linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card
@ 2018-07-25 15:25 Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 1/5] misc: rtsx_usb: Use USB remote wakeup signaling for card insertion detection Kai-Heng Feng
                   ` (4 more replies)
  0 siblings, 5 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

Hi,

This is based on Ulf's work [1] [2].

This patch series can keep rtsx_usb suspended, to save ~0.5W on Intel
platforms and ~1.5W on AMD platforms.

[1] https://patchwork.kernel.org/patch/10440583/
[2] https://patchwork.kernel.org/patch/10445725/

Kai-Heng Feng (5):
  misc: rtsx_usb: Use USB remote wakeup signaling for card insertion
    detection
  memstick: Prevent memstick host from getting runtime suspended during
    card detection
  memstick: rtsx_usb_ms: Use ms_dev() helper
  memstick: rtsx_usb_ms: Support runtime power management
  misc: rtsx_usb: / memstick: rtsx_usb_ms: Avoid long delay before
    system suspend
--
v3: Skip parent device check in rtsx_usb_resume_child() .
    Remove dev_dbg() if it only prints function name.
    Use ms_dev() helper at more places.
    Remove const qualifier for UNIVERSAL_DEV_PM_OPS.

v2: Resend to linux-usb and LKML.

 drivers/memstick/core/memstick.c    |   4 +
 drivers/memstick/host/rtsx_usb_ms.c | 148 +++++++++++++++-------------
 drivers/misc/cardreader/rtsx_usb.c  |   9 ++
 3 files changed, 91 insertions(+), 70 deletions(-)

-- 
2.17.1


^ permalink raw reply	[flat|nested] 12+ messages in thread

* [PATCH 1/5] misc: rtsx_usb: Use USB remote wakeup signaling for card insertion detection
  2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
@ 2018-07-25 15:25 ` Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection Kai-Heng Feng
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

Although rtsx_usb doesn't support card removal detection, card insertion
will resume rtsx_usb by USB remote wakeup signaling.

When rtsx_usb gets resumed, also resumes its child devices,
rtsx_usb_sdmmc and rtsx_usb_ms, to notify them there's a card in its
slot.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/misc/cardreader/rtsx_usb.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/misc/cardreader/rtsx_usb.c b/drivers/misc/cardreader/rtsx_usb.c
index b97903ff1a72..f7a66f614085 100644
--- a/drivers/misc/cardreader/rtsx_usb.c
+++ b/drivers/misc/cardreader/rtsx_usb.c
@@ -723,8 +723,15 @@ static int rtsx_usb_suspend(struct usb_interface *intf, pm_message_t message)
 	return 0;
 }
 
+static int rtsx_usb_resume_child(struct device *dev, void *data)
+{
+	pm_request_resume(dev);
+	return 0;
+}
+
 static int rtsx_usb_resume(struct usb_interface *intf)
 {
+	device_for_each_child(&intf->dev, NULL, rtsx_usb_resume_child);
 	return 0;
 }
 
@@ -734,6 +741,7 @@ static int rtsx_usb_reset_resume(struct usb_interface *intf)
 		(struct rtsx_ucr *)usb_get_intfdata(intf);
 
 	rtsx_usb_reset_chip(ucr);
+	device_for_each_child(&intf->dev, NULL, rtsx_usb_resume_child);
 	return 0;
 }
 
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection
  2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 1/5] misc: rtsx_usb: Use USB remote wakeup signaling for card insertion detection Kai-Heng Feng
@ 2018-07-25 15:25 ` Kai-Heng Feng
  2018-07-25 17:56   ` Alan Stern
  2018-07-25 15:25 ` [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper Kai-Heng Feng
                   ` (2 subsequent siblings)
  4 siblings, 1 reply; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

We can use MEMSTICK_POWER_{ON,OFF} along with pm_runtime_{get,put}
helpers to let memstick host support runtime pm.

There's a small window between memstick_detect_change() and its queued
work, memstick_check(). In this window the rpm count may go down to zero
before the memstick host powers on, so the host can be inadvertently
suspended.

