public inbox for linux-mmc@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue
@ 2016-07-26 17:01 Bhaktipriya Shridhar
  2016-07-26 17:02 ` Tejun Heo
  2016-07-27 12:47 ` Ulf Hansson
  0 siblings, 2 replies; 3+ messages in thread
From: Bhaktipriya Shridhar @ 2016-07-26 17:01 UTC (permalink / raw)
  To: Ulf Hansson; +Cc: linux-mmc, linux-kernel, Tejun Heo

The workqueue "workq" provides support for sd/mmc async request, which
makes next request do dma_map_sg() while previous request transferring
data.

The workqueue has a single workitem(&host->work) and hence doesn't require
ordering. Also, it is not being used on a memory reclaim path. Hence,
the singlethreaded workqueue has been replaced with the use of system_wq.

System workqueues have been able to handle high level of concurrency
for a long time now and hence it's not required to have a singlethreaded
workqueue just to gain concurrency. Unlike a dedicated per-cpu workqueue
created with create_singlethread_workqueue(), system_wq allows multiple
work items to overlap executions even on the same CPU; however, a
per-cpu workqueue doesn't have any CPU locality or global ordering
guarantee unless the target CPU is explicitly specified and thus the
increase of local concurrency shouldn't make any difference.

Work item has been flushed in rtsx_pci_sdmmc_drv_remove() to ensure that
there are no pending tasks while disconnecting the driver.

Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>
---
 drivers/mmc/host/rtsx_pci_sdmmc.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/drivers/mmc/host/rtsx_pci_sdmmc.c b/drivers/mmc/host/rtsx_pci_sdmmc.c
index 9313748..49516dc 100644
--- a/drivers/mmc/host/rtsx_pci_sdmmc.c
+++ b/drivers/mmc/host/rtsx_pci_sdmmc.c
@@ -38,7 +38,6 @@ struct realtek_pci_sdmmc {
 	struct rtsx_pcr		*pcr;
 	struct mmc_host		*mmc;
 	struct mmc_request	*mrq;
-	struct workqueue_struct *workq;
 #define SDMMC_WORKQ_NAME	"rtsx_pci_sdmmc_workq"

 	struct work_struct	work;
@@ -885,7 +884,7 @@ static void sdmmc_request(struct mmc_host *mmc, struct mmc_request *mrq)
 	if (sd_rw_cmd(mrq->cmd) || sdio_extblock_cmd(mrq->cmd, data))
 		host->using_cookie = sd_pre_dma_transfer(host, data, false);

-	queue_work(host->workq, &host->work);
+	schedule_work(&host->work);
 }

 static int sd_set_bus_width(struct realtek_pci_sdmmc *host,
@@ -1404,11 +1403,6 @@ static int rtsx_pci_sdmmc_drv_probe(struct platform_device *pdev)
 		return -ENOMEM;

 	host = mmc_priv(mmc);
-	host->workq = create_singlethread_workqueue(SDMMC_WORKQ_NAME);
-	if (!host->workq) {
-		mmc_free_host(mmc);
-		return -ENOMEM;
-	}
 	host->pcr = pcr;
 	host->mmc = mmc;
 	host->pdev = pdev;
@@ -1462,9 +1456,7 @@ static int rtsx_pci_sdmmc_drv_remove(struct platform_device *pdev)
 	mmc_remove_host(mmc);
 	host->eject = true;

-	flush_workqueue(host->workq);
-	destroy_workqueue(host->workq);
-	host->workq = NULL;
+	flush_work(&host->work);

 	mmc_free_host(mmc);

--
2.1.4


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue
  2016-07-26 17:01 [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue Bhaktipriya Shridhar
@ 2016-07-26 17:02 ` Tejun Heo
  2016-07-27 12:47 ` Ulf Hansson
  1 sibling, 0 replies; 3+ messages in thread
From: Tejun Heo @ 2016-07-26 17:02 UTC (permalink / raw)
  To: Bhaktipriya Shridhar; +Cc: Ulf Hansson, linux-mmc, linux-kernel

On Tue, Jul 26, 2016 at 10:31:06PM +0530, Bhaktipriya Shridhar wrote:
> The workqueue "workq" provides support for sd/mmc async request, which
> makes next request do dma_map_sg() while previous request transferring
> data.
> 
> The workqueue has a single workitem(&host->work) and hence doesn't require
> ordering. Also, it is not being used on a memory reclaim path. Hence,
> the singlethreaded workqueue has been replaced with the use of system_wq.
> 
> System workqueues have been able to handle high level of concurrency
> for a long time now and hence it's not required to have a singlethreaded
> workqueue just to gain concurrency. Unlike a dedicated per-cpu workqueue
> created with create_singlethread_workqueue(), system_wq allows multiple
> work items to overlap executions even on the same CPU; however, a
> per-cpu workqueue doesn't have any CPU locality or global ordering
> guarantee unless the target CPU is explicitly specified and thus the
> increase of local concurrency shouldn't make any difference.
> 
> Work item has been flushed in rtsx_pci_sdmmc_drv_remove() to ensure that
> there are no pending tasks while disconnecting the driver.
> 
> Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>

Acked-by: Tejun Heo <tj@kernel.org>

Thanks.

