Re: [PATCH 2/6] blk-mq: replace timeout synchronization with a RCU and generation based scheme

[Bart Van Assche]

On Tue, 2017-12-12 at 11:01 -0800, Tejun Heo wrote:
> +/*
> + * Bits for request->gstate.  The lower two bits carry MQ_RQ_* state value
> + * and the upper bits the generation number.
> + */
> +enum mq_rq_state {
> +	MQ_RQ_IDLE		= 0,
> +	MQ_RQ_IN_FLIGHT		= 1,
> +
> +	MQ_RQ_STATE_BITS	= 2,
> +	MQ_RQ_STATE_MASK	= (1 << MQ_RQ_STATE_BITS) - 1,
> +	MQ_RQ_GEN_INC		= 1 << MQ_RQ_STATE_BITS,
> +};
> +
> @@ -85,6 +98,38 @@ extern void blk_mq_rq_timed_out(struct request *req, bool reserved);
> +/**
> + * blk_mq_rq_state() - read the current MQ_RQ_* state of a request
> + * @rq: target request.
> + */
> +static inline int blk_mq_rq_state(struct request *rq)
> +{
> +	return READ_ONCE(rq->gstate) & MQ_RQ_STATE_MASK;
> +}
> +
> +/**
> + * blk_mq_rq_update_state() - set the current MQ_RQ_* state of a request
> + * @rq: target request.
> + * @state: new state to set.
> + *
> + * Set @rq's state to @state.  The caller is responsible for ensuring that
> + * there are no other updaters.  A request can transition into IN_FLIGHT
> + * only from IDLE and doing so increments the generation number.
> + */
> +static inline void blk_mq_rq_update_state(struct request *rq,
> +					  enum mq_rq_state state)
> +{
> +	u64 new_val = (READ_ONCE(rq->gstate) & ~MQ_RQ_STATE_MASK) | state;
> +
> +	if (state == MQ_RQ_IN_FLIGHT) {
> +		WARN_ON_ONCE(blk_mq_rq_state(rq) != MQ_RQ_IDLE);
> +		new_val += MQ_RQ_GEN_INC;
> +	}
> +
> +	/* avoid exposing interim values */
> +	WRITE_ONCE(rq->gstate, new_val);
> +}

Hello Tejun,

Have you considered the following instead of introducing MQ_RQ_IDLE and
MQ_RQ_IN_FLIGHT? I think this could help to limit the number of new atomic
operations introduced in the hot path by this patch series.

static inline bool blk_mq_rq_in_flight(struct request *rq)
{
	return list_empty(&rq->queuelist);
}

Thanks,

Bart.