Paolo Bonzini wrote:
> This makes it possible to do blocking writes directly to the socket,
> with no buffer in the middle. For RAM, only the migration_bitmap_sync()
> call needs the iothread lock. For block migration, it is needed by
> the block layer (including bdrv_drain_all and dirty bitmap
Il 19/02/2013 13:34, Orit Wasserman ha scritto:
> Looks good but locking is complicated ..
You mean as a fact of life :) or this locking in particular?
Paolo
> Reviewed-by: Orit Wasserman
>
>
On 02/19/2013 02:43 PM, Paolo Bonzini wrote:
> Il 19/02/2013 13:34, Orit Wasserman ha scritto:
>> Looks good but locking is complicated ..
>
> You mean as a fact of life :) or this locking in particular?
both :)
>
> Paolo
>
>> Reviewed-by: Orit Wasserman
>>
>>
>
On 02/15/2013 07:46 PM, Paolo Bonzini wrote:
> This makes it possible to do blocking writes directly to the socket,
> with no buffer in the middle. For RAM, only the migration_bitmap_sync()
> call needs the iothread lock. For block migration, it is needed by
> the block layer (including bdrv_drai
This makes it possible to do blocking writes directly to the socket,
with no buffer in the middle. For RAM, only the migration_bitmap_sync()
call needs the iothread lock. For block migration, it is needed by
the block layer (including bdrv_drain_all and dirty bitmap access),
but because some code
