>
>>>> +TRACE_EVENT(kvm_inject_nested_exception,
>>>> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long esr_el2,
>>>> + unsigned long pc),
>>>> + TP_ARGS(vcpu, esr_el2, pc),
>>>> +
>>>> + TP_STRUCT__entry(
>>>> + __field(struct kvm_vcpu *, vcpu)
>>>> + __field(unsigned long, esr_el2)
>>>> + __field(unsigned long, pc)
>>>> + ),
>>>> +
>>>> + TP_fast_assign(
>>>> + __entry->vcpu = vcpu;
>>>> + __entry->esr_el2 = esr_el2;
>>>> + __entry->pc = pc;
>>>> + ),
>>>> +
>>>> + TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
>>>> + __entry->vcpu, __entry->esr_el2, __entry->pc)
>>>> +);
>>>> #endif /* _TRACE_ARM64_KVM_H */
>>>>
>>>> #undef TRACE_INCLUDE_PATH
>>> _______________________________________________
>>> kvmarm mailing list
>>> [email protected]
>>> https://lists.cs.columbia.edu/mailman/listinfo/kvmarm

2017-06-07 17:21:07

by Bandan Das

[permalink] [raw]
Subject: Re: [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor

Jintack Lim <[email protected]> writes:

> Compilation on 32bit arm architecture will fail without them.
...
>> It seems these functions are
>> defined separately in 32/64 bit specific header files. Or is it that
>> 64 bit compilation also depends on the 32 bit header file ?
>
> It's only for 32bit architecture. For example, kvm_inject_nested_irq()
> is called in virt/kvm/arm/vgic/vgic.c which is shared between 32 and
> 64 bit.

Ah, that's the catch! Thanks for clearing this up!

