Re: [PATCH] target/arm/kvm: Check supported feature per accelerator (not per vCPU)

[Philippe Mathieu-Daudé]

On 6/18/20 11:22 AM, Andrew Jones wrote:
> On Wed, Jun 17, 2020 at 03:08:00PM +0200, Philippe Mathieu-Daudé wrote:
>> Since commit d70c996df23f, when enabling the PMU we get:
>>
>>   $ qemu-system-aarch64 -cpu host,pmu=on -M virt,accel=kvm,gic-version=3
>>   Segmentation fault (core dumped)
>>
>>   Thread 1 "qemu-system-aar" received signal SIGSEGV, Segmentation fault.
>>   0x0000aaaaaae356d0 in kvm_ioctl (s=0x0, type=44547) at 
>> accel/kvm/kvm-all.c:2588
>>   2588        ret = ioctl(s->fd, type, arg);
>>   (gdb) bt
>>   #0  0x0000aaaaaae356d0 in kvm_ioctl (s=0x0, type=44547) at 
>> accel/kvm/kvm-all.c:2588
>>   #1  0x0000aaaaaae31568 in kvm_check_extension (s=0x0, extension=126) at 
>> accel/kvm/kvm-all.c:916
>>   #2  0x0000aaaaaafce254 in kvm_arm_pmu_supported (cpu=0xaaaaac214ab0) at 
>> target/arm/kvm.c:213
>>   #3  0x0000aaaaaafc0f94 in arm_set_pmu (obj=0xaaaaac214ab0, value=true, 
>> errp=0xffffffffe438) at target/arm/cpu.c:1111
>>   #4  0x0000aaaaab5533ac in property_set_bool (obj=0xaaaaac214ab0, 
>> v=0xaaaaac223a80, name=0xaaaaac11a970 "pmu", opaque=0xaaaaac222730, 
>> errp=0xffffffffe438) at qom/object.c:2170
>>   #5  0x0000aaaaab5512f0 in object_property_set (obj=0xaaaaac214ab0, 
>> v=0xaaaaac223a80, name=0xaaaaac11a970 "pmu", errp=0xffffffffe438) at 
>> qom/object.c:1328
>>   #6  0x0000aaaaab551e10 in object_property_parse (obj=0xaaaaac214ab0, 
>> string=0xaaaaac11b4c0 "on", name=0xaaaaac11a970 "pmu", errp=0xffffffffe438) 
>> at qom/object.c:1561
>>   #7  0x0000aaaaab54ee8c in object_apply_global_props (obj=0xaaaaac214ab0, 
>> props=0xaaaaac018e20, errp=0xaaaaabd6fd88 <error_fatal>) at qom/object.c:407
>>   #8  0x0000aaaaab1dd5a4 in qdev_prop_set_globals (dev=0xaaaaac214ab0) at 
>> hw/core/qdev-properties.c:1218
>>   #9  0x0000aaaaab1d9fac in device_post_init (obj=0xaaaaac214ab0) at 
>> hw/core/qdev.c:1050
>>   ...
>>   #15 0x0000aaaaab54f310 in object_initialize_with_type (obj=0xaaaaac214ab0, 
>> size=52208, type=0xaaaaabe237f0) at qom/object.c:512
>>   #16 0x0000aaaaab54fa24 in object_new_with_type (type=0xaaaaabe237f0) at 
>> qom/object.c:687
>>   #17 0x0000aaaaab54fa80 in object_new (typename=0xaaaaabe23970 
>> "host-arm-cpu") at qom/object.c:702
>>   #18 0x0000aaaaaaf04a74 in machvirt_init (machine=0xaaaaac0a8550) at 
>> hw/arm/virt.c:1770
>>   #19 0x0000aaaaab1e8720 in machine_run_board_init (machine=0xaaaaac0a8550) 
>> at hw/core/machine.c:1138
>>   #20 0x0000aaaaaaf95394 in qemu_init (argc=5, argv=0xffffffffea58, 
>> envp=0xffffffffea88) at softmmu/vl.c:4348
>>   #21 0x0000aaaaaada3f74 in main (argc=<optimized out>, argv=<optimized 
>> out>, envp=<optimized out>) at softmmu/main.c:48
>>
>> This is because in frame #2, cpu->kvm_state is still NULL
>> (the vCPU is not yet realized).
>>
>> KVM has a hard requirement of all cores supporting the same
>> feature set. We only need to check if the accelerator supports
>> a feature, not each vCPU individually.
>>
>> Fix by kvm_arm_<FEATURE>_supported() functions take a AccelState
>> argument (already realized/valid at this point) instead of a
>> CPUState argument.
>>
>> Reported-by: Haibo Xu <haibo.xu@linaro.org>
>> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> ---
>>  target/arm/kvm_arm.h | 25 +++++++++++++------------
>>  target/arm/cpu.c     |  2 +-
>>  target/arm/cpu64.c   | 10 +++++-----
>>  target/arm/kvm.c     |  4 ++--
>>  target/arm/kvm64.c   | 14 +++++---------
>>  5 files changed, 26 insertions(+), 29 deletions(-)
>>
>> diff --git a/target/arm/kvm_arm.h b/target/arm/kvm_arm.h
>> index 48bf5e16d5..8209525f20 100644
