On 7/7/2016 11:27 AM, Jan Beulich wrote:
On 06.07.16 at 17:54, <cz...@bitdefender.com> wrote:
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -475,7 +475,7 @@ void hvm_do_resume(struct vcpu *v)
if ( unlikely(v->arch.vm_event) )
{
- if ( v->arch.vm_event->emulate_flags )
+ if ( unlikely(v->arch.vm_event->emulate_flags) )
{
enum emul_kind kind = EMUL_KIND_NORMAL;
Acked-by: Jan Beulich <jbeul...@suse.com>
--- a/xen/arch/x86/vm_event.c
+++ b/xen/arch/x86/vm_event.c
@@ -96,14 +96,16 @@ void vm_event_register_write_resume(struct vcpu *v,
vm_event_response_t *rsp)
{
if ( rsp->flags & VM_EVENT_FLAG_DENY )
{
- struct monitor_write_data *w = &v->arch.vm_event->write_data;
+ struct monitor_write_data *w;
- ASSERT(w);
+ ASSERT(v->arch.vm_event);
/* deny flag requires the vCPU to be paused */
if ( !atomic_read(&v->vm_event_pause_count) )
return;
+ w = &v->arch.vm_event->write_data;
+
switch ( rsp->reason )
{
case VM_EVENT_REASON_MOV_TO_MSR:
I'd have preferred for you to leave alone the initializer, but we'll see
what the maintainers are going to say.
Jan
It's 'cleaner' this way, doesn't it? Not assigning a pointer to a
possibly invalid address...
Anyway, I'm preparing a v4, I'll probably drop this change if you won't
*subdue* to ack it (kidding).
Thanks,
Corneliu.
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