Increment rpm count before calling memstick_check(), and decrement rpm
count afterward, as now we are sure the memstick host should be
suspended or not.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/core/memstick.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
index 76382c858c35..944fe0c93fa7 100644
--- a/drivers/memstick/core/memstick.c
+++ b/drivers/memstick/core/memstick.c
@@ -18,6 +18,7 @@
 #include <linux/delay.h>
 #include <linux/slab.h>
 #include <linux/module.h>
+#include <linux/pm_runtime.h>
 
 #define DRIVER_NAME "memstick"
 
@@ -209,6 +210,7 @@ static int memstick_dummy_check(struct memstick_dev *card)
  */
 void memstick_detect_change(struct memstick_host *host)
 {
+	pm_runtime_get_noresume(host->dev.parent);
 	queue_work(workqueue, &host->media_checker);
 }
 EXPORT_SYMBOL(memstick_detect_change);
@@ -479,6 +481,8 @@ static void memstick_check(struct work_struct *work)
 		host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_OFF);
 
 	mutex_unlock(&host->lock);
+
+	pm_runtime_put_noidle(host->dev.parent);
 	dev_dbg(&host->dev, "memstick_check finished\n");
 }
 
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper
  2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 1/5] misc: rtsx_usb: Use USB remote wakeup signaling for card insertion detection Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection Kai-Heng Feng
@ 2018-07-25 15:25 ` Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management Kai-Heng Feng
  2018-07-25 15:25 ` [PATCH 5/5] misc: rtsx_usb: / memstick: rtsx_usb_ms: Avoid long delay before system suspend Kai-Heng Feng
  4 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

Use ms_dev() helper for consistency.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/host/rtsx_usb_ms.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index 4f64563df7de..cd12f3d1c088 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -785,7 +785,7 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 
 	mutex_lock(&host->host_mutex);
 	if (host->req) {
-		dev_dbg(&(pdev->dev),
+		dev_dbg(ms_dev(host),
 			"%s: Controller removed during transfer\n",
 			dev_name(&msh->dev));
 		host->req->error = -ENOMEDIUM;
@@ -807,10 +807,10 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	if (pm_runtime_active(ms_dev(host)))
 		pm_runtime_put(ms_dev(host));
 
-	pm_runtime_disable(&pdev->dev);
+	pm_runtime_disable(ms_dev(host));
 	platform_set_drvdata(pdev, NULL);
 
-	dev_dbg(&(pdev->dev),
+	dev_dbg(ms_dev(host),
 		": Realtek USB Memstick controller has been removed\n");
 
 	return 0;
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management
  2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
                   ` (2 preceding siblings ...)
  2018-07-25 15:25 ` [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper Kai-Heng Feng
@ 2018-07-25 15:25 ` Kai-Heng Feng
  2018-07-25 17:59   ` Alan Stern
  2018-07-25 15:25 ` [PATCH 5/5] misc: rtsx_usb: / memstick: rtsx_usb_ms: Avoid long delay before system suspend Kai-Heng Feng
  4 siblings, 1 reply; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

In order to let host's parent device, rtsx_usb, to use USB remote wake
up signaling to do card detection, it needs to be suspended. Hence it's
necessary to add runtime PM support for the memstick host.

To keep memstick host stays suspended when it's not in use, convert the
card detection function from kthread to delayed_work, which can be
scheduled when the host is resumed and can be canceled when the host is
suspended.

Use an idle function check if there's no card and the power mode is
MEMSTICK_POWER_OFF. If both criteria are met, put the device to suspend.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/host/rtsx_usb_ms.c | 138 ++++++++++++++--------------
 1 file changed, 71 insertions(+), 67 deletions(-)

diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index cd12f3d1c088..126eb310980a 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -40,15 +40,14 @@ struct rtsx_usb_ms {
 
 	struct mutex		host_mutex;
 	struct work_struct	handle_req;
-
-	struct task_struct	*detect_ms;
-	struct completion	detect_ms_exit;
+	struct delayed_work	poll_card;
 
 	u8			ssc_depth;
 	unsigned int		clock;
 	int			power_mode;
 	unsigned char           ifmode;
 	bool			eject;
+	bool			suspend;
 };
 
