[v2,2/2] KVM: selftests: Add logic to detect if ioctl() failed because VM was killed

  Add yet another macro to the VM/vCPU ioctl() framework to detect when an
ioctl() failed because KVM killed/bugged the VM, i.e. when there was
nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
-EIO, which can be quite misleading and ultimately waste user/developer
time.

Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
heuristic is obviously less than ideal, but practically speaking the logic
is bulletproof barring a KVM change, and any such change would arguably
break userspace, e.g. if KVM returns something other than -EIO.

Without the detection, tearing down a bugged VM yields a cryptic failure
when deleting memslots:

  ==== Test Assertion Failure ====
  lib/kvm_util.c:689: !ret
  pid=45131 tid=45131 errno=5 - Input/output error
     1	0x00000000004036c3: __vm_mem_region_delete at kvm_util.c:689
     2	0x00000000004042f0: kvm_vm_free at kvm_util.c:724 (discriminator 12)
     3	0x0000000000402929: race_sync_regs at sync_regs_test.c:193
     4	0x0000000000401cab: main at sync_regs_test.c:334 (discriminator 6)
     5	0x0000000000416f13: __libc_start_call_main at libc-start.o:?
     6	0x000000000041855f: __libc_start_main_impl at ??:?
     7	0x0000000000401d40: _start at ??:?
  KVM_SET_USER_MEMORY_REGION failed, rc: -1 errno: 5 (Input/output error)

Which morphs into a more pointed error message with the detection:

  ==== Test Assertion Failure ====
  lib/kvm_util.c:689: false
  pid=80347 tid=80347 errno=5 - Input/output error
     1	0x00000000004039ab: __vm_mem_region_delete at kvm_util.c:689 (discriminator 5)
     2	0x0000000000404660: kvm_vm_free at kvm_util.c:724 (discriminator 12)
     3	0x0000000000402ac9: race_sync_regs at sync_regs_test.c:193
     4	0x0000000000401cb7: main at sync_regs_test.c:334 (discriminator 6)
     5	0x0000000000418263: __libc_start_call_main at libc-start.o:?
     6	0x00000000004198af: __libc_start_main_impl at ??:?
     7	0x0000000000401d90: _start at ??:?
  KVM killed/bugged the VM, check the kernel log for clues

Suggested-by: Michal Luczaj <mhal@rbox.co>
Cc: Oliver Upton <oliver.upton@linux.dev>
Cc: Colton Lewis <coltonlewis@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
---
 .../selftests/kvm/include/kvm_util_base.h     | 39 ++++++++++++++++---
 tools/testing/selftests/kvm/lib/kvm_util.c    |  2 +-
 2 files changed, 35 insertions(+), 6 deletions(-)
  

On 11/8/2023 9:09 AM, Sean Christopherson wrote:
> Add yet another macro to the VM/vCPU ioctl() framework to detect when an
> ioctl() failed because KVM killed/bugged the VM, i.e. when there was
> nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
> a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
> -EIO, which can be quite misleading and ultimately waste user/developer
> time.
> 
> Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
> dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
> heuristic is obviously less than ideal, but practically speaking the logic
> is bulletproof barring a KVM change, and any such change would arguably
> break userspace, e.g. if KVM returns something other than -EIO.

We hit similar issue when testing TDX VMs. Most failure of SEMCALL is 
handled with a KVM_BUG_ON(), which leads to vm dead. Then the following 
IOCTL from userspace (QEMU) and gets -EIO.

Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD? and replace 
-EIO with 0? yes, it's a ABI change. But I'm wondering if any userspace 
relies on -EIO behavior for VM DEAD case.

