Peter Maydell [off-list ref] writes:

On 4 March 2015 at 14:35, Alex Benn?e [off-list ref] wrote:

quoted

This adds the saving and restore of the current Multi-Processing state
of the machine. While the KVM_GET/SET_MP_STATE API exposes a number of
potential states for x86 we only use two for ARM. Either the process is
running or not. We then save this state into the cpu_powered TCG state
to avoid changing the serialisation format.

Signed-off-by: Alex Benn?e <redacted>

---
v2
  - make mpstate field runtime dependant (kvm_enabled())
  - drop initial KVM_CAP_MP_STATE requirement
  - re-use cpu_powered instead of new field

diff --git a/target-arm/machine.c b/target-arm/machine.c
index 9446e5a..185f9a2 100644
--- a/target-arm/machine.c
+++ b/target-arm/machine.c

@@ -161,6 +161,7 @@ static const VMStateInfo vmstate_cpsr = {
     .put = put_cpsr,
 };

+
 static void cpu_pre_save(void *opaque)
 {
     ARMCPU *cpu = opaque;

@@ -170,6 +171,20 @@ static void cpu_pre_save(void *opaque)
             /* This should never fail */
             abort();
         }
+#if defined CONFIG_KVM
+        if (kvm_check_extension(CPU(cpu)->kvm_state, KVM_CAP_MP_STATE)) {
+            struct kvm_mp_state mp_state;
+            int ret = kvm_vcpu_ioctl(CPU(cpu), KVM_GET_MP_STATE, &mp_state);
+            if (ret) {
+                fprintf(stderr, "%s: failed to get MP_STATE %d/%s\n",
+                        __func__, ret, strerror(ret));
+                abort();
+            }
+            cpu->powered_off =
+                (mp_state.mp_state == KVM_MP_STATE_RUNNABLE)
+                ? false : true;

Ternary operator to produce a true-or-false result is a bit
redundant...

quoted

+        }
+#endif

Why is this in pre-save/post-load rather than in the
kvm_arch_get/put_registers functions like all the other
syncing code?

Yeah the #ifdefs should have waved the red flag - I'll move it ;-)

-- PMM

-- 
Alex Benn?e