 static inline struct device *ms_dev(struct rtsx_usb_ms *host)
@@ -524,7 +523,7 @@ static void rtsx_usb_ms_handle_req(struct work_struct *work)
 	int rc;
 
 	if (!host->req) {
-		pm_runtime_get_sync(ms_dev(host));
+		pm_runtime_get_noresume(ms_dev(host));
 		do {
 			rc = memstick_next_req(msh, &host->req);
 			dev_dbg(ms_dev(host), "next req %d\n", rc);
@@ -545,7 +544,7 @@ static void rtsx_usb_ms_handle_req(struct work_struct *work)
 						host->req->error);
 			}
 		} while (!rc);
-		pm_runtime_put(ms_dev(host));
+		pm_runtime_put_noidle(ms_dev(host));
 	}
 
 }
@@ -572,7 +571,7 @@ static int rtsx_usb_ms_set_param(struct memstick_host *msh,
 	dev_dbg(ms_dev(host), "%s: param = %d, value = %d\n",
 			__func__, param, value);
 
-	pm_runtime_get_sync(ms_dev(host));
+	pm_runtime_get_noresume(ms_dev(host));
 	mutex_lock(&ucr->dev_mutex);
 
 	err = rtsx_usb_card_exclusive_check(ucr, RTSX_USB_MS_CARD);
@@ -585,14 +584,14 @@ static int rtsx_usb_ms_set_param(struct memstick_host *msh,
 			break;
 
 		if (value == MEMSTICK_POWER_ON) {
-			pm_runtime_get_sync(ms_dev(host));
+			pm_runtime_get_noresume(ms_dev(host));
 			err = ms_power_on(host);
+			if (err)
+				pm_runtime_put_noidle(ms_dev(host));
 		} else if (value == MEMSTICK_POWER_OFF) {
 			err = ms_power_off(host);
-			if (host->msh->card)
+			if (!err)
 				pm_runtime_put_noidle(ms_dev(host));
-			else
-				pm_runtime_put(ms_dev(host));
 		} else
 			err = -EINVAL;
 		if (!err)
@@ -638,7 +637,7 @@ static int rtsx_usb_ms_set_param(struct memstick_host *msh,
 	}
 out:
 	mutex_unlock(&ucr->dev_mutex);
-	pm_runtime_put(ms_dev(host));
+	pm_runtime_put_noidle(ms_dev(host));
 
 	/* power-on delay */
 	if (param == MEMSTICK_POWER && value == MEMSTICK_POWER_ON)
@@ -648,75 +647,86 @@ static int rtsx_usb_ms_set_param(struct memstick_host *msh,
 	return err;
 }
 
-#ifdef CONFIG_PM_SLEEP
-static int rtsx_usb_ms_suspend(struct device *dev)
+#ifdef CONFIG_PM
+static int rtsx_usb_ms_runtime_suspend(struct device *dev)
 {
 	struct rtsx_usb_ms *host = dev_get_drvdata(dev);
 	struct memstick_host *msh = host->msh;
 
-	dev_dbg(ms_dev(host), "--> %s\n", __func__);
-
+	host->suspend = true;
 	memstick_suspend_host(msh);
+
 	return 0;
 }
 
-static int rtsx_usb_ms_resume(struct device *dev)
+static int rtsx_usb_ms_runtime_resume(struct device *dev)
 {
 	struct rtsx_usb_ms *host = dev_get_drvdata(dev);
 	struct memstick_host *msh = host->msh;
 
-	dev_dbg(ms_dev(host), "--> %s\n", __func__);
-
 	memstick_resume_host(msh);
+	host->suspend = false;
+	schedule_delayed_work(&host->poll_card, 100);
+
 	return 0;
 }
-#endif /* CONFIG_PM_SLEEP */
 
