On 4/17/21 10:49 PM, Chaitanya Kulkarni wrote:
> On 4/17/21 08:30, Jens Axboe wrote:
>> + cmd->error = null_process_cmd(cmd, req_op(req), blk_rq_pos(req),
>> + blk_rq_sectors(req));
>
> How about following on the top of this patch ?
>
> diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c
> index 8efaf21cc053..4c27e37ccc51 100644
> --- a/drivers/block/null_blk/main.c
> +++ b/drivers/block/null_blk/main.c
> @@ -1496,6 +1496,7 @@ static int null_map_queues(struct blk_mq_tag_set *set)
> static int null_poll(struct blk_mq_hw_ctx *hctx)
> {
> struct nullb_queue *nq = hctx->driver_data;
> + blk_status_t sts;
> LIST_HEAD(list);
> int nr = 0;
>
> @@ -1510,8 +1511,16 @@ static int null_poll(struct blk_mq_hw_ctx *hctx)
> req = list_first_entry(&list, struct request, queuelist);
> list_del_init(&req->queuelist);
> cmd = blk_mq_rq_to_pdu(req);
> - cmd->error = null_process_cmd(cmd, req_op(req),
> blk_rq_pos(req),
> - blk_rq_sectors(req));
> + if (cmd->nq->dev->zoned)
> + sts = null_process_zoned_cmd(cmd, req_op(req),
> + blk_rq_pos(req),
> + blk_rq_sectors(req));
> + else
> + sts = null_process_cmd(cmd, req_op(req),
> blk_rq_pos(req),
> + blk_rq_sectors(req));
> +
> + cmd->error = sts;
> +
> nullb_complete_cmd(cmd);
> nr++;
> }
>
> If you are okay I can send a well tested patch with little bit code
> cleanup once this is in the tree.
Yes, that might be a good idea. I'll just fold it in, I've got it
sitting separately so far. Just let me know when you've tested it.
--
Jens Axboe
