Re: [PATCH 26/30] mmc: update workqueue usages

Re: [PATCH 26/30] mmc: update workqueue usages

[Chris Ball]

Hi linux-omap folks,

I'd like to merge this patch via the MMC tree; please review/ACK.

Thanks,

- Chris.

From: Tejun Heo <tj@kernel.org>

Workqueue creation API has been updated and flush_scheduled_work() is
deprecated and scheduled to be removed.

* core/core.c: Use alloc_ordered_workqueue() instead of
  create_singlethread_workqueue().  This removes an unnecessary
  rescuer.

* host/omap.c: Create, use and flush mmc_omap_wq instead of the
  system_wq.

* Flush host->mmc_carddetect_work directly on removal instead of using
  flush_scheduled_work().

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Chris Ball <cjb@laptop.org>
Cc: linux-mmc@vger.kernel.org
---
This is part of a series to remove flush_scheduled_work() usage to
prepare for deprecation of flush_scheduled_work().  Patches in this
series are self contained and mostly straight-forward.

Please feel free to take it into the appropriate tree, or just ack it.
In the latter case, I'll merge the patch through the workqueue tree
during the next merge window.

If you're seeing this patch for the second time, it's because the
commit hasn't showed up in mainline yet.  Please let me know what
should be done.

Thank you.

 drivers/mmc/core/core.c       |    2 +-
 drivers/mmc/host/omap.c       |   24 ++++++++++++++++++------
 drivers/mmc/host/omap_hsmmc.c |    2 +-
 3 files changed, 20 insertions(+), 8 deletions(-)

diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
index 31ae07a..c76f492 100644
--- a/drivers/mmc/core/core.c
+++ b/drivers/mmc/core/core.c
@@ -1789,7 +1789,7 @@ static int __init mmc_init(void)
 {
 	int ret;
 
-	workqueue = create_singlethread_workqueue("kmmcd");
+	workqueue = alloc_ordered_workqueue("kmmcd", 0);
 	if (!workqueue)
 		return -ENOMEM;
 
diff --git a/drivers/mmc/host/omap.c b/drivers/mmc/host/omap.c
index 0c7e37f..379d2ff 100644
--- a/drivers/mmc/host/omap.c
+++ b/drivers/mmc/host/omap.c
@@ -173,6 +173,8 @@ struct mmc_omap_host {
 	struct omap_mmc_platform_data *pdata;
 };
 
+static struct workqueue_struct *mmc_omap_wq;
+
 static void mmc_omap_fclk_offdelay(struct mmc_omap_slot *slot)
 {
 	unsigned long tick_ns;
@@ -289,7 +291,7 @@ static void mmc_omap_release_slot(struct mmc_omap_slot *slot, int clk_enabled)
 		host->next_slot = new_slot;
 		host->mmc = new_slot->mmc;
 		spin_unlock_irqrestore(&host->slot_lock, flags);
-		schedule_work(&host->slot_release_work);
+		queue_work(mmc_omap_wq, &host->slot_release_work);
 		return;
 	}
 
@@ -457,7 +459,7 @@ mmc_omap_xfer_done(struct mmc_omap_host *host, struct mmc_data *data)
 	}
 
 	host->stop_data = data;
-	schedule_work(&host->send_stop_work);
+	queue_work(mmc_omap_wq, &host->send_stop_work);
 }
 
 static void
@@ -637,7 +639,7 @@ mmc_omap_cmd_timer(unsigned long data)
 		OMAP_MMC_WRITE(host, IE, 0);
 		disable_irq(host->irq);
 		host->abort = 1;
-		schedule_work(&host->cmd_abort_work);
+		queue_work(mmc_omap_wq, &host->cmd_abort_work);
 	}
 	spin_unlock_irqrestore(&host->slot_lock, flags);
 }
@@ -826,7 +828,7 @@ static irqreturn_t mmc_omap_irq(int irq, void *dev_id)
 		host->abort = 1;
 		OMAP_MMC_WRITE(host, IE, 0);
 		disable_irq_nosync(host->irq);
-		schedule_work(&host->cmd_abort_work);
+		queue_work(mmc_omap_wq, &host->cmd_abort_work);
 		return IRQ_HANDLED;
 	}
 