> Without the detection, tearing down a bugged VM yields a cryptic failure
> when deleting memslots:
> 
>    ==== Test Assertion Failure ====
>    lib/kvm_util.c:689: !ret
>    pid=45131 tid=45131 errno=5 - Input/output error
>       1	0x00000000004036c3: __vm_mem_region_delete at kvm_util.c:689
>       2	0x00000000004042f0: kvm_vm_free at kvm_util.c:724 (discriminator 12)
>       3	0x0000000000402929: race_sync_regs at sync_regs_test.c:193
>       4	0x0000000000401cab: main at sync_regs_test.c:334 (discriminator 6)
>       5	0x0000000000416f13: __libc_start_call_main at libc-start.o:?
>       6	0x000000000041855f: __libc_start_main_impl at ??:?
>       7	0x0000000000401d40: _start at ??:?
>    KVM_SET_USER_MEMORY_REGION failed, rc: -1 errno: 5 (Input/output error)
> 
> Which morphs into a more pointed error message with the detection:
> 
>    ==== Test Assertion Failure ====
>    lib/kvm_util.c:689: false
>    pid=80347 tid=80347 errno=5 - Input/output error
>       1	0x00000000004039ab: __vm_mem_region_delete at kvm_util.c:689 (discriminator 5)
>       2	0x0000000000404660: kvm_vm_free at kvm_util.c:724 (discriminator 12)
>       3	0x0000000000402ac9: race_sync_regs at sync_regs_test.c:193
>       4	0x0000000000401cb7: main at sync_regs_test.c:334 (discriminator 6)
>       5	0x0000000000418263: __libc_start_call_main at libc-start.o:?
>       6	0x00000000004198af: __libc_start_main_impl at ??:?
>       7	0x0000000000401d90: _start at ??:?
>    KVM killed/bugged the VM, check the kernel log for clues
> 
> Suggested-by: Michal Luczaj <mhal@rbox.co>
> Cc: Oliver Upton <oliver.upton@linux.dev>
> Cc: Colton Lewis <coltonlewis@google.com>
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---
>   .../selftests/kvm/include/kvm_util_base.h     | 39 ++++++++++++++++---
>   tools/testing/selftests/kvm/lib/kvm_util.c    |  2 +-
>   2 files changed, 35 insertions(+), 6 deletions(-)
> 
> diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
> index 1f6193dc7d3a..c7717942ddbb 100644
> --- a/tools/testing/selftests/kvm/include/kvm_util_base.h
> +++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
> @@ -282,11 +282,40 @@ static __always_inline void static_assert_is_vm(struct kvm_vm *vm) { }
>   	kvm_do_ioctl((vm)->fd, cmd, arg);			\
>   })
>   
> +/*
> + * Assert that a VM or vCPU ioctl() succeeded, with extra magic to detect if
> + * the ioctl() failed because KVM killed/bugged the VM.  To detect a dead VM,
> + * probe KVM_CAP_USER_MEMORY, which (a) has been supported by KVM since before
> + * selftests existed and (b) should never outright fail, i.e. is supposed to
> + * return 0 or 1.  If KVM kills a VM, KVM returns -EIO for all ioctl()s for the
> + * VM and its vCPUs, including KVM_CHECK_EXTENSION.
> + */
> +#define __TEST_ASSERT_VM_VCPU_IOCTL(cond, name, ret, vm)				\
> +do {											\
> +	int __errno = errno;								\
> +											\
> +	static_assert_is_vm(vm);							\
> +											\
> +	if (cond)									\
> +		break;									\
> +											\
> +	if (errno == EIO &&								\
> +	    __vm_ioctl(vm, KVM_CHECK_EXTENSION, (void *)KVM_CAP_USER_MEMORY) < 0) {	\
> +		TEST_ASSERT(errno == EIO, "KVM killed the VM, should return -EIO");	\
> +		TEST_FAIL("KVM killed/bugged the VM, check the kernel log for clues");	\
> +	}										\
> +	errno = __errno;								\
> +	TEST_ASSERT(cond, __KVM_IOCTL_ERROR(name, ret));				\
> +} while (0)
> +
> +#define TEST_ASSERT_VM_VCPU_IOCTL(cond, cmd, ret, vm)		\
> +	__TEST_ASSERT_VM_VCPU_IOCTL(cond, #cmd, ret, vm)
> +
>   #define vm_ioctl(vm, cmd, arg)					\
>   ({								\
>   	int ret = __vm_ioctl(vm, cmd, arg);			\
>   								\
> -	TEST_ASSERT(!ret, __KVM_IOCTL_ERROR(#cmd, ret));	\
> +	__TEST_ASSERT_VM_VCPU_IOCTL(!ret, #cmd, ret, vm);		\
>   })
>   
>   static __always_inline void static_assert_is_vcpu(struct kvm_vcpu *vcpu) { }
> @@ -301,7 +330,7 @@ static __always_inline void static_assert_is_vcpu(struct kvm_vcpu *vcpu) { }
>   ({								\
>   	int ret = __vcpu_ioctl(vcpu, cmd, arg);			\
>   								\
> -	TEST_ASSERT(!ret, __KVM_IOCTL_ERROR(#cmd, ret));	\
> +	__TEST_ASSERT_VM_VCPU_IOCTL(!ret, #cmd, ret, (vcpu)->vm);	\
>   })
>   
>   /*
> @@ -312,7 +341,7 @@ static inline int vm_check_cap(struct kvm_vm *vm, long cap)
>   {
>   	int ret =  __vm_ioctl(vm, KVM_CHECK_EXTENSION, (void *)cap);
>   
> -	TEST_ASSERT(ret >= 0, KVM_IOCTL_ERROR(KVM_CHECK_EXTENSION, ret));
> +	TEST_ASSERT_VM_VCPU_IOCTL(ret >= 0, KVM_CHECK_EXTENSION, ret, vm);
>   	return ret;
>   }
>   
> @@ -371,7 +400,7 @@ static inline int vm_get_stats_fd(struct kvm_vm *vm)
>   {
>   	int fd = __vm_ioctl(vm, KVM_GET_STATS_FD, NULL);
>   
> -	TEST_ASSERT(fd >= 0, KVM_IOCTL_ERROR(KVM_GET_STATS_FD, fd));
> +	TEST_ASSERT_VM_VCPU_IOCTL(fd >= 0, KVM_GET_STATS_FD, fd, vm);
>   	return fd;
>   }
>   
> @@ -583,7 +612,7 @@ static inline int vcpu_get_stats_fd(struct kvm_vcpu *vcpu)
>   {
>   	int fd = __vcpu_ioctl(vcpu, KVM_GET_STATS_FD, NULL);
>   
> -	TEST_ASSERT(fd >= 0, KVM_IOCTL_ERROR(KVM_GET_STATS_FD, fd));
> +	TEST_ASSERT_VM_VCPU_IOCTL(fd >= 0, KVM_CHECK_EXTENSION, fd, vcpu->vm);
>   	return fd;
>   }
>   
> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
> index 7a8af1821f5d..c847f942cd38 100644
> --- a/tools/testing/selftests/kvm/lib/kvm_util.c
> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c
> @@ -1227,7 +1227,7 @@ struct kvm_vcpu *__vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpu_id)
>   	vcpu->vm = vm;
>   	vcpu->id = vcpu_id;
>   	vcpu->fd = __vm_ioctl(vm, KVM_CREATE_VCPU, (void *)(unsigned long)vcpu_id);
> -	TEST_ASSERT(vcpu->fd >= 0, KVM_IOCTL_ERROR(KVM_CREATE_VCPU, vcpu->fd));
> +	TEST_ASSERT_VM_VCPU_IOCTL(vcpu->fd >= 0, KVM_CREATE_VCPU, vcpu->fd, vm);
>   
>   	TEST_ASSERT(vcpu_mmap_sz() >= sizeof(*vcpu->run), "vcpu mmap size "
>   		"smaller than expected, vcpu_mmap_sz: %i expected_min: %zi",
  