-/*
- * Thread function of ms card slot detection. The thread starts right after
- * successful host addition. It stops while the driver removal function sets
- * host->eject true.
- */
-static int rtsx_usb_detect_ms_card(void *__host)
+static int rtsx_usb_ms_runtime_idle(struct device *dev)
+{
+	struct rtsx_usb_ms *host = dev_get_drvdata(dev);
+
+	if (!host->msh->card && host->power_mode == MEMSTICK_POWER_OFF) {
+		pm_schedule_suspend(dev, 0);
+		return 0;
+	}
+
+	return -EAGAIN;
+}
+
+static UNIVERSAL_DEV_PM_OPS(rtsx_usb_ms_pm_ops,
+		rtsx_usb_ms_runtime_suspend, rtsx_usb_ms_runtime_resume,
+		rtsx_usb_ms_runtime_idle);
+#endif /* CONFIG_PM */
+
+static void rtsx_usb_ms_poll_card(struct work_struct *work)
 {
-	struct rtsx_usb_ms *host = (struct rtsx_usb_ms *)__host;
+	struct rtsx_usb_ms *host = container_of(work, struct rtsx_usb_ms,
+			poll_card.work);
 	struct rtsx_ucr *ucr = host->ucr;
-	u8 val = 0;
 	int err;
+	u8 val;
 
-	for (;;) {
-		pm_runtime_get_sync(ms_dev(host));
-		mutex_lock(&ucr->dev_mutex);
+	if (host->eject || host->suspend)
+		return;
 
-		/* Check pending MS card changes */
-		err = rtsx_usb_read_register(ucr, CARD_INT_PEND, &val);
-		if (err) {
-			mutex_unlock(&ucr->dev_mutex);
-			goto poll_again;
-		}
-
-		/* Clear the pending */
-		rtsx_usb_write_register(ucr, CARD_INT_PEND,
-				XD_INT | MS_INT | SD_INT,
-				XD_INT | MS_INT | SD_INT);
+	pm_runtime_get_noresume(ms_dev(host));
+	mutex_lock(&ucr->dev_mutex);
 
+	/* Check pending MS card changes */
+	err = rtsx_usb_read_register(ucr, CARD_INT_PEND, &val);
+	if (err) {
 		mutex_unlock(&ucr->dev_mutex);
+		goto poll_again;
+	}
 
-		if (val & MS_INT) {
-			dev_dbg(ms_dev(host), "MS slot change detected\n");
-			memstick_detect_change(host->msh);
-		}
+	/* Clear the pending */
+	rtsx_usb_write_register(ucr, CARD_INT_PEND,
+			XD_INT | MS_INT | SD_INT,
+			XD_INT | MS_INT | SD_INT);
 
-poll_again:
-		pm_runtime_put(ms_dev(host));
-		if (host->eject)
-			break;
+	mutex_unlock(&ucr->dev_mutex);
+	pm_runtime_put_noidle(ms_dev(host));
 
-		schedule_timeout_idle(HZ);
+	if (val & MS_INT) {
+		dev_dbg(ms_dev(host), "MS slot change detected\n");
+		memstick_detect_change(host->msh);
 	}
 
-	complete(&host->detect_ms_exit);
-	return 0;
+	pm_runtime_idle(ms_dev(host));
+
+poll_again:
+	if (!host->eject && !host->suspend)
+		schedule_delayed_work(&host->poll_card, 100);
 }
 
 static int rtsx_usb_ms_drv_probe(struct platform_device *pdev)
@@ -747,26 +757,21 @@ static int rtsx_usb_ms_drv_probe(struct platform_device *pdev)
 	mutex_init(&host->host_mutex);
 	INIT_WORK(&host->handle_req, rtsx_usb_ms_handle_req);
 
-	init_completion(&host->detect_ms_exit);
-	host->detect_ms = kthread_create(rtsx_usb_detect_ms_card, host,
-			"rtsx_usb_ms_%d", pdev->id);
-	if (IS_ERR(host->detect_ms)) {
-		dev_dbg(&(pdev->dev),
-				"Unable to create polling thread.\n");
-		err = PTR_ERR(host->detect_ms);
-		goto err_out;
-	}
+	INIT_DELAYED_WORK(&host->poll_card, rtsx_usb_ms_poll_card);
 
 	msh->request = rtsx_usb_ms_request;
 	msh->set_param = rtsx_usb_ms_set_param;
 	msh->caps = MEMSTICK_CAP_PAR4;
 
-	pm_runtime_enable(&pdev->dev);
+	pm_runtime_set_active(ms_dev(host));
+	pm_runtime_enable(ms_dev(host));
+
 	err = memstick_add_host(msh);
 	if (err)
 		goto err_out;
 