@@ -1387,7 +1389,7 @@ static void mmc_omap_remove_slot(struct mmc_omap_slot *slot)
 
 	tasklet_kill(&slot->cover_tasklet);
 	del_timer_sync(&slot->cover_timer);
-	flush_scheduled_work();
+	flush_workqueue(mmc_omap_wq);
 
 	mmc_remove_host(mmc);
 	mmc_free_host(mmc);
@@ -1608,12 +1610,22 @@ static struct platform_driver mmc_omap_driver = {
 
 static int __init mmc_omap_init(void)
 {
-	return platform_driver_probe(&mmc_omap_driver, mmc_omap_probe);
+	int ret;
+
+	mmc_omap_wq = alloc_workqueue("mmc_omap", 0, 0);
+	if (!mmc_omap_wq)
+		return -ENOMEM;
+
+	ret = platform_driver_probe(&mmc_omap_driver, mmc_omap_probe);
+	if (ret)
+		destroy_workqueue(mmc_omap_wq);
+	return ret;
 }
 
 static void __exit mmc_omap_exit(void)
 {
 	platform_driver_unregister(&mmc_omap_driver);
+	destroy_workqueue(mmc_omap_wq);
 }
 
 module_init(mmc_omap_init);
diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c
index 5d46021..078fdf1 100644
--- a/drivers/mmc/host/omap_hsmmc.c
+++ b/drivers/mmc/host/omap_hsmmc.c
@@ -2290,7 +2290,7 @@ static int omap_hsmmc_remove(struct platform_device *pdev)
 		free_irq(host->irq, host);
 		if (mmc_slot(host).card_detect_irq)
 			free_irq(mmc_slot(host).card_detect_irq, host);
-		flush_scheduled_work();
+		flush_work_sync(&host->mmc_carddetect_work);
 
 		mmc_host_disable(host->mmc);
 		clk_disable(host->iclk);
-- 
Chris Ball   <cjb@laptop.org>   <http://printf.net/>
One Laptop Per Child

Re: [PATCH 26/30] mmc: update workqueue usages

[Tony Lindgren]

Hi,

* Chris Ball <cjb@laptop.org> [110108 18:33]:
> Hi linux-omap folks,
> 
> I'd like to merge this patch via the MMC tree; please review/ACK.
> 
> Thanks,
> 
> - Chris.
> 
> From: Tejun Heo <tj@kernel.org>
> 
> Workqueue creation API has been updated and flush_scheduled_work() is
> deprecated and scheduled to be removed.
> 
> * core/core.c: Use alloc_ordered_workqueue() instead of
>   create_singlethread_workqueue().  This removes an unnecessary
>   rescuer.
> 
> * host/omap.c: Create, use and flush mmc_omap_wq instead of the
>   system_wq.
> 
> * Flush host->mmc_carddetect_work directly on removal instead of using
>   flush_scheduled_work().
> 
> Signed-off-by: Tejun Heo <tj@kernel.org>
> Cc: Chris Ball <cjb@laptop.org>
> Cc: linux-mmc@vger.kernel.org
> ---
> This is part of a series to remove flush_scheduled_work() usage to
> prepare for deprecation of flush_scheduled_work().  Patches in this
> series are self contained and mostly straight-forward.
> 
> Please feel free to take it into the appropriate tree, or just ack it.
> In the latter case, I'll merge the patch through the workqueue tree
> during the next merge window.

Seems to work just fine, tested on n800 for omap.c and omap4 panda
for omap_hsmmc.c. Please feel free to queue via workqueue tree as
long as the MMC people are fine with that:

Tested-by: Tony Lindgren <tony@atomide.com>