On Wed, Nov 08, 2023, Xiaoyao Li wrote:
> On 11/8/2023 9:09 AM, Sean Christopherson wrote:
> > Add yet another macro to the VM/vCPU ioctl() framework to detect when an
> > ioctl() failed because KVM killed/bugged the VM, i.e. when there was
> > nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
> > a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
> > -EIO, which can be quite misleading and ultimately waste user/developer
> > time.
> > 
> > Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
> > dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
> > heuristic is obviously less than ideal, but practically speaking the logic
> > is bulletproof barring a KVM change, and any such change would arguably
> > break userspace, e.g. if KVM returns something other than -EIO.
> 
> We hit similar issue when testing TDX VMs. Most failure of SEMCALL is
> handled with a KVM_BUG_ON(), which leads to vm dead. Then the following
> IOCTL from userspace (QEMU) and gets -EIO.
> 
> Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD?

Why?  Even if KVM_EXIT_VM_DEAD somehow provided enough information to be useful
from an automation perspective, the VM is obviously dead.  I don't see how the
VMM can do anything but log the error and tear down the VM.  KVM_BUG_ON() comes
with a WARN, which will be far more helpful for a human debugger, e.g. because
all vCPUs would exit with KVM_EXIT_VM_DEAD, it wouldn't even identify which vCPU
initially triggered the issue.

Using an exit reason is a also bit tricky because it requires a vCPU, whereas a
dead VM blocks anything and everything.

> and replace -EIO with 0? yes, it's a ABI change.

Definitely a "no" on this one.  As has been established by the guest_memfd series,
it's ok to return -1/errno with a valid exit_reason.

> But I'm wondering if any userspace relies on -EIO behavior for VM DEAD case.

