Re: [PATCH V2 2/2] mmc: hsq: dynamic adjustment of hsq->depth

From: Ulf Hansson
Date: Thu Aug 24 2023 - 06:38:15 EST


On Wed, 23 Aug 2023 at 08:18, Wenchao Chen <wenchao.chen@xxxxxxxxxx> wrote:
>
> Increasing hsq_depth improves random write performance.
>
> Signed-off-by: Wenchao Chen <wenchao.chen@xxxxxxxxxx>
> ---
> drivers/mmc/host/mmc_hsq.c | 26 ++++++++++++++++++++++++++
> drivers/mmc/host/mmc_hsq.h | 2 ++
> 2 files changed, 28 insertions(+)
>
> diff --git a/drivers/mmc/host/mmc_hsq.c b/drivers/mmc/host/mmc_hsq.c
> index 8556cacb21a1..8682a3d16a76 100644
> --- a/drivers/mmc/host/mmc_hsq.c
> +++ b/drivers/mmc/host/mmc_hsq.c
> @@ -21,6 +21,30 @@ static void mmc_hsq_retry_handler(struct work_struct *work)
> mmc->ops->request(mmc, hsq->mrq);
> }
>
> +static void mmc_hsq_modify_threshold(struct mmc_hsq *hsq)
> +{
> + struct mmc_host *mmc = hsq->mmc;
> + struct mmc_request *mrq;
> + struct hsq_slot *slot;
> + int need_change = 0;
> + int tag;
> +
> + for (tag = 0; tag < HSQ_NUM_SLOTS; tag++) {
> + slot = &hsq->slot[tag];
> + mrq = slot->mrq;
> + if (mrq && mrq->data && (mrq->data->blocks == HSQ_DATA_IS_4K)

This assumes mrq->data->blksz is 512 (which at least for now is always
the case), but perhaps better to compute the request size instead?
Hence:

"mrq->data->blksz * mrq->data->blocks == 4096"

> + && (mrq->data->flags & MMC_DATA_WRITE))
> + need_change++;
> + else
> + break;
> + }
> +
> + if (need_change > 1)
> + mmc->hsq_depth = HSQ_PERFORMANCE_DEPTH;
> + else
> + mmc->hsq_depth = HSQ_NORMAL_DEPTH;
> +}
> +
> static void mmc_hsq_pump_requests(struct mmc_hsq *hsq)
> {
> struct mmc_host *mmc = hsq->mmc;
> @@ -42,6 +66,8 @@ static void mmc_hsq_pump_requests(struct mmc_hsq *hsq)
> return;
> }
>
> + mmc_hsq_modify_threshold(hsq);
> +
> slot = &hsq->slot[hsq->next_tag];
> hsq->mrq = slot->mrq;
> hsq->qcnt--;
> diff --git a/drivers/mmc/host/mmc_hsq.h b/drivers/mmc/host/mmc_hsq.h
> index aa5c4543b55f..fc031e38f1e0 100644
> --- a/drivers/mmc/host/mmc_hsq.h
> +++ b/drivers/mmc/host/mmc_hsq.h
> @@ -10,6 +10,8 @@
> * flight to avoid a long latency.
> */
> #define HSQ_NORMAL_DEPTH 2
> +#define HSQ_PERFORMANCE_DEPTH 5
> +#define HSQ_DATA_IS_4K 8

Perhaps re-phrase the comment a few lines above to explain why/when
'5' can be good too.

>
> struct hsq_slot {
> struct mmc_request *mrq;

Kind regards
Uffe