[PATCH AUTOSEL 3.18 35/61] arm64: KVM: Skip MMIO insn after emulation

[Sasha Levin]

From: Mark Rutland <mark.rutland@xxxxxxx>

[ Upstream commit 0d640732dbebed0f10f18526de21652931f0b2f2 ]

When we emulate an MMIO instruction, we advance the CPU state within
decode_hsr(), before emulating the instruction effects.

Having this logic in decode_hsr() is opaque, and advancing the state
before emulation is problematic. It gets in the way of applying
consistent single-step logic, and it prevents us from being able to fail
an MMIO instruction with a synchronous exception.

Clean this up by only advancing the CPU state *after* the effects of the
instruction are emulated.

Cc: Peter Maydell <peter.maydell@xxxxxxxxxx>
Reviewed-by: Alex BennÃe <alex.bennee@xxxxxxxxxx>
Reviewed-by: Christoffer Dall <christoffer.dall@xxxxxxx>
Signed-off-by: Mark Rutland <mark.rutland@xxxxxxx>
Signed-off-by: Marc Zyngier <marc.zyngier@xxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
 arch/arm/kvm/mmio.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/arch/arm/kvm/mmio.c b/arch/arm/kvm/mmio.c
index 4cb5a93182e9..c215bb4423ac 100644
--- a/arch/arm/kvm/mmio.c
+++ b/arch/arm/kvm/mmio.c
@@ -118,6 +118,12 @@ int kvm_handle_mmio_return(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		*vcpu_reg(vcpu, vcpu->arch.mmio_decode.rt) = data;
 	}
 
+	/*
+	 * The MMIO instruction is emulated and should not be re-executed
+	 * in the guest.
+	 */
+	kvm_skip_instr(vcpu, kvm_vcpu_trap_il_is32bit(vcpu));
+
 	return 0;
 }
 
@@ -154,11 +160,6 @@ static int decode_hsr(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	vcpu->arch.mmio_decode.sign_extend = sign_extend;
 	vcpu->arch.mmio_decode.rt = rt;
 
-	/*
-	 * The MMIO instruction is emulated and should not be re-executed
-	 * in the guest.
-	 */
-	kvm_skip_instr(vcpu, kvm_vcpu_trap_il_is32bit(vcpu));
 	return 0;
 }
 
-- 
2.19.1