I doubt userspace relies on -EIO, but userpsace definitely relies on -1/errno being
returned when a fatal error.
  

On 11/9/2023 12:07 AM, Sean Christopherson wrote:
> On Wed, Nov 08, 2023, Xiaoyao Li wrote:
>> On 11/8/2023 9:09 AM, Sean Christopherson wrote:
>>> Add yet another macro to the VM/vCPU ioctl() framework to detect when an
>>> ioctl() failed because KVM killed/bugged the VM, i.e. when there was
>>> nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
>>> a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
>>> -EIO, which can be quite misleading and ultimately waste user/developer
>>> time.
>>>
>>> Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
>>> dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
>>> heuristic is obviously less than ideal, but practically speaking the logic
>>> is bulletproof barring a KVM change, and any such change would arguably
>>> break userspace, e.g. if KVM returns something other than -EIO.
>>
>> We hit similar issue when testing TDX VMs. Most failure of SEMCALL is
>> handled with a KVM_BUG_ON(), which leads to vm dead. Then the following
>> IOCTL from userspace (QEMU) and gets -EIO.
>>
>> Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD?
> 
> Why?  Even if KVM_EXIT_VM_DEAD somehow provided enough information to be useful
> from an automation perspective, the VM is obviously dead.  I don't see how the
> VMM can do anything but log the error and tear down the VM.  KVM_BUG_ON() comes
> with a WARN, which will be far more helpful for a human debugger, e.g. because
> all vCPUs would exit with KVM_EXIT_VM_DEAD, it wouldn't even identify which vCPU
> initially triggered the issue.

It's not about providing more helpful debugging info, but to provide a 
dedicated notification for VMM that "the VM is dead, all the following 
command may not response". With it, VMM can get rid of the tricky 
detection like this patch.

> Using an exit reason is a also bit tricky because it requires a vCPU, whereas a
> dead VM blocks anything and everything.

No argue of it. It cannot work for all the case, but at least it can 
make some case happier.

>> and replace -EIO with 0? yes, it's a ABI change.
> 
> Definitely a "no" on this one.  As has been established by the guest_memfd series,
> it's ok to return -1/errno with a valid exit_reason.
> 
>> But I'm wondering if any userspace relies on -EIO behavior for VM DEAD case.
> 
> I doubt userspace relies on -EIO, but userpsace definitely relies on -1/errno being
> returned when a fatal error.

what about KVM_EXIT_SHUTDOWN? Or KVM_EXIT_INTERNAL_ERROR?
  

On Mon, Nov 13, 2023, Xiaoyao Li wrote:
> On 11/9/2023 12:07 AM, Sean Christopherson wrote:
> > On Wed, Nov 08, 2023, Xiaoyao Li wrote:
> > > On 11/8/2023 9:09 AM, Sean Christopherson wrote:
> > > > Add yet another macro to the VM/vCPU ioctl() framework to detect when an
> > > > ioctl() failed because KVM killed/bugged the VM, i.e. when there was
> > > > nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
> > > > a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
> > > > -EIO, which can be quite misleading and ultimately waste user/developer
> > > > time.
> > > > 
> > > > Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
> > > > dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
> > > > heuristic is obviously less than ideal, but practically speaking the logic
> > > > is bulletproof barring a KVM change, and any such change would arguably
> > > > break userspace, e.g. if KVM returns something other than -EIO.
> > > 
> > > We hit similar issue when testing TDX VMs. Most failure of SEMCALL is
> > > handled with a KVM_BUG_ON(), which leads to vm dead. Then the following
> > > IOCTL from userspace (QEMU) and gets -EIO.
> > > 
> > > Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD?
> > 
> > Why?  Even if KVM_EXIT_VM_DEAD somehow provided enough information to be useful
> > from an automation perspective, the VM is obviously dead.  I don't see how the
> > VMM can do anything but log the error and tear down the VM.  KVM_BUG_ON() comes
> > with a WARN, which will be far more helpful for a human debugger, e.g. because
> > all vCPUs would exit with KVM_EXIT_VM_DEAD, it wouldn't even identify which vCPU
> > initially triggered the issue.
> 
> It's not about providing more helpful debugging info, but to provide a
> dedicated notification for VMM that "the VM is dead, all the following
> command may not response". With it, VMM can get rid of the tricky detection
> like this patch.

But a VMM doesn't need this tricky detection, because this tricky detections isn't
about detecting that the VM is dead, it's all about helping a human debug why a
test failed.