>> --- a/target/arm/kvm_arm.h
>> +++ b/target/arm/kvm_arm.h
>> @@ -12,6 +12,7 @@
>>  #define QEMU_KVM_ARM_H
>>  
>>  #include "sysemu/kvm.h"
>> +#include "sysemu/accel.h"
>>  #include "exec/memory.h"
>>  #include "qemu/error-report.h"
>>  
>> @@ -269,29 +270,29 @@ void kvm_arm_add_vcpu_properties(Object *obj);
>>  
>>  /**
>>   * kvm_arm_aarch32_supported:
>> - * @cs: CPUState
>> + * @as: AccelState
>>   *
>> - * Returns: true if the KVM VCPU can enable AArch32 mode
>> + * Returns: true if the KVM accelerator can enable AArch32 mode
>>   * and false otherwise.
>>   */
>> -bool kvm_arm_aarch32_supported(CPUState *cs);
>> +bool kvm_arm_aarch32_supported(AccelState *as);
>>  
>>  /**
>>   * kvm_arm_pmu_supported:
>> - * @cs: CPUState
>> + * @as: AccelState
>>   *
>> - * Returns: true if the KVM VCPU can enable its PMU
>> + * Returns: true if the KVM accelerator can enable its PMU
>>   * and false otherwise.
>>   */
>> -bool kvm_arm_pmu_supported(CPUState *cs);
>> +bool kvm_arm_pmu_supported(AccelState *as);
>>  
>>  /**
>>   * kvm_arm_sve_supported:
>> - * @cs: CPUState
>> + * @as: AccelState
>>   *
>> - * Returns true if the KVM VCPU can enable SVE and false otherwise.
>> + * Returns true if the KVM accelerator can enable SVE and false otherwise.
>>   */
>> -bool kvm_arm_sve_supported(CPUState *cs);
>> +bool kvm_arm_sve_supported(AccelState *as);
>>  
>>  /**
>>   * kvm_arm_get_max_vm_ipa_size:
>> @@ -359,17 +360,17 @@ static inline void 
>> kvm_arm_set_cpu_features_from_host(ARMCPU *cpu)
>>  
>>  static inline void kvm_arm_add_vcpu_properties(Object *obj) {}
>>  
>> -static inline bool kvm_arm_aarch32_supported(CPUState *cs)
>> +static inline bool kvm_arm_aarch32_supported(AccelState *as)
>>  {
>>      return false;
>>  }
>>  
>> -static inline bool kvm_arm_pmu_supported(CPUState *cs)
>> +static inline bool kvm_arm_pmu_supported(AccelState *as)
>>  {
>>      return false;
>>  }
>>  
>> -static inline bool kvm_arm_sve_supported(CPUState *cs)
>> +static inline bool kvm_arm_sve_supported(AccelState *as)
>>  {
>>      return false;
>>  }
>> diff --git a/target/arm/cpu.c b/target/arm/cpu.c
>> index 5b7a36b5d7..29b314427c 100644
>> --- a/target/arm/cpu.c
>> +++ b/target/arm/cpu.c
>> @@ -1108,7 +1108,7 @@ static void arm_set_pmu(Object *obj, bool value, Error 
>> **errp)
>>      ARMCPU *cpu = ARM_CPU(obj);
>>  
>>      if (value) {
>> -        if (kvm_enabled() && !kvm_arm_pmu_supported(CPU(cpu))) {
>> +        if (kvm_enabled() && !kvm_arm_pmu_supported(current_accel())) {
>>              error_setg(errp, "'pmu' feature not supported by KVM on this 
>> host");
>>              return;
>>          }
>> diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c
>> index 778cecc2e6..13835768ab 100644
>> --- a/target/arm/cpu64.c
>> +++ b/target/arm/cpu64.c
>> @@ -266,7 +266,7 @@ void arm_cpu_sve_finalize(ARMCPU *cpu, Error **errp)
>>  
>>      /* Collect the set of vector lengths supported by KVM. */
>>      bitmap_zero(kvm_supported, ARM_MAX_VQ);
>> -    if (kvm_enabled() && kvm_arm_sve_supported(CPU(cpu))) {
>> +    if (kvm_enabled() && kvm_arm_sve_supported(current_accel())) {
>>          kvm_arm_sve_get_vls(CPU(cpu), kvm_supported);
>>      } else if (kvm_enabled()) {
>>          assert(!cpu_isar_feature(aa64_sve, cpu));
>> @@ -473,7 +473,7 @@ static void cpu_max_set_sve_max_vq(Object *obj, Visitor 
>> *v, const char *name,
>>          return;
>>      }
>>  
>> -    if (kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) {
>> +    if (kvm_enabled() && !kvm_arm_sve_supported(current_accel())) {
>>          error_setg(errp, "cannot set sve-max-vq");
>>          error_append_hint(errp, "SVE not supported by KVM on this host\n");
