linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v1 0/3] ethernet: sun: use generic power management
@ 2020-07-06  8:57 Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 1/3] sun/sungem: " Vaibhav Gupta
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Vaibhav Gupta @ 2020-07-06  8:57 UTC (permalink / raw)
  To: Bjorn Helgaas, Bjorn Helgaas, bjorn, Vaibhav Gupta,
	David S. Miller, Jakub Kicinski
  Cc: Vaibhav Gupta, netdev, linux-kernel, linux-kernel-mentees, skhan

Linux Kernel Mentee: Remove Legacy Power Management.

The purpose of this patch series is to remove legacy power management callbacks
from sun ethernet drivers.

The callbacks performing suspend() and resume() operations are still calling
pci_save_state(), pci_set_power_state(), etc. and handling the power management
themselves, which is not recommended.

The conversion requires the removal of the those function calls and change the
callback definition accordingly and make use of dev_pm_ops structure.

All patches are compile-tested only.

Vaibhav Gupta (3):
  sun/sungem: use generic power management
  sun/niu: use generic power management
  sun/cassini: use generic power management

 drivers/net/ethernet/sun/cassini.c | 17 +++----
 drivers/net/ethernet/sun/niu.c     | 17 +++----
 drivers/net/ethernet/sun/sungem.c  | 76 +++++++++++++++++-------------
 3 files changed, 57 insertions(+), 53 deletions(-)

-- 
2.27.0


^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PATCH v1 1/3] sun/sungem: use generic power management
  2020-07-06  8:57 [PATCH v1 0/3] ethernet: sun: use generic power management Vaibhav Gupta
@ 2020-07-06  8:57 ` Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 2/3] sun/niu: " Vaibhav Gupta
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Vaibhav Gupta @ 2020-07-06  8:57 UTC (permalink / raw)
  To: Bjorn Helgaas, Bjorn Helgaas, bjorn, Vaibhav Gupta,
	David S. Miller, Jakub Kicinski
  Cc: Vaibhav Gupta, netdev, linux-kernel, linux-kernel-mentees, skhan

With legacy PM, drivers themselves were responsible for managing the
device's power states and takes care of register states. And they use PCI
helper functions to do it.

After upgrading to the generic structure, PCI core will take care of
required tasks and drivers should do only device-specific operations.

In this driver:
gem_suspend() calls gem_do_stop() which in turn invokes
pci_disable_device(). As the PCI helper function is not called at the
end/start of the function body, breaking the function in two parts
may change its behavior.

The only other function invoking gem_do_stop() is gem_close(). Hence,
gem_close() and gem_suspend() can do the required end steps on their own.

The same case is with gem_resume(). Both gem_resume() and gem_open()
invoke gem_do_start(). Again, make the caller functions do the required
steps on their own.

Compile-tested only.

Signed-off-by: Vaibhav Gupta <vaibhavgupta40@gmail.com>
---
 drivers/net/ethernet/sun/sungem.c | 76 +++++++++++++++++--------------
 1 file changed, 43 insertions(+), 33 deletions(-)

diff --git a/drivers/net/ethernet/sun/sungem.c b/drivers/net/ethernet/sun/sungem.c
index 2d392a7b179a..0e9d2cf7979b 100644
--- a/drivers/net/ethernet/sun/sungem.c
+++ b/drivers/net/ethernet/sun/sungem.c
@@ -2139,20 +2139,6 @@ static int gem_do_start(struct net_device *dev)
 	struct gem *gp = netdev_priv(dev);
 	int rc;
 
-	/* Enable the cell */
-	gem_get_cell(gp);
-
-	/* Make sure PCI access and bus master are enabled */
-	rc = pci_enable_device(gp->pdev);
-	if (rc) {
-		netdev_err(dev, "Failed to enable chip on PCI bus !\n");
-
-		/* Put cell and forget it for now, it will be considered as
-		 * still asleep, a new sleep cycle may bring it back
-		 */
-		gem_put_cell(gp);
-		return -ENXIO;
-	}
 	pci_set_master(gp->pdev);
 
 	/* Init & setup chip hardware */
@@ -2230,13 +2216,6 @@ static void gem_do_stop(struct net_device *dev, int wol)
 
