On 04/23/2010 02:14 PM, Takuya Yoshikawa wrote:
Do you have performance numbers? I'm interested in both measurements of
KVM_SWITCH_DIRTY_LOG under various conditions and macro benchmarks (for
example, total guest throughput improvement under Kemari).
Currently, I'm just checking the performa
2010/4/23 Takuya Yoshikawa :
> (2010/04/23 19:28), Avi Kivity wrote:
>
>>>
>>> OK, I will do in the next version. In this RFC, I would be happy if I can
>>> know the overall design is right or not.
>>>
>>
>> Everything looks reasonable to me.
>>
>
> Thank you!
>
>> Do you have performance numbers?
(2010/04/23 19:28), Avi Kivity wrote:
OK, I will do in the next version. In this RFC, I would be happy if I can
know the overall design is right or not.
Everything looks reasonable to me.
Thank you!
Do you have performance numbers? I'm interested in both measurements of
KVM_SWITCH_DIRTY
On 04/22/2010 12:07 PM, Takuya Yoshikawa wrote:
+ slots->memslots[i].dirty_bitmap = NULL;
+ slots->memslots[i].dirty_bitmap_old = NULL;
kvm_free_physmem_slot(&slots->memslots[i], NULL);
+ }
+/*
+ * Please use generic *_user bitops once they become available.
+ * Be careful setting the bit w
(2010/04/21 20:26), Avi Kivity wrote:
r = 0;
@@ -1858,7 +1866,7 @@ int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
if (memslot->is_dirty) {
kvm_flush_remote_tlbs(kvm);
n = kvm_dirty_bitmap_bytes(memslot);
- memset(memslot->dirty_bitmap, 0, n);
+ clear_user(memslot->dirty_bitmap, n);
memslot->is_
On 04/20/2010 02:02 PM, Takuya Yoshikawa wrote:
We move dirty bitmaps to user space.
- Allocation and destruction: we use do_mmap() and do_munmap().
The new bitmap space is twice longer than the original one and we
use the additional space for double buffering: this makes it
possib
(2010/04/20 20:10), Alexander Graf wrote:
On 20.04.2010, at 13:02, Takuya Yoshikawa wrote:
We move dirty bitmaps to user space.
- Allocation and destruction: we use do_mmap() and do_munmap().
The new bitmap space is twice longer than the original one and we
use the additional space for
On 20.04.2010, at 13:02, Takuya Yoshikawa wrote:
> We move dirty bitmaps to user space.
>
> - Allocation and destruction: we use do_mmap() and do_munmap().
> The new bitmap space is twice longer than the original one and we
> use the additional space for double buffering: this makes it
> p
We move dirty bitmaps to user space.
- Allocation and destruction: we use do_mmap() and do_munmap().
The new bitmap space is twice longer than the original one and we
use the additional space for double buffering: this makes it
possible to update the active bitmap while letting the user
