[PATCH for-next v3 3/4] block: add helper to map bvec iterator for passthrough

Jens Axboe axboe at kernel.dk
Fri Sep 2 16:14:41 PDT 2022


On 9/2/22 9:16 AM, Kanchan Joshi wrote:
> Add blk_rq_map_user_bvec which maps the bvec iterator into a bio and
> places that into the request.
> This helper is to be used in nvme for uring-passthrough with
> fixed-buffer.
> 
> Signed-off-by: Kanchan Joshi <joshi.k at samsung.com>
> Signed-off-by: Anuj Gupta <anuj20.g at samsung.com>
> ---
>  block/blk-map.c        | 71 ++++++++++++++++++++++++++++++++++++++++++
>  include/linux/blk-mq.h |  1 +
>  2 files changed, 72 insertions(+)
> 
> diff --git a/block/blk-map.c b/block/blk-map.c
> index f3768876d618..0f7dc568e34b 100644
> --- a/block/blk-map.c
> +++ b/block/blk-map.c
> @@ -612,6 +612,77 @@ int blk_rq_map_user(struct request_queue *q, struct request *rq,
>  }
>  EXPORT_SYMBOL(blk_rq_map_user);
>  
> +/* Prepare bio for passthrough IO given an existing bvec iter */
> +int blk_rq_map_user_bvec(struct request *rq, struct iov_iter *iter)
> +{
> +	struct request_queue *q = rq->q;
> +	size_t iter_count, nr_segs;
> +	struct bio *bio;
> +	struct bio_vec *bv, *bvec_arr, *bvprvp = NULL;
> +	struct queue_limits *lim = &q->limits;
> +	unsigned int nsegs = 0, bytes = 0;
> +	int ret, i;
> +
> +	iter_count = iov_iter_count(iter);
> +	nr_segs = iter->nr_segs;
> +
> +	if (!iter_count || (iter_count >> 9) > queue_max_hw_sectors(q))
> +		return -EINVAL;
> +	if (nr_segs > queue_max_segments(q))
> +		return -EINVAL;
> +	if (rq->cmd_flags & REQ_POLLED) {
> +		blk_opf_t opf = rq->cmd_flags | REQ_ALLOC_CACHE;
> +
> +		/* no iovecs to alloc, as we already have a BVEC iterator */
> +		bio = bio_alloc_bioset(NULL, 0, opf, GFP_KERNEL,
> +					&fs_bio_set);
> +		if (!bio)
> +			return -ENOMEM;
> +	} else {
> +		bio = bio_kmalloc(0, GFP_KERNEL);
> +		if (!bio)
> +			return -ENOMEM;
> +		bio_init(bio, NULL, bio->bi_inline_vecs, 0, req_op(rq));
> +	}

I think this should be a helper at this point, as it's the same
duplicated code we have in the normal map path.

-- 
Jens Axboe



More information about the Linux-nvme mailing list