From: Zeng Guang <guang.zeng@intel.com>

For VM exits caused by events (NMI, #DB, and #PF) delivered by FRED, the
event data is saved in the exit-qualification field. (FRED spec. 10.6.2)

Expand FRED KVM entry interface to include the event data obtained from
the exit qualification.

Signed-off-by: Zeng Guang <guang.zeng@intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
---
 arch/x86/include/asm/fred.h | 8 ++++----
 arch/x86/kvm/vmx/vmx.c      | 4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/fred.h b/arch/x86/include/asm/fred.h
index e86c7ba32435..15f5d2eabd1d 100644
--- a/arch/x86/include/asm/fred.h
+++ b/arch/x86/include/asm/fred.h
@@ -63,14 +63,14 @@ static __always_inline unsigned long fred_event_data(struct pt_regs *regs)
 
 void asm_fred_entrypoint_user(void);
 void asm_fred_entrypoint_kernel(void);
-void asm_fred_entry_from_kvm(struct fred_ss);
+void asm_fred_entry_from_kvm(struct fred_ss, unsigned long edata);
 
 __visible void fred_entry_from_user(struct pt_regs *regs);
 __visible void fred_entry_from_kernel(struct pt_regs *regs);
 __visible void __fred_entry_from_kvm(struct pt_regs *regs);
 
 /* Can be called from noinstr code, thus __always_inline */
-static __always_inline void fred_entry_from_kvm(unsigned int type, unsigned int vector)
+static __always_inline void fred_entry_from_kvm(unsigned int type, unsigned int vector, unsigned long edata)
 {
 	struct fred_ss ss = {
 		.ss     =__KERNEL_DS,
@@ -80,7 +80,7 @@ static __always_inline void fred_entry_from_kvm(unsigned int type, unsigned int
 		.lm     = 1,
 	};
 
-	asm_fred_entry_from_kvm(ss);
+	asm_fred_entry_from_kvm(ss, edata);
 }
 
 void cpu_init_fred_exceptions(void);
@@ -90,7 +90,7 @@ void fred_complete_exception_setup(void);
 static __always_inline unsigned long fred_event_data(struct pt_regs *regs) { return 0; }
 static inline void cpu_init_fred_exceptions(void) { }
 static inline void fred_complete_exception_setup(void) { }
-static __always_inline void fred_entry_from_kvm(unsigned int type, unsigned int vector) { }
+static __always_inline void fred_entry_from_kvm(unsigned int type, unsigned int vector, unsigned long edata) { }
 #endif /* CONFIG_X86_FRED */
 #endif /* !__ASSEMBLY__ */
 
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index b3c83c06f826..4e7b36081b76 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -7024,7 +7024,7 @@ static void handle_external_interrupt_irqoff(struct kvm_vcpu *vcpu,
 
 	kvm_before_interrupt(vcpu, KVM_HANDLING_IRQ);
 	if (cpu_feature_enabled(X86_FEATURE_FRED))
-		fred_entry_from_kvm(EVENT_TYPE_EXTINT, vector);
+		fred_entry_from_kvm(EVENT_TYPE_EXTINT, vector, 0);
 	else
 		vmx_do_interrupt_irqoff(gate_offset((gate_desc *)host_idt_base + vector));
 	kvm_after_interrupt(vcpu);
@@ -7332,7 +7332,7 @@ static noinstr void vmx_vcpu_enter_exit(struct kvm_vcpu *vcpu,
 	    is_nmi(vmx_get_intr_info(vcpu))) {
 		kvm_before_interrupt(vcpu, KVM_HANDLING_NMI);
 		if (cpu_feature_enabled(X86_FEATURE_FRED))
-			fred_entry_from_kvm(EVENT_TYPE_NMI, NMI_VECTOR);
+			fred_entry_from_kvm(EVENT_TYPE_NMI, NMI_VECTOR, 0);
 		else
 			vmx_do_nmi_irqoff();
 		kvm_after_interrupt(vcpu);
-- 
2.25.1

On 6/28/2024 1:18 PM, Jacob Pan wrote:
> From: Zeng Guang <guang.zeng@intel.com>
> 
> For VM exits caused by events (NMI, #DB, and #PF) delivered by FRED, the
> event data is saved in the exit-qualification field. (FRED spec. 10.6.2)

I don't like mentioning #DB/#PF here, they belong to the guest that was 
running, and KVM handles them.

While NMIs belong to host, and the host NMI handler needs event data
saved in VMCS in NMI induced VM exits.

Or you paint a full picture.

> Expand FRED KVM entry interface to include the event data obtained from
> the exit qualification.
> 
> Signed-off-by: Zeng Guang <guang.zeng@intel.com>
> Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>

On Fri, 28 Jun 2024 21:01:28 -0700, Xin Li <xin@zytor.com> wrote:

> On 6/28/2024 1:18 PM, Jacob Pan wrote:
> > From: Zeng Guang <guang.zeng@intel.com>
> > 
> > For VM exits caused by events (NMI, #DB, and #PF) delivered by FRED, the
> > event data is saved in the exit-qualification field. (FRED spec.
> > 10.6.2)  
> 
> I don't like mentioning #DB/#PF here, they belong to the guest that was 
> running, and KVM handles them.
It is part of the spec.

> While NMIs belong to host, and the host NMI handler needs event data
> saved in VMCS in NMI induced VM exits.
> 
> Or you paint a full picture.
I will add your explanation that #DB/#PF belong to the guest so readers can
get a full picture of both usage and spec.
> 
> > Expand FRED KVM entry interface to include the event data obtained from
> > the exit qualification.
> > 
> > Signed-off-by: Zeng Guang <guang.zeng@intel.com>
> > Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>  
> 


Thanks,

Jacob