-EIO already effectively says "the VM is dead", e.g. QEMU isn't going to keep trying
to run vCPUs.  Similarly, selftests assert either way, the goal is purely to print
out a unique error message to minimize the chances of confusing the human running
the test (or looking at results).

> > Definitely a "no" on this one.  As has been established by the guest_memfd series,
> > it's ok to return -1/errno with a valid exit_reason.
> > 
> > > But I'm wondering if any userspace relies on -EIO behavior for VM DEAD case.
> > 
> > I doubt userspace relies on -EIO, but userpsace definitely relies on -1/errno being
> > returned when a fatal error.
> 
> what about KVM_EXIT_SHUTDOWN? Or KVM_EXIT_INTERNAL_ERROR?

I don't follow, those are vcpu_run.exit_reason values, not errno values.  Returning
any flavor of KVM_EXIT_*, which are positive values, would break userspace, e.g.
QEMU explicitly looks for "ret < 0", and glibc only treats small-ish negative
values as errors, i.e. a postive return value will be propagated verbatim up to
QEMU.
  

On 11/30/2023 3:22 AM, Sean Christopherson wrote:
> On Mon, Nov 13, 2023, Xiaoyao Li wrote:
>> On 11/9/2023 12:07 AM, Sean Christopherson wrote:
>>> On Wed, Nov 08, 2023, Xiaoyao Li wrote:
>>>> On 11/8/2023 9:09 AM, Sean Christopherson wrote:
>>>>> Add yet another macro to the VM/vCPU ioctl() framework to detect when an
>>>>> ioctl() failed because KVM killed/bugged the VM, i.e. when there was
>>>>> nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
>>>>> a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
>>>>> -EIO, which can be quite misleading and ultimately waste user/developer
>>>>> time.
>>>>>
>>>>> Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
>>>>> dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
>>>>> heuristic is obviously less than ideal, but practically speaking the logic
>>>>> is bulletproof barring a KVM change, and any such change would arguably
>>>>> break userspace, e.g. if KVM returns something other than -EIO.
>>>>
>>>> We hit similar issue when testing TDX VMs. Most failure of SEMCALL is
>>>> handled with a KVM_BUG_ON(), which leads to vm dead. Then the following
>>>> IOCTL from userspace (QEMU) and gets -EIO.
>>>>
>>>> Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD?
>>>
>>> Why?  Even if KVM_EXIT_VM_DEAD somehow provided enough information to be useful
>>> from an automation perspective, the VM is obviously dead.  I don't see how the
>>> VMM can do anything but log the error and tear down the VM.  KVM_BUG_ON() comes
>>> with a WARN, which will be far more helpful for a human debugger, e.g. because
>>> all vCPUs would exit with KVM_EXIT_VM_DEAD, it wouldn't even identify which vCPU
>>> initially triggered the issue.
>>
>> It's not about providing more helpful debugging info, but to provide a
>> dedicated notification for VMM that "the VM is dead, all the following
>> command may not response". With it, VMM can get rid of the tricky detection
>> like this patch.
> 
> But a VMM doesn't need this tricky detection, because this tricky detections isn't
> about detecting that the VM is dead, it's all about helping a human debug why a
> test failed.
> 
> -EIO already effectively says "the VM is dead", e.g. QEMU isn't going to keep trying
> to run vCPUs.  

If -EIO for KVM ioctl denotes "the VM is dead" is to be the officially 
announced API, I'm fine.


> Similarly, selftests assert either way, the goal is purely to print
> out a unique error message to minimize the chances of confusing the human running
> the test (or looking at results).
> 
>>> Definitely a "no" on this one.  As has been established by the guest_memfd series,
>>> it's ok to return -1/errno with a valid exit_reason.
>>>
>>>> But I'm wondering if any userspace relies on -EIO behavior for VM DEAD case.
>>>
>>> I doubt userspace relies on -EIO, but userpsace definitely relies on -1/errno being
>>> returned when a fatal error.
>>
>> what about KVM_EXIT_SHUTDOWN? Or KVM_EXIT_INTERNAL_ERROR?
> 
> I don't follow,

I was trying to ask if KVM_EXIT_SHUTDOWN and KVM_EXIT_INTERNAL_ERROR are 
treated as fatal error by userspace.

> those are vcpu_run.exit_reason values, not errno values.  Returning
> any flavor of KVM_EXIT_*, which are positive values, would break userspace, e.g.
> QEMU explicitly looks for "ret < 0", and glibc only treats small-ish negative
> values as errors, i.e. a postive return value will be propagated verbatim up to
> QEMU.
  

