KVM: x86: skip host CPUID call for hypervisor leaves
Paolo Bonzini
maintainerPATCH 1/1 KVM: x86: skip host CPUID call for hypervisor leaves
Hypervisor leaves are always synthesized by __do_cpuid_func. Just return zeroes and do not ask the host, it would return a bogus value anyway if it were used.
--- arch/x86/kvm/cpuid.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
arch/x86/kvm/cpuid.c +9 -1
@@ -692,9 +692,17 @@ static struct kvm_cpuid_entry2 *do_host_cpuid(struct kvm_cpuid_array *array, entry = &array->entries[array->nent++]; +memset (entry,0 ,sizeof (*entry)); entry->function = function; entry->index = index; - entry->flags =0 ; +switch (function &0xC0000000 ) { +case 0x40000000 : +/* Hypervisor leaves are always synthesized by __do_cpuid_func. */ +return entry; + +default : +break ; + }cpuid_count (entry->function, entry->index, &entry->eax, &entry->ebx, &entry->ecx, &entry->edx);
Signature · 1 line
2.31.1
Quoted 3 lines · Paolo Bonzini
Hypervisor leaves are always synthesized by
__do_cpuid_func. Just return zeroes and do not ask the host, it would return a bogus value anyway if it were used.
Why always bogus? Nested virtualization is a thing, isn't it? :-) It is, however, true that __do_cpuid_func() will throw the result away.
Quoted 12 lines
Signed-off-by Paolo Bonzini <[email protected]>--- arch/x86/kvm/cpuid.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)arch/x86/kvm/cpuid.c +0
-0@@ -692,9 +692,17 @@ static struct kvm_cpuid_entry2 *do_host_cpuid(struct kvm_cpuid_array *array,
arch/x86/kvm/cpuid.c static struct kvm_cpuid_entry2 *do_host_cpuid(struct kvm_cpuid_array *array, · line 701
Quoted 10 lines
arch/x86/kvm/cpuid.c +5
-1entry = &array->entries[array->nent++]; +memset (entry,0 ,sizeof (*entry)); entry->function = function; entry->index = index; - entry->flags =0 ; +switch (function &0xC0000000 ) { +case 0x40000000 : +/* Hypervisor leaves are always synthesized by __do_cpuid_func. */ +return entry;
FWIW, 0x40000XXX leaves are not the only ones where we don't use do_host_cpuid() result at all, e.g. I can see that we also return constant values for 0x3, 0x5, 0x6, 0xC0000002 - 0xC0000004.
Out of pure curiosity, what's the motivation for the patch? We seem to only use __do_cpuid_func() to serve KVM_GET_SUPPORTED_CPUID/KVM_GET_EMULATED_CPUID, not for kvm_emulate_cpuid() so these few CPUID calls we save here should not give us any performace gain..
arch/x86/kvm/cpuid.c static struct kvm_cpuid_entry2 *do_host_cpuid(struct kvm_cpuid_array *array, · line 708
Quoted 7 lines
arch/x86/kvm/cpuid.c +4
-0+ +default : +break ; + }cpuid_count (entry->function, entry->index, &entry->eax, &entry->ebx, &entry->ecx, &entry->edx);
The patch seems to be correct, so
Signature · 1 line
Vitaly
Paolo Bonzini
maintainerQuoted 8 lines · Vitaly Kuznetsov
Quoted 3 lines · Paolo Bonzini
Hypervisor leaves are always synthesized by
__do_cpuid_func. Just return zeroes and do not ask the host, it would return a bogus value anyway if it were used.Why always bogus? Nested virtualization is a thing, isn't it? :-) It is, however, true that
__do_cpuid_func()will throw the result away.
Well, bogus because all hypercalls and MSRs would go through us so it makes little if any sense (given the current hypercall and MSR code) for the host values to be used in KVM_GET_SUPPORTED_CPUID.
Quoted 8 lines
FWIW, 0x40000XXX leaves are not the only ones where we don't use
do_host_cpuid()result at all, e.g. I can see that we also return constant values for0x3,0x5,0x6,0xC0000002-0xC0000004.Out of pure curiosity, what's the motivation for the patch? We seem to only use
__do_cpuid_func()to serve KVM_GET_SUPPORTED_CPUID/KVM_GET_EMULATED_CPUID, not forkvm_emulate_cpuid()so these few CPUID calls we save here should not give us any performace gain..
I just have it in queue because of another change that I have not submitted yet.
Paolo
Quoted 12 lines
arch/x86/kvm/cpuid.c static struct kvm_cpuid_entry2 *do_host_cpuid(struct kvm_cpuid_array *array, · line 708
Quoted 7 lines
arch/x86/kvm/cpuid.c +4
-0+ +default : +break ; + }cpuid_count (entry->function, entry->index, &entry->eax, &entry->ebx, &entry->ecx, &entry->edx);The patch seems to be correct, so
Reviewed-by Vitaly Kuznetsov <[email protected]>