Re: [PATCH v16 06/99] qtest/arm-cpu-features: Remove TCG fallback to KVM specific tests

[Philippe Mathieu-Daudé]

On 6/7/21 3:28 PM, Thomas Huth wrote:
> On 04/06/2021 17.51, Alex Bennée wrote:
>> From: Philippe Mathieu-Daudé <philmd@redhat.com>
>>
>> sve_tests_sve_off_kvm() and test_query_cpu_model_expansion_kvm()
>> tests are now only being run if KVM is available. Drop the TCG
>> fallback.
>>
>> Suggested-by: Andrew Jones <drjones@redhat.com>
>> Reviewed-by: Andrew Jones <drjones@redhat.com>
>> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
>> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> Message-Id: <20210505125806.1263441-7-philmd@redhat.com>
>> ---
>>   tests/qtest/arm-cpu-features.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/tests/qtest/arm-cpu-features.c
>> b/tests/qtest/arm-cpu-features.c
>> index 66300c3bc2..b1d406542f 100644
>> --- a/tests/qtest/arm-cpu-features.c
>> +++ b/tests/qtest/arm-cpu-features.c
>> @@ -21,7 +21,7 @@
>>   #define SVE_MAX_VQ 16
>>     #define MACHINE     "-machine virt,gic-version=max -accel tcg "
>> -#define MACHINE_KVM "-machine virt,gic-version=max -accel kvm -accel
>> tcg "
>> +#define MACHINE_KVM "-machine virt,gic-version=max -accel kvm "
> 
> Same comment as with patch 04/99: I think this is wrong. You're mixing
> up whether an accelerator has been built into the binary with the fact
> whether an accelerator is available and usable. There are plenty of
> cases where e.g. kvm is built into the binary but not usable during
> runtime, e.g. because:
> 1) The kernel does not support it
> 2) The current host CPU does not support it
> 3) There are problems with the permission to /dev/kvm
> etc.

Yes.

> I think we either need the fallback mechanism to tcg

No, this is precisely what we want to test.

> or you need to
> properly check whether KVM is usable, too.

Yes.