>>          return;
>> @@ -519,7 +519,7 @@ static void cpu_arm_set_sve_vq(Object *obj, Visitor *v, 
>> const char *name,
>>          return;
>>      }
>>  
>> -    if (value && kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) {
>> +    if (value && kvm_enabled() && !kvm_arm_sve_supported(current_accel())) {
>>          error_setg(errp, "cannot enable %s", name);
>>          error_append_hint(errp, "SVE not supported by KVM on this host\n");
>>          return;
>> @@ -556,7 +556,7 @@ static void cpu_arm_set_sve(Object *obj, Visitor *v, 
>> const char *name,
>>          return;
>>      }
>>  
>> -    if (value && kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) {
>> +    if (value && kvm_enabled() && !kvm_arm_sve_supported(current_accel())) {
>>          error_setg(errp, "'sve' feature not supported by KVM on this host");
>>          return;
>>      }
>> @@ -751,7 +751,7 @@ static void aarch64_cpu_set_aarch64(Object *obj, bool 
>> value, Error **errp)
>>       * uniform execution state like do_interrupt.
>>       */
>>      if (value == false) {
>> -        if (!kvm_enabled() || !kvm_arm_aarch32_supported(CPU(cpu))) {
>> +        if (!kvm_enabled() || !kvm_arm_aarch32_supported(current_accel())) {
>>              error_setg(errp, "'aarch64' feature cannot be disabled "
>>                               "unless KVM is enabled and 32-bit EL1 "
>>                               "is supported");
>> diff --git a/target/arm/kvm.c b/target/arm/kvm.c
>> index eef3bbd1cc..2247a96757 100644
>> --- a/target/arm/kvm.c
>> +++ b/target/arm/kvm.c
>> @@ -208,9 +208,9 @@ void kvm_arm_add_vcpu_properties(Object *obj)
>>      }
>>  }
>>  
>> -bool kvm_arm_pmu_supported(CPUState *cpu)
>> +bool kvm_arm_pmu_supported(AccelState *as)
>>  {
>> -    return kvm_check_extension(cpu->kvm_state, KVM_CAP_ARM_PMU_V3);
>> +    return kvm_check_extension(KVM_STATE(as), KVM_CAP_ARM_PMU_V3);
>>  }
>>  
>>  int kvm_arm_get_max_vm_ipa_size(MachineState *ms)
>> diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c
>> index f09ed9f4df..ae4e37ce78 100644
>> --- a/target/arm/kvm64.c
>> +++ b/target/arm/kvm64.c
>> @@ -652,18 +652,14 @@ bool kvm_arm_get_host_cpu_features(ARMHostCPUFeatures 
>> *ahcf)
>>      return true;
>>  }
>>  
>> -bool kvm_arm_aarch32_supported(CPUState *cpu)
>> +bool kvm_arm_aarch32_supported(AccelState *as)
>>  {
>> -    KVMState *s = KVM_STATE(current_accel());
>> -
>> -    return kvm_check_extension(s, KVM_CAP_ARM_EL1_32BIT);
>> +    return kvm_check_extension(KVM_STATE(as), KVM_CAP_ARM_EL1_32BIT);
>>  }
>>  
>> -bool kvm_arm_sve_supported(CPUState *cpu)
>> +bool kvm_arm_sve_supported(AccelState *as)
>>  {
>> -    KVMState *s = KVM_STATE(current_accel());
>> -
>> -    return kvm_check_extension(s, KVM_CAP_ARM_SVE);
>> +    return kvm_check_extension(KVM_STATE(as), KVM_CAP_ARM_SVE);
>>  }
>>  
>>  QEMU_BUILD_BUG_ON(KVM_ARM64_SVE_VQ_MIN != 1);
>> @@ -798,7 +794,7 @@ int kvm_arch_init_vcpu(CPUState *cs)
>>          env->features &= ~(1ULL << ARM_FEATURE_PMU);
>>      }
>>      if (cpu_isar_feature(aa64_sve, cpu)) {
>> -        assert(kvm_arm_sve_supported(cs));
>> +        assert(kvm_arm_sve_supported(ACCEL(cs->kvm_state)));
> 
> Might as well use current_accel() here too, right?

I was not sure, I used kvm_state since at this point we are
sure it is initialized.

> 
>>          cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_SVE;
>>      }
>>  
>> -- 
>> 2.21.3
>>
>>
> 
> At all callsites we pass current_accel() to the kvm_arm_<feat>_supported()
> functions. Is there any reason not to drop their input parameter and just
> use current_accel() internally?

Clever idea :)

> 
> Thanks,
> drew
>