On Wed, Mar 19, 2014 at 3:54 PM, Christoph Hellwig <h...@infradead.org> wrote: > On Wed, Mar 19, 2014 at 11:08:20AM +0800, Ming Lei wrote: >> > + blk_rq_init(q, q->flush_rq); >> > + if (reg->cmd_size) >> > + q->flush_rq->special = >> > + blk_mq_rq_to_pdu(q->flush_rq); >> > + >> > + if (reg->ops->init_request(driver_data, >> > + NULL, q->flush_rq, -1)) >> > + goto err_flush_rq; >> > + } >> >> The above looks a bit weird because q->flush_rq is invisible to >> driver and should always be initialized no matter if driver defines >> its .init_request callback or not. > > You mean the blk_rq_init? We already do a real initialization before > actually using it, it's just there to prevent passing a half-initialized > one to the driver. >
OK. Actually it doesn't matter because flush_rq will be reinitialized before using, not like common request. And driver should only initialize req->special instead of the request itself, maybe it is better to document the fact. Thanks, -- Ming Lei -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