-	wake_up_process(host->detect_ms);
+	schedule_delayed_work(&host->poll_card, 100);
+
 	return 0;
 err_out:
 	memstick_free_host(msh);
@@ -782,6 +787,7 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	msh = host->msh;
 	host->eject = true;
 	cancel_work_sync(&host->handle_req);
+	cancel_delayed_work_sync(&host->poll_card);
 
 	mutex_lock(&host->host_mutex);
 	if (host->req) {
@@ -797,7 +803,6 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	}
 	mutex_unlock(&host->host_mutex);
 
-	wait_for_completion(&host->detect_ms_exit);
 	memstick_remove_host(msh);
 	memstick_free_host(msh);
 
@@ -816,9 +821,6 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static SIMPLE_DEV_PM_OPS(rtsx_usb_ms_pm_ops,
-		rtsx_usb_ms_suspend, rtsx_usb_ms_resume);
-
 static struct platform_device_id rtsx_usb_ms_ids[] = {
 	{
 		.name = "rtsx_usb_ms",
@@ -834,7 +836,9 @@ static struct platform_driver rtsx_usb_ms_driver = {
 	.id_table       = rtsx_usb_ms_ids,
 	.driver		= {
 		.name	= "rtsx_usb_ms",
+#ifdef CONFIG_PM
 		.pm	= &rtsx_usb_ms_pm_ops,
+#endif
 	},
 };
 module_platform_driver(rtsx_usb_ms_driver);
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* [PATCH 5/5] misc: rtsx_usb: / memstick: rtsx_usb_ms: Avoid long delay before system suspend
  2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
                   ` (3 preceding siblings ...)
  2018-07-25 15:25 ` [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management Kai-Heng Feng
@ 2018-07-25 15:25 ` Kai-Heng Feng
  4 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-25 15:25 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

There's a long power-on delay at the end of rtsx_usb_ms_set_param().
This delay is noticeable right before system suspend.

To prevent already suspended memstick host from getting powered on by PM
core, use DPM_FLAG_SMART_SUSPEND to avoid the situation.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/host/rtsx_usb_ms.c | 4 ++++
 drivers/misc/cardreader/rtsx_usb.c  | 1 +
 2 files changed, 5 insertions(+)

diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index 126eb310980a..e3b635d1220f 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -763,6 +763,10 @@ static int rtsx_usb_ms_drv_probe(struct platform_device *pdev)
 	msh->set_param = rtsx_usb_ms_set_param;
 	msh->caps = MEMSTICK_CAP_PAR4;
 
+	/* DPM_FLAG_LEAVE_SUSPENDED is not needed, the parent device will wake
+	 * up memstick host.
+	 */
+	dev_pm_set_driver_flags(ms_dev(host), DPM_FLAG_SMART_SUSPEND);
 	pm_runtime_set_active(ms_dev(host));
 	pm_runtime_enable(ms_dev(host));
 
diff --git a/drivers/misc/cardreader/rtsx_usb.c b/drivers/misc/cardreader/rtsx_usb.c
index f7a66f614085..98bb878a6ade 100644
--- a/drivers/misc/cardreader/rtsx_usb.c
+++ b/drivers/misc/cardreader/rtsx_usb.c
@@ -671,6 +671,7 @@ static int rtsx_usb_probe(struct usb_interface *intf,
 		goto out_init_fail;
 
 #ifdef CONFIG_PM
+	dev_pm_set_driver_flags(&intf->dev, DPM_FLAG_SMART_SUSPEND | DPM_FLAG_LEAVE_SUSPENDED);
 	intf->needs_remote_wakeup = 1;
 	usb_enable_autosuspend(usb_dev);
 #endif
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* Re: [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection
  2018-07-25 15:25 ` [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection Kai-Heng Feng
@ 2018-07-25 17:56   ` Alan Stern
  2018-07-26  7:01     ` Kai-Heng Feng
  0 siblings, 1 reply; 12+ messages in thread
From: Alan Stern @ 2018-07-25 17:56 UTC (permalink / raw)
  To: Kai-Heng Feng
  Cc: arnd, gregkh, ulf.hansson, bauer.chen, ricky_wu, linux-kernel, linux-usb

