On 1/12/22 12:19, Li RongQing wrote:
After support paravirtualized TLB shootdowns, steal_time.preempted
includes not only KVM_VCPU_PREEMPTED, but also KVM_VCPU_FLUSH_TLB
and kvm_vcpu_is_preempted should test only with KVM_VCPU_PREEMPTED
The combination of PREEMPTED=0,FLUSH_TLB=1 is invalid and can only
happens if the guest malfunctions (which it doesn't, it uses cmpxchg to
set KVM_VCPU_PREEMPTED); the host only does an xchg with 0 as the new
value. Since this is guest code, this patch does not change an actual
error in the code, does it?
Paolo
Fixes: 858a43aae2367 ("KVM: X86: use paravirtualized TLB Shootdown")
Signed-off-by: Li RongQing<lirongqing@xxxxxxxxx>
---
arch/x86/kernel/kvm.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index 59abbda..a9202d9 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -1025,8 +1025,8 @@ asm(
".type __raw_callee_save___kvm_vcpu_is_preempted, @function;"
"__raw_callee_save___kvm_vcpu_is_preempted:"
"movq __per_cpu_offset(,%rdi,8), %rax;"
-"cmpb $0, " __stringify(KVM_STEAL_TIME_preempted) "+steal_time(%rax);"
-"setne %al;"
+"movb " __stringify(KVM_STEAL_TIME_preempted) "+steal_time(%rax), %al;"
+"andb $" __stringify(KVM_VCPU_PREEMPTED) ", %al;"
"ret;"
".size __raw_callee_save___kvm_vcpu_is_preempted, .-__raw_callee_save___kvm_vcpu_is_preempted;"
".popsection");
