Re: [PATCH] xen-blkfront: check feature-persitent on resume

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

On 23/05/14 19:51, Konrad Rzeszutek Wilk wrote:
> On Thu, May 22, 2014 at 04:40:07PM +0200, Roger Pau Monne wrote:
>> We are missing a check to see if the backend supports persistent
>> grants on resume, meaning we will always run with the value fetched
>> from the firsts host on which we run on.
>>
>> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
>> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
>> Cc: David Vrabel <david.vrabel@citrix.com>
>> ---
>>  drivers/block/xen-blkfront.c |   23 ++++++++++++++++-------
>>  1 files changed, 16 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
>> index efe1b47..00ae36b 100644
>> --- a/drivers/block/xen-blkfront.c
>> +++ b/drivers/block/xen-blkfront.c
>> @@ -1430,6 +1430,20 @@ static void split_bio_end(struct bio *bio, int error)
>>  	bio_put(bio);
>>  }
>>  
>> +static void blkfront_setup_persistent(struct blkfront_info *info)
>> +{
>> +	unsigned int persistent;
>> +	int err;
>> +
>> +	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
>> +			    "feature-persistent", "%u", &persistent,
>> +			    NULL);
>> +	if (err)
>> +		info->feature_persistent = 0;
>> +	else
>> +		info->feature_persistent = persistent;
>> +}
>> +
>>  static int blkif_recover(struct blkfront_info *info)
>>  {
>>  	int i;
>> @@ -1456,6 +1470,7 @@ static int blkif_recover(struct blkfront_info *info)
>>  	info->shadow_free = info->ring.req_prod_pvt;
>>  	info->shadow[BLK_RING_SIZE-1].req.u.rw.id = 0x0fffffff;
>>  
>> +	blkfront_setup_persistent(info);
>>  	rc = blkfront_setup_indirect(info);
>>  	if (rc) {
>>  		kfree(copy);
>> @@ -1851,13 +1866,7 @@ static void blkfront_connect(struct blkfront_info *info)
>>  	if (!err && discard)
>>  		blkfront_setup_discard(info);
>>  
>> -	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
>> -			    "feature-persistent", "%u", &persistent,
>> -			    NULL);
>> -	if (err)
>> -		info->feature_persistent = 0;
>> -	else
>> -		info->feature_persistent = persistent;
>> +	blkfront_setup_persistent(info);
> 
> Should we also flush the ones that are on the info->grants list if
> we turning it off?

This path is only taken on the first connection, further calls to
blkfront_connect from resume code don't get here, because they take the
BLKIF_STATE_SUSPENDED path and just call blkif_recover and return. And
on disconnection blkfront_connect doesn't even get called.

Roger.