On Wed, 25 Jul 2018, Kai-Heng Feng wrote:

> We can use MEMSTICK_POWER_{ON,OFF} along with pm_runtime_{get,put}
> helpers to let memstick host support runtime pm.
> 
> There's a small window between memstick_detect_change() and its queued
> work, memstick_check(). In this window the rpm count may go down to zero
> before the memstick host powers on, so the host can be inadvertently
> suspended.
> 
> Increment rpm count before calling memstick_check(), and decrement rpm
> count afterward, as now we are sure the memstick host should be
> suspended or not.
> 
> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
> ---
>  drivers/memstick/core/memstick.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
> index 76382c858c35..944fe0c93fa7 100644
> --- a/drivers/memstick/core/memstick.c
> +++ b/drivers/memstick/core/memstick.c
> @@ -18,6 +18,7 @@
>  #include <linux/delay.h>
>  #include <linux/slab.h>
>  #include <linux/module.h>
> +#include <linux/pm_runtime.h>
>  
>  #define DRIVER_NAME "memstick"
>  
> @@ -209,6 +210,7 @@ static int memstick_dummy_check(struct memstick_dev *card)
>   */
>  void memstick_detect_change(struct memstick_host *host)
>  {
> +	pm_runtime_get_noresume(host->dev.parent);
>  	queue_work(workqueue, &host->media_checker);
>  }
>  EXPORT_SYMBOL(memstick_detect_change);
> @@ -479,6 +481,8 @@ static void memstick_check(struct work_struct *work)
>  		host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_OFF);
>  
>  	mutex_unlock(&host->lock);
> +
> +	pm_runtime_put_noidle(host->dev.parent);
>  	dev_dbg(&host->dev, "memstick_check finished\n");
>  }

This should be pm_runtime_put(), not _put_noidle().  The reason is
because if this call causes the PM runtime usage count to drop to 0,
you do want the device to go into runtime suspend.

Alan Stern


^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management
  2018-07-25 15:25 ` [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management Kai-Heng Feng
@ 2018-07-25 17:59   ` Alan Stern
  2018-07-26  7:38     ` Kai-Heng Feng
  0 siblings, 1 reply; 12+ messages in thread
From: Alan Stern @ 2018-07-25 17:59 UTC (permalink / raw)
  To: Kai-Heng Feng
  Cc: arnd, gregkh, ulf.hansson, bauer.chen, ricky_wu, linux-kernel, linux-usb

On Wed, 25 Jul 2018, Kai-Heng Feng wrote:

> In order to let host's parent device, rtsx_usb, to use USB remote wake
> up signaling to do card detection, it needs to be suspended. Hence it's
> necessary to add runtime PM support for the memstick host.
> 
> To keep memstick host stays suspended when it's not in use, convert the
> card detection function from kthread to delayed_work, which can be
> scheduled when the host is resumed and can be canceled when the host is
> suspended.
> 
> Use an idle function check if there's no card and the power mode is
> MEMSTICK_POWER_OFF. If both criteria are met, put the device to suspend.
> 
> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>

I'm not familiar enough with this driver to comment on specific
instances, but you should carefully go through the code and make sure
that you use pm_runtime_get_noresume() and pm_runtime_put_noidle()
correctly.

The basic idea is to use these when you know beforehand that the device
is already at full power (for _get_noresume) and the usage count will
not go to 0 (for _put_noidle).  If you aren't certain of these
requirements then you should call pm_runtime_get_sync() and
pm_runtime_put().

