vmx: use mov instead of lea in __vmx_vcpu_run()
PATCH 1/1 vmx: use mov instead of lea in __vmx_vcpu_run()
"mov rsi, rsp" is equivalent to "lea rsi, [rsp]" but 1 byte shorter.
Signed-off-by Alexey Dobriyan <[email protected]>
---
arch/x86/kvm/vmx/vmenter.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/vmx/vmenter.S +1 -1
@@ -72,7 +72,7 @@ SYM_FUNC_START(__vmx_vcpu_run) /* Copy @flags to BL, _ASM_ARG3 is volatile. */ mov %_ASM_ARG3B, %bl - lea (%_ASM_SP), %_ASM_ARG2 + mov %_ASM_SP, %_ASM_ARG2 call vmx_update_host_rsp ALTERNATIVE "jmp .Lspec_ctrl_done", "", X86_FEATURE_MSR_SPEC_CTRL
Sean Christopherson
maintainerKVM: VMX: for the shortlog please.
Quoted 1 line · Alexey Dobriyan
"mov rsi, rsp" is equivalent to "lea rsi, [rsp]" but 1 byte shorter.
Eww, Intel syntax ;-)
Quoted 9 lines
Signed-off-by Alexey Dobriyan <[email protected]>---
arch/x86/kvm/vmx/vmenter.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/vmx/vmenter.S SYM_FUNC_START(__vmx_vcpu_run) · line 75
Quoted 5 lines
arch/x86/kvm/vmx/vmenter.S +1
-1@@ -72,7 +72,7 @@ SYM_FUNC_START(__vmx_vcpu_run) /* Copy @flags to BL, _ASM_ARG3 is volatile. */ mov %_ASM_ARG3B, %bl - lea (%_ASM_SP), %_ASM_ARG2 + mov %_ASM_SP, %_ASM_ARG2
I don't have a strong preference. It's probably worth converting, e.g. move elimination on modern CPUs might shave a whole uop.
I'm pretty sure LEA is a holdover from when this code pre-calculated RSP before a series of pushes.
Quoted 3 lines
arch/x86/kvm/vmx/vmenter.S +0
-0call vmx_update_host_rsp ALTERNATIVE "jmp .Lspec_ctrl_done", "", X86_FEATURE_MSR_SPEC_CTRL