target-arm: kvm: use AddressSpace-specific listener

The only address space where the GIC devices are added is
address_space_memory.  There is no need to use a global
MemoryListener.

This removes the only user of global MemoryListeners.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 target-arm/kvm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

On 30 September 2016 at 08:17, Paolo Bonzini <pbonzini@redhat.com> wrote:
> The only address space where the GIC devices are added is
> address_space_memory.  There is no need to use a global
> MemoryListener.
>
> This removes the only user of global MemoryListeners.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  target-arm/kvm.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/target-arm/kvm.c b/target-arm/kvm.c
> index dbe393c..41f2038 100644
> --- a/target-arm/kvm.c
> +++ b/target-arm/kvm.c
> @@ -283,7 +283,7 @@ void kvm_arm_register_device(MemoryRegion *mr, uint64_t devid, uint64_t group,
>      }
>
>      if (QSLIST_EMPTY(&kvm_devices_head)) {
> -        memory_listener_register(&devlistener, NULL);
> +        memory_listener_register(&devlistener, &address_space_memory);
>          qemu_add_machine_init_done_notifier(&notify);
>      }
>      kd = g_new0(KVMDevice, 1);

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

The code effectively assumes there's only one address
space the devices will live in anyway...

thanks
-- PMM

On 6 October 2016 at 17:02, Peter Maydell <peter.maydell@linaro.org> wrote:
> On 30 September 2016 at 08:17, Paolo Bonzini <pbonzini@redhat.com> wrote:
>> The only address space where the GIC devices are added is
>> address_space_memory.  There is no need to use a global
>> MemoryListener.
>>
>> This removes the only user of global MemoryListeners.
>>
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>> ---
>>  target-arm/kvm.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/target-arm/kvm.c b/target-arm/kvm.c
>> index dbe393c..41f2038 100644
>> --- a/target-arm/kvm.c
>> +++ b/target-arm/kvm.c
>> @@ -283,7 +283,7 @@ void kvm_arm_register_device(MemoryRegion *mr, uint64_t devid, uint64_t group,
>>      }
>>
>>      if (QSLIST_EMPTY(&kvm_devices_head)) {
>> -        memory_listener_register(&devlistener, NULL);
>> +        memory_listener_register(&devlistener, &address_space_memory);
>>          qemu_add_machine_init_done_notifier(&notify);
>>      }
>>      kd = g_new0(KVMDevice, 1);
>
> Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
>
> The code effectively assumes there's only one address
> space the devices will live in anyway...



Applied to target-arm.next, thanks.

-- PMM

On 30 September 2016 at 08:17, Paolo Bonzini <pbonzini@redhat.com> wrote:
> The only address space where the GIC devices are added is
> address_space_memory.  There is no need to use a global
> MemoryListener.
>
> This removes the only user of global MemoryListeners.
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  target-arm/kvm.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/target-arm/kvm.c b/target-arm/kvm.c
> index dbe393c..41f2038 100644
> --- a/target-arm/kvm.c
> +++ b/target-arm/kvm.c
> @@ -283,7 +283,7 @@ void kvm_arm_register_device(MemoryRegion *mr, uint64_t devid, uint64_t group,
>      }
>
>      if (QSLIST_EMPTY(&kvm_devices_head)) {
> -        memory_listener_register(&devlistener, NULL);
> +        memory_listener_register(&devlistener, &address_space_memory);
>          qemu_add_machine_init_done_notifier(&notify);
>      }
>      kd = g_new0(KVMDevice, 1);
> --
> 2.7.4

Belatedly noticed this doesn't build on ARM hosts because
there's no #include "exec/address-spaces.h". Will fix up
in target-arm.next tree.

thanks
-- PMM