On Thu, Nov 30, 2023, Xiaoyao Li wrote:
> On 11/30/2023 3:22 AM, Sean Christopherson wrote:
> > On Mon, Nov 13, 2023, Xiaoyao Li wrote:
> > > On 11/9/2023 12:07 AM, Sean Christopherson wrote:
> > > > On Wed, Nov 08, 2023, Xiaoyao Li wrote:
> > > > > On 11/8/2023 9:09 AM, Sean Christopherson wrote:
> > > > > > Add yet another macro to the VM/vCPU ioctl() framework to detect when an
> > > > > > ioctl() failed because KVM killed/bugged the VM, i.e. when there was
> > > > > > nothing wrong with the ioctl() itself.  If KVM kills a VM, e.g. by way of
> > > > > > a failed KVM_BUG_ON(), all subsequent VM and vCPU ioctl()s will fail with
> > > > > > -EIO, which can be quite misleading and ultimately waste user/developer
> > > > > > time.
> > > > > > 
> > > > > > Use KVM_CHECK_EXTENSION on KVM_CAP_USER_MEMORY to detect if the VM is
> > > > > > dead and/or bug, as KVM doesn't provide a dedicated ioctl().  Using a
> > > > > > heuristic is obviously less than ideal, but practically speaking the logic
> > > > > > is bulletproof barring a KVM change, and any such change would arguably
> > > > > > break userspace, e.g. if KVM returns something other than -EIO.
> > > > > 
> > > > > We hit similar issue when testing TDX VMs. Most failure of SEMCALL is
> > > > > handled with a KVM_BUG_ON(), which leads to vm dead. Then the following
> > > > > IOCTL from userspace (QEMU) and gets -EIO.
> > > > > 
> > > > > Can we return a new KVM_EXIT_VM_DEAD on KVM_REQ_VM_DEAD?
> > > > 
> > > > Why?  Even if KVM_EXIT_VM_DEAD somehow provided enough information to be useful
> > > > from an automation perspective, the VM is obviously dead.  I don't see how the
> > > > VMM can do anything but log the error and tear down the VM.  KVM_BUG_ON() comes
> > > > with a WARN, which will be far more helpful for a human debugger, e.g. because
> > > > all vCPUs would exit with KVM_EXIT_VM_DEAD, it wouldn't even identify which vCPU
> > > > initially triggered the issue.
> > > 
> > > It's not about providing more helpful debugging info, but to provide a
> > > dedicated notification for VMM that "the VM is dead, all the following
> > > command may not response". With it, VMM can get rid of the tricky detection
> > > like this patch.
> > 
> > But a VMM doesn't need this tricky detection, because this tricky detections isn't
> > about detecting that the VM is dead, it's all about helping a human debug why a
> > test failed.
> > 
> > -EIO already effectively says "the VM is dead", e.g. QEMU isn't going to keep trying
> > to run vCPUs.
> 
> If -EIO for KVM ioctl denotes "the VM is dead" is to be the officially
> announced API, I'm fine.

Yes, -EIO is effectively ABI at this point.  Though there is the caveat that -EIO
doesn't guarantee KVM killed the VM, i.e. KVM could return -EIO for some other
reason (though that's highly unlikely for KVM_RUN at least).

> > Similarly, selftests assert either way, the goal is purely to print
> > out a unique error message to minimize the chances of confusing the human running
> > the test (or looking at results).
> > 
> > > > Definitely a "no" on this one.  As has been established by the guest_memfd series,
> > > > it's ok to return -1/errno with a valid exit_reason.
> > > > 
> > > > > But I'm wondering if any userspace relies on -EIO behavior for VM DEAD case.
> > > > 
> > > > I doubt userspace relies on -EIO, but userpsace definitely relies on -1/errno being
> > > > returned when a fatal error.
> > > 
> > > what about KVM_EXIT_SHUTDOWN? Or KVM_EXIT_INTERNAL_ERROR?
> > 
> > I don't follow,
> 
> I was trying to ask if KVM_EXIT_SHUTDOWN and KVM_EXIT_INTERNAL_ERROR are
> treated as fatal error by userspace.

Ah.  Not really.  SHUTDOWN isn't fatal per se, e.g. QEMU emulates a RESET if a
vCPU hits shutdown.  INTERNAL_ERROR isn't always fatal on x86, e.g. QEMU ignores
(I think that's what happens) emulation failure when the vCPU is at CPL > 0 so
that guest userspace can't DoS the VM.