-- 
tejun

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue
  2016-07-26 17:01 [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue Bhaktipriya Shridhar
  2016-07-26 17:02 ` Tejun Heo
@ 2016-07-27 12:47 ` Ulf Hansson
  1 sibling, 0 replies; 3+ messages in thread
From: Ulf Hansson @ 2016-07-27 12:47 UTC (permalink / raw)
  To: Bhaktipriya Shridhar; +Cc: linux-mmc, linux-kernel@vger.kernel.org, Tejun Heo

On 26 July 2016 at 19:01, Bhaktipriya Shridhar <bhaktipriya96@gmail.com> wrote:
> The workqueue "workq" provides support for sd/mmc async request, which
> makes next request do dma_map_sg() while previous request transferring
> data.
>
> The workqueue has a single workitem(&host->work) and hence doesn't require
> ordering. Also, it is not being used on a memory reclaim path. Hence,
> the singlethreaded workqueue has been replaced with the use of system_wq.
>
> System workqueues have been able to handle high level of concurrency
> for a long time now and hence it's not required to have a singlethreaded
> workqueue just to gain concurrency. Unlike a dedicated per-cpu workqueue
> created with create_singlethread_workqueue(), system_wq allows multiple
> work items to overlap executions even on the same CPU; however, a
> per-cpu workqueue doesn't have any CPU locality or global ordering
> guarantee unless the target CPU is explicitly specified and thus the
> increase of local concurrency shouldn't make any difference.
>
> Work item has been flushed in rtsx_pci_sdmmc_drv_remove() to ensure that
> there are no pending tasks while disconnecting the driver.
>
> Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>

Thanks, applied for next!

Kind regards
Uffe

> ---
>  drivers/mmc/host/rtsx_pci_sdmmc.c | 12 ++----------
>  1 file changed, 2 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/mmc/host/rtsx_pci_sdmmc.c b/drivers/mmc/host/rtsx_pci_sdmmc.c
> index 9313748..49516dc 100644
> --- a/drivers/mmc/host/rtsx_pci_sdmmc.c
> +++ b/drivers/mmc/host/rtsx_pci_sdmmc.c
> @@ -38,7 +38,6 @@ struct realtek_pci_sdmmc {
>         struct rtsx_pcr         *pcr;
>         struct mmc_host         *mmc;
>         struct mmc_request      *mrq;
> -       struct workqueue_struct *workq;
>  #define SDMMC_WORKQ_NAME       "rtsx_pci_sdmmc_workq"
>
>         struct work_struct      work;
> @@ -885,7 +884,7 @@ static void sdmmc_request(struct mmc_host *mmc, struct mmc_request *mrq)
>         if (sd_rw_cmd(mrq->cmd) || sdio_extblock_cmd(mrq->cmd, data))
>                 host->using_cookie = sd_pre_dma_transfer(host, data, false);
>
> -       queue_work(host->workq, &host->work);
> +       schedule_work(&host->work);
>  }
>
>  static int sd_set_bus_width(struct realtek_pci_sdmmc *host,
> @@ -1404,11 +1403,6 @@ static int rtsx_pci_sdmmc_drv_probe(struct platform_device *pdev)
>                 return -ENOMEM;
>
>         host = mmc_priv(mmc);
> -       host->workq = create_singlethread_workqueue(SDMMC_WORKQ_NAME);
> -       if (!host->workq) {
> -               mmc_free_host(mmc);
> -               return -ENOMEM;
> -       }
>         host->pcr = pcr;
>         host->mmc = mmc;
>         host->pdev = pdev;
> @@ -1462,9 +1456,7 @@ static int rtsx_pci_sdmmc_drv_remove(struct platform_device *pdev)
>         mmc_remove_host(mmc);
>         host->eject = true;
>
> -       flush_workqueue(host->workq);
> -       destroy_workqueue(host->workq);
> -       host->workq = NULL;
> +       flush_work(&host->work);
>
>         mmc_free_host(mmc);
>
> --
> 2.1.4
>

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2016-07-27 12:47 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-07-26 17:01 [PATCH] mmc: rtsx: Remove deprecated create_singlethread_workqueue Bhaktipriya Shridhar
2016-07-26 17:02 ` Tejun Heo
2016-07-27 12:47 ` Ulf Hansson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox