Re: [Xen-devel] [PATCH v4 2/2] xen/block: add multi-page ring support

["Roger Pau Monné" <roger.pau@citrix.com>]

El 20/05/15 a les 15.21, Julien Grall ha escrit:
> Hi,
> 
> On 20/05/15 14:10, Bob Liu wrote:
>> ---
>>  drivers/block/xen-blkback/blkback.c |  12 ++++
>>  drivers/block/xen-blkback/common.h  |   3 +-
>>  drivers/block/xen-blkback/xenbus.c  |  85 +++++++++++++++++++++-------
>>  drivers/block/xen-blkfront.c        | 110 ++++++++++++++++++++++++++----------
>>  4 files changed, 161 insertions(+), 49 deletions(-)
>>
>> diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c
>> index 713fc9f..057890f 100644
>> --- a/drivers/block/xen-blkback/blkback.c
>> +++ b/drivers/block/xen-blkback/blkback.c
>> @@ -84,6 +84,12 @@ MODULE_PARM_DESC(max_persistent_grants,
>>                   "Maximum number of grants to map persistently");
>>  
>>  /*
>> + * Maximum number of pages to be used as the ring between front and backend
>> + */
>> +unsigned int xen_blkif_max_ring_order = XENBUS_MAX_RING_PAGE_ORDER;
> 
> We will soon support 64KB page granularity with ARM64, although the PV
> protocol will keep a 4KB page granularity.
> 
> Can you clarify with granularity is used here? The one of the host or
> the one of the PV protocol?

It's using 4K pages, because those are then granted to the domain
handling the backend.

Roger.