>>
>> Bandan
>>
>>> Thanks,
>>> Jintack
>>>
>>>>
>>>> Bandan
>>>>
>>>>> static inline void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu) { };
>>>>> static inline void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu) { };
>>>>> static inline void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt) { };
>>>>> diff --git a/arch/arm64/include/asm/kvm_emulate.h b/arch/arm64/include/asm/kvm_emulate.h
>>>>> index 8892c82..0987ee4 100644
>>>>> --- a/arch/arm64/include/asm/kvm_emulate.h
>>>>> +++ b/arch/arm64/include/asm/kvm_emulate.h
>>>>> @@ -42,6 +42,25 @@
>>>>> void kvm_inject_dabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>>>> void kvm_inject_pabt(struct kvm_vcpu *vcpu, unsigned long addr);
>>>>>
>>>>> +#ifdef CONFIG_KVM_ARM_NESTED_HYP
>>>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2);
>>>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu);
>>>>> +#else
>>>>> +static inline int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>>>> +{
>>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>>> + __func__);
>>>>> + return -EINVAL;
>>>>> +}
>>>>> +
>>>>> +static inline int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>>>> +{
>>>>> + kvm_err("Unexpected call to %s for the non-nesting configuration\n",
>>>>> + __func__);
>>>>> + return -EINVAL;
>>>>> +}
>>>>> +#endif
>>>>> +
>>>>> void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu);
>>>>> void kvm_arm_restore_shadow_state(struct kvm_vcpu *vcpu);
>>>>> void kvm_arm_init_cpu_context(kvm_cpu_context_t *cpu_ctxt);
>>>>> diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile
>>>>> index 7811d27..b342bdd 100644
>>>>> --- a/arch/arm64/kvm/Makefile
>>>>> +++ b/arch/arm64/kvm/Makefile
>>>>> @@ -34,3 +34,5 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o
>>>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/irqchip.o
>>>>> kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o
>>>>> kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o
>>>>> +
>>>>> +kvm-$(CONFIG_KVM_ARM_NESTED_HYP) += emulate-nested.o
>>>>> diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c
>>>>> new file mode 100644
>>>>> index 0000000..59d147f
>>>>> --- /dev/null
>>>>> +++ b/arch/arm64/kvm/emulate-nested.c
>>>>> @@ -0,0 +1,66 @@
>>>>> +/*
>>>>> + * Copyright (C) 2016 - Columbia University
>>>>> + * Author: Jintack Lim <[email protected]>
>>>>> + *
>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>> + * published by the Free Software Foundation.
>>>>> + *
>>>>> + * This program is distributed in the hope that it will be useful,
>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>>>>> + * GNU General Public License for more details.
>>>>> + *
>>>>> + * You should have received a copy of the GNU General Public License
>>>>> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
>>>>> + */
>>>>> +
>>>>> +#include <linux/kvm.h>
>>>>> +#include <linux/kvm_host.h>
>>>>> +
>>>>> +#include <asm/kvm_emulate.h>
>>>>> +
>>>>> +#include "trace.h"
>>>>> +
>>>>> +#define EL2_EXCEPT_SYNC_OFFSET 0x400
>>>>> +#define EL2_EXCEPT_ASYNC_OFFSET 0x480
>>>>> +
>>>>> +
>>>>> +/*
>>>>> + * Emulate taking an exception. See ARM ARM J8.1.2 AArch64.TakeException()
>>>>> + */
>>>>> +static int kvm_inject_nested(struct kvm_vcpu *vcpu, u64 esr_el2,
>>>>> + int exception_offset)
>>>>> +{
>>>>> + int ret = 1;
>>>>> + kvm_cpu_context_t *ctxt = &vcpu->arch.ctxt;
>>>>> +
>>>>> + /* We don't inject an exception recursively to virtual EL2 */
>>>>> + if (vcpu_mode_el2(vcpu))
>>>>> + BUG();
>>>>> +
>>>>> + ctxt->el2_regs[SPSR_EL2] = *vcpu_cpsr(vcpu);
>>>>> + ctxt->el2_regs[ELR_EL2] = *vcpu_pc(vcpu);
>>>>> + ctxt->el2_regs[ESR_EL2] = esr_el2;
>>>>> +
>>>>> + /* On an exception, PSTATE.SP = 1 */
>>>>> + *vcpu_cpsr(vcpu) = PSR_MODE_EL2h;
>>>>> + *vcpu_cpsr(vcpu) |= (PSR_A_BIT | PSR_F_BIT | PSR_I_BIT | PSR_D_BIT);
>>>>> + *vcpu_pc(vcpu) = ctxt->el2_regs[VBAR_EL2] + exception_offset;
>>>>> +
>>>>> + trace_kvm_inject_nested_exception(vcpu, esr_el2, *vcpu_pc(vcpu));
>>>>> +
>>>>> + return ret;
>>>>> +}
>>>>> +
>>>>> +int kvm_inject_nested_sync(struct kvm_vcpu *vcpu, u64 esr_el2)
>>>>> +{
>>>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_SYNC_OFFSET);
>>>>> +}
>>>>> +
>>>>> +int kvm_inject_nested_irq(struct kvm_vcpu *vcpu)
>>>>> +{
>>>>> + u64 esr_el2 = kvm_vcpu_get_hsr(vcpu);
>>>>> + /* We supports only IRQ and FIQ, so the esr_el2 is not updated. */
>>>>> + return kvm_inject_nested(vcpu, esr_el2, EL2_EXCEPT_ASYNC_OFFSET);
>>>>> +}
>>>>> diff --git a/arch/arm64/kvm/trace.h b/arch/arm64/kvm/trace.h
>>>>> index 7fb0008..7c86cfb 100644
>>>>> --- a/arch/arm64/kvm/trace.h
>>>>> +++ b/arch/arm64/kvm/trace.h
>>>>> @@ -167,6 +167,26 @@
>>>>> );
>>>>>
>>>>>
>>>>> +TRACE_EVENT(kvm_inject_nested_exception,
>>>>> + TP_PROTO(struct kvm_vcpu *vcpu, unsigned long esr_el2,
>>>>> + unsigned long pc),
>>>>> + TP_ARGS(vcpu, esr_el2, pc),
>>>>> +
>>>>> + TP_STRUCT__entry(
>>>>> + __field(struct kvm_vcpu *, vcpu)
>>>>> + __field(unsigned long, esr_el2)
>>>>> + __field(unsigned long, pc)
>>>>> + ),
>>>>> +
>>>>> + TP_fast_assign(
>>>>> + __entry->vcpu = vcpu;
>>>>> + __entry->esr_el2 = esr_el2;
>>>>> + __entry->pc = pc;
>>>>> + ),
>>>>> +
>>>>> + TP_printk("vcpu: %p, inject exception to vEL2: ESR_EL2 0x%lx, vector: 0x%016lx",
>>>>> + __entry->vcpu, __entry->esr_el2, __entry->pc)
>>>>> +);
>>>>> #endif /* _TRACE_ARM64_KVM_H */
>>>>>
>>>>> #undef TRACE_INCLUDE_PATH
>>>> _______________________________________________
>>>> kvmarm mailing list
>>>> [email protected]
>>>> https://lists.cs.columbia.edu/mailman/listinfo/kvmarm