Re: [Qemu-devel] [PATCH v2 04/11] memory: add getter for owner

[Paolo Bonzini <pbonzini@redhat.com>]

Il 01/07/2013 16:24, Jan Kiszka ha scritto:
> On 2013-06-28 18:58, Paolo Bonzini wrote:
>> Whenever memory regions are accessed outside the BQL, they need to be
>> preserved against hot-unplug.  MemoryRegions actually do not have their
>> own reference count; they piggyback on a QOM object, their "owner".
>> The owner is set at creation time, and there is a function to retrieve
>> the owner.
>>
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>> ---
>>  include/exec/memory.h | 7 +++++++
>>  memory.c              | 6 ++++++
>>  2 files changed, 13 insertions(+)
>>
>> diff --git a/include/exec/memory.h b/include/exec/memory.h
>> index be3d39f..1ad9c19 100644
>> --- a/include/exec/memory.h
>> +++ b/include/exec/memory.h
>> @@ -380,6 +380,13 @@ void memory_region_init_iommu(MemoryRegion *mr,
>>  void memory_region_destroy(MemoryRegion *mr);
>>  
>>  /**
>> + * memory_region_owner: get a memory region's owner.
>> + *
>> + * @mr: the memory region being queried.
>> + */
>> +struct Object *memory_region_owner(MemoryRegion *mr);
>> +
>> +/**
>>   * memory_region_size: get a memory region's size.
>>   *
>>   * @mr: the memory region being queried.
>> diff --git a/memory.c b/memory.c
>> index 6250bec1..4d396c3 100644
>> --- a/memory.c
>> +++ b/memory.c
>> @@ -738,6 +738,7 @@ void memory_region_init(MemoryRegion *mr,
>>      mr->owner = owner;
>>      mr->iommu_ops = NULL;
>>      mr->parent = NULL;
>> +    mr->owner = NULL;
>>      mr->size = int128_make64(size);
>>      if (size == UINT64_MAX) {
>>          mr->size = int128_2_64();
>> @@ -1011,6 +1012,11 @@ void memory_region_destroy(MemoryRegion *mr)
>>      g_free(mr->ioeventfds);
>>  }
>>  
>> +Object *memory_region_owner(MemoryRegion *mr)
>> +{
>> +    return mr->owner;
>> +}
>> +
>>  uint64_t memory_region_size(MemoryRegion *mr)
>>  {
>>      if (int128_eq(mr->size, int128_2_64())) {
>>
> 
> OK, fine... and who is using this? ;)

There'll be a couple users in hw/acpi/core.c.

> Also, as this is trivial and the structure publicly known anyway, I'd go
> for a static inline function instead.

Ok.  Though it's not a fast path anyway.

Paolo