Alan Stern


^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection
  2018-07-25 17:56   ` Alan Stern
@ 2018-07-26  7:01     ` Kai-Heng Feng
  0 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-26  7:01 UTC (permalink / raw)
  To: Alan Stern
  Cc: arnd, gregkh, ulf.hansson, bauer.chen, ricky_wu, linux-kernel, linux-usb



> On 2018Jul26, at 01:56, Alan Stern <stern@rowland.harvard.edu> wrote:
> 
> On Wed, 25 Jul 2018, Kai-Heng Feng wrote:
> 
>> We can use MEMSTICK_POWER_{ON,OFF} along with pm_runtime_{get,put}
>> helpers to let memstick host support runtime pm.
>> 
>> There's a small window between memstick_detect_change() and its queued
>> work, memstick_check(). In this window the rpm count may go down to zero
>> before the memstick host powers on, so the host can be inadvertently
>> suspended.
>> 
>> Increment rpm count before calling memstick_check(), and decrement rpm
>> count afterward, as now we are sure the memstick host should be
>> suspended or not.
>> 
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
>> ---
>> drivers/memstick/core/memstick.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>> 
>> diff --git a/drivers/memstick/core/memstick.c b/drivers/memstick/core/memstick.c
>> index 76382c858c35..944fe0c93fa7 100644
>> --- a/drivers/memstick/core/memstick.c
>> +++ b/drivers/memstick/core/memstick.c
>> @@ -18,6 +18,7 @@
>> #include <linux/delay.h>
>> #include <linux/slab.h>
>> #include <linux/module.h>
>> +#include <linux/pm_runtime.h>
>> 
>> #define DRIVER_NAME "memstick"
>> 
>> @@ -209,6 +210,7 @@ static int memstick_dummy_check(struct memstick_dev *card)
>>  */
>> void memstick_detect_change(struct memstick_host *host)
>> {
>> +	pm_runtime_get_noresume(host->dev.parent);
>> 	queue_work(workqueue, &host->media_checker);
>> }
>> EXPORT_SYMBOL(memstick_detect_change);
>> @@ -479,6 +481,8 @@ static void memstick_check(struct work_struct *work)
>> 		host->set_param(host, MEMSTICK_POWER, MEMSTICK_POWER_OFF);
>> 
>> 	mutex_unlock(&host->lock);
>> +
>> +	pm_runtime_put_noidle(host->dev.parent);
>> 	dev_dbg(&host->dev, "memstick_check finished\n");
>> }
> 
> This should be pm_runtime_put(), not _put_noidle().  The reason is
> because if this call causes the PM runtime usage count to drop to 0,
> you do want the device to go into runtime suspend.

Will update in next version, thanks!

Kai-Heng

> 
> Alan Stern


^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management
  2018-07-25 17:59   ` Alan Stern
@ 2018-07-26  7:38     ` Kai-Heng Feng
  0 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-26  7:38 UTC (permalink / raw)
  To: Alan Stern
  Cc: arnd, gregkh, ulf.hansson, bauer.chen, ricky_wu, linux-kernel, linux-usb



> On 2018Jul26, at 01:59, Alan Stern <stern@rowland.harvard.edu> wrote:
> 
> On Wed, 25 Jul 2018, Kai-Heng Feng wrote:
> 
>> In order to let host's parent device, rtsx_usb, to use USB remote wake
>> up signaling to do card detection, it needs to be suspended. Hence it's
>> necessary to add runtime PM support for the memstick host.
>> 
>> To keep memstick host stays suspended when it's not in use, convert the
>> card detection function from kthread to delayed_work, which can be
>> scheduled when the host is resumed and can be canceled when the host is
>> suspended.
>> 
>> Use an idle function check if there's no card and the power mode is
>> MEMSTICK_POWER_OFF. If both criteria are met, put the device to suspend.
>> 
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
> 
> I'm not familiar enough with this driver to comment on specific
> instances, but you should carefully go through the code and make sure
> that you use pm_runtime_get_noresume() and pm_runtime_put_noidle()
> correctly.
> 
> The basic idea is to use these when you know beforehand that the device
> is already at full power (for _get_noresume) and the usage count will
> not go to 0 (for _put_noidle).  If you aren't certain of these
> requirements then you should call pm_runtime_get_sync() and
> pm_runtime_put().

There are two different uses for pm_runtime helpers.

The first use case is to prevent parent device, rtsx_usb, from suspending:
pm_runtime_get_noresume()
rtsx_usb_read_register()
pm_runtime_put_noidle()

The second use case, to manage memstick host itself.
memstick_detect_change() powers the host on if there’s a card in slot, and it powers host off when there’s no card.
Once memstick_detect_change() is done, we can check if the host can be suspended.

