On Fri, Feb 17, 2023, Aaron Lewis wrote:
On Tue, Feb 14, 2023 at 6:46 PM Mingwei Zhang mizhang@google.com wrote:
Add check of CR0.TS[bit 3] before the check of IA32_XFD_ERR in the #NM handler in amx_test. This is because XFD may not be the only reason of the IA32_XFD MSR and the bitmap corresponding to the state components required by the faulting instruction." (Intel SDM vol 1. Section 13.14)
Add the missing check of CR0.TS.
Signed-off-by: Mingwei Zhang mizhang@google.com
tools/testing/selftests/kvm/x86_64/amx_test.c | 1 + 1 file changed, 1 insertion(+)
diff --git a/tools/testing/selftests/kvm/x86_64/amx_test.c b/tools/testing/selftests/kvm/x86_64/amx_test.c index aac727ff7cf8..847752998660 100644 --- a/tools/testing/selftests/kvm/x86_64/amx_test.c +++ b/tools/testing/selftests/kvm/x86_64/amx_test.c @@ -215,6 +215,7 @@ void guest_nm_handler(struct ex_regs *regs) { /* Check if #NM is triggered by XFEATURE_MASK_XTILEDATA */ GUEST_SYNC(7);
GUEST_ASSERT((get_cr0() & X86_CR0_TS) == 0);Can't we infer that the #NM is the result of an XFD error due to the fact that IA32_XFD_ERR is set? Is this check needed? SDM vol 1, 13.14, EXTENDED FEATURE DISABLE (XFD)
- Device-not-available exceptions that are not due to XFD - those resulting from setting CR0.TS to 1 - do not modify the IA32_XFD_ERR MSR.
We don't infer from the reasons of #NM and that is the purpose of this selftest. Yes, this looks a little bit pedantic. But still, it is worth adding the check since violation of that indicates either 1) the selftest mistakenly did not clear XFD_ERR prior to #NM or 2) hardware is broken.
GUEST_ASSERT(rdmsr(MSR_IA32_XFD_ERR) == XFEATURE_MASK_XTILEDATA); GUEST_SYNC(8); GUEST_ASSERT(rdmsr(MSR_IA32_XFD_ERR) == XFEATURE_MASK_XTILEDATA);-- 2.39.1.581.gbfd45094c4-goog