Re: [PATCH 07/18] dm: always defer request allocation to the owner of the request_queue

2017-01-27 Thread Mike Snitzer
On Fri, Jan 27 2017 at 11:36am -0500, Christoph Hellwig wrote: > On Fri, Jan 27, 2017 at 11:34:34AM -0500, Mike Snitzer wrote: > > Noticed after further review that it seems a bit weird to have the non > > blk-mq support in drivers calling blk_mq_rq_to_pdu(). But I'm not sure > > a

Re: [PATCH 07/18] dm: always defer request allocation to the owner of the request_queue

2017-01-27 Thread Christoph Hellwig
On Fri, Jan 27, 2017 at 11:34:34AM -0500, Mike Snitzer wrote: > Noticed after further review that it seems a bit weird to have the non > blk-mq support in drivers calling blk_mq_rq_to_pdu(). But I'm not sure > a blk_rq_to_pdu() macro to blk_mq_rq_to_pdu() is the right thing. What > do you guys

Re: [PATCH 07/18] dm: always defer request allocation to the owner of the request_queue

2017-01-27 Thread Mike Snitzer
On Wed, Jan 25 2017 at 12:25pm -0500, Christoph Hellwig wrote: > DM already calls blk_mq_alloc_request on the request_queue of the > underlying device if it is a blk-mq device. But now that we allow drivers > to allocate additional data and initialize it ahead of time we need to do

[PATCH 07/18] dm: always defer request allocation to the owner of the request_queue

2017-01-25 Thread Christoph Hellwig
DM already calls blk_mq_alloc_request on the request_queue of the underlying device if it is a blk-mq device. But now that we allow drivers to allocate additional data and initialize it ahead of time we need to do the same for all drivers. Doing so and using the new cmd_size infrastructure in