So we are sure the timing when the power is on or off.

Kai-Heng

> 
> Alan Stern
> 


^ permalink raw reply	[flat|nested] 12+ messages in thread

* [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper
  2018-07-31  6:17 [PATCH 0/5 v4] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
@ 2018-07-31  6:17 ` Kai-Heng Feng
  0 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-31  6:17 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

Use ms_dev() helper for consistency.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/host/rtsx_usb_ms.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index 4f64563df7de..cd12f3d1c088 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -785,7 +785,7 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 
 	mutex_lock(&host->host_mutex);
 	if (host->req) {
-		dev_dbg(&(pdev->dev),
+		dev_dbg(ms_dev(host),
 			"%s: Controller removed during transfer\n",
 			dev_name(&msh->dev));
 		host->req->error = -ENOMEDIUM;
@@ -807,10 +807,10 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	if (pm_runtime_active(ms_dev(host)))
 		pm_runtime_put(ms_dev(host));
 
-	pm_runtime_disable(&pdev->dev);
+	pm_runtime_disable(ms_dev(host));
 	platform_set_drvdata(pdev, NULL);
 
-	dev_dbg(&(pdev->dev),
+	dev_dbg(ms_dev(host),
 		": Realtek USB Memstick controller has been removed\n");
 
 	return 0;
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

* [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper
  2018-07-23 10:27 [PATCH 0/5 v2] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
@ 2018-07-23 10:27 ` Kai-Heng Feng
  0 siblings, 0 replies; 12+ messages in thread
From: Kai-Heng Feng @ 2018-07-23 10:27 UTC (permalink / raw)
  To: arnd, gregkh
  Cc: ulf.hansson, stern, bauer.chen, ricky_wu, linux-kernel,
	linux-usb, Kai-Heng Feng

Use ms_dev() helper for consistency.

Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
---
 drivers/memstick/host/rtsx_usb_ms.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/memstick/host/rtsx_usb_ms.c b/drivers/memstick/host/rtsx_usb_ms.c
index 4f64563df7de..d56f0863653f 100644
--- a/drivers/memstick/host/rtsx_usb_ms.c
+++ b/drivers/memstick/host/rtsx_usb_ms.c
@@ -807,7 +807,7 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	if (pm_runtime_active(ms_dev(host)))
 		pm_runtime_put(ms_dev(host));
 
-	pm_runtime_disable(&pdev->dev);
+	pm_runtime_disable(ms_dev(host));
 	platform_set_drvdata(pdev, NULL);
 
 	dev_dbg(&(pdev->dev),
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 12+ messages in thread

end of thread, other threads:[~2018-07-31  6:17 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-07-25 15:25 [PATCH 0/5 v3] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
2018-07-25 15:25 ` [PATCH 1/5] misc: rtsx_usb: Use USB remote wakeup signaling for card insertion detection Kai-Heng Feng
2018-07-25 15:25 ` [PATCH 2/5] memstick: Prevent memstick host from getting runtime suspended during card detection Kai-Heng Feng
2018-07-25 17:56   ` Alan Stern
2018-07-26  7:01     ` Kai-Heng Feng
2018-07-25 15:25 ` [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper Kai-Heng Feng
2018-07-25 15:25 ` [PATCH 4/5] memstick: rtsx_usb_ms: Support runtime power management Kai-Heng Feng
2018-07-25 17:59   ` Alan Stern
2018-07-26  7:38     ` Kai-Heng Feng
2018-07-25 15:25 ` [PATCH 5/5] misc: rtsx_usb: / memstick: rtsx_usb_ms: Avoid long delay before system suspend Kai-Heng Feng
  -- strict thread matches above, loose matches on Subject: below --
2018-07-31  6:17 [PATCH 0/5 v4] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
2018-07-31  6:17 ` [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper Kai-Heng Feng
2018-07-23 10:27 [PATCH 0/5 v2] Keep rtsx_usb suspended when there's no card Kai-Heng Feng
2018-07-23 10:27 ` [PATCH 3/5] memstick: rtsx_usb_ms: Use ms_dev() helper Kai-Heng Feng

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).