 	/* Shut the PHY down eventually and setup WOL */
 	gem_stop_phy(gp, wol);
-
-	/* Make sure bus master is disabled */
-	pci_disable_device(gp->pdev);
-
-	/* Cell not needed neither if no WOL */
-	if (!wol)
-		gem_put_cell(gp);
 }
 
 static void gem_reset_task(struct work_struct *work)
@@ -2288,26 +2267,53 @@ static void gem_reset_task(struct work_struct *work)
 
 static int gem_open(struct net_device *dev)
 {
+	struct gem *gp = netdev_priv(dev);
+	int rc;
+
 	/* We allow open while suspended, we just do nothing,
 	 * the chip will be initialized in resume()
 	 */
-	if (netif_device_present(dev))
+	if (netif_device_present(dev)) {
+		/* Enable the cell */
+		gem_get_cell(gp);
+
+		/* Make sure PCI access and bus master are enabled */
+		rc = pci_enable_device(gp->pdev);
+		if (rc) {
+			netdev_err(dev, "Failed to enable chip on PCI bus !\n");
+
+			/* Put cell and forget it for now, it will be considered
+			 *as still asleep, a new sleep cycle may bring it back
+			 */
+			gem_put_cell(gp);
+			return -ENXIO;
+		}
 		return gem_do_start(dev);
+	}
+
 	return 0;
 }
 
 static int gem_close(struct net_device *dev)
 {
-	if (netif_device_present(dev))
+	struct gem *gp = netdev_priv(dev);
+
+	if (netif_device_present(dev)) {
 		gem_do_stop(dev, 0);
 
+		/* Make sure bus master is disabled */
+		pci_disable_device(gp->pdev);
+
+		/* Cell not needed neither if no WOL */
+		if (!gp->asleep_wol)
+			gem_put_cell(gp);
+	}
 	return 0;
 }
 
-#ifdef CONFIG_PM
-static int gem_suspend(struct pci_dev *pdev, pm_message_t state)
+static int __maybe_unused gem_suspend(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct gem *gp = netdev_priv(dev);
 
 	/* Lock the network stack first to avoid racing with open/close,
@@ -2336,15 +2342,19 @@ static int gem_suspend(struct pci_dev *pdev, pm_message_t state)
 	gp->asleep_wol = !!gp->wake_on_lan;
 	gem_do_stop(dev, gp->asleep_wol);
 
+	/* Cell not needed neither if no WOL */
+	if (!gp->asleep_wol)
+		gem_put_cell(gp);
+
 	/* Unlock the network stack */
 	rtnl_unlock();
 
 	return 0;
 }
 
-static int gem_resume(struct pci_dev *pdev)
+static int __maybe_unused gem_resume(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct gem *gp = netdev_priv(dev);
 
 	/* See locking comment in gem_suspend */
@@ -2359,6 +2369,9 @@ static int gem_resume(struct pci_dev *pdev)
 		return 0;
 	}
 
+	/* Enable the cell */
+	gem_get_cell(gp);
+
 	/* Restart chip. If that fails there isn't much we can do, we
 	 * leave things stopped.
 	 */
@@ -2375,7 +2388,6 @@ static int gem_resume(struct pci_dev *pdev)
 
 	return 0;
 }
-#endif /* CONFIG_PM */
 
 static struct net_device_stats *gem_get_stats(struct net_device *dev)
 {
@@ -3019,16 +3031,14 @@ static int gem_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 }
 
+static SIMPLE_DEV_PM_OPS(gem_pm_ops, gem_suspend, gem_resume);
 
 static struct pci_driver gem_driver = {
 	.name		= GEM_MODULE_NAME,
 	.id_table	= gem_pci_tbl,
 	.probe		= gem_init_one,
 	.remove		= gem_remove_one,
-#ifdef CONFIG_PM
-	.suspend	= gem_suspend,
-	.resume		= gem_resume,
-#endif /* CONFIG_PM */
+	.driver.pm	= &gem_pm_ops,
 };
 
 module_pci_driver(gem_driver);
-- 
2.27.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH v1 2/3] sun/niu: use generic power management
  2020-07-06  8:57 [PATCH v1 0/3] ethernet: sun: use generic power management Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 1/3] sun/sungem: " Vaibhav Gupta
@ 2020-07-06  8:57 ` Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 3/3] sun/cassini: " Vaibhav Gupta
  2020-07-06 19:24 ` [PATCH v1 0/3] ethernet: sun: " David Miller
  3 siblings, 0 replies; 5+ messages in thread
From: Vaibhav Gupta @ 2020-07-06  8:57 UTC (permalink / raw)
  To: Bjorn Helgaas, Bjorn Helgaas, bjorn, Vaibhav Gupta,
	David S. Miller, Jakub Kicinski
  Cc: Vaibhav Gupta, netdev, linux-kernel, linux-kernel-mentees, skhan

With legacy PM, drivers themselves were responsible for managing the
device's power states and takes care of register states.

After upgrading to the generic structure, PCI core will take care of
required tasks and drivers should do only device-specific operations.

The driver was calling pci_save/restore_state() which is no more needed.

Compile-tested only.

Signed-off-by: Vaibhav Gupta <vaibhavgupta40@gmail.com>
---
 drivers/net/ethernet/sun/niu.c | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/sun/niu.c b/drivers/net/ethernet/sun/niu.c
index 9a5004f674c7..68541c823245 100644
--- a/drivers/net/ethernet/sun/niu.c
+++ b/drivers/net/ethernet/sun/niu.c
@@ -9873,9 +9873,9 @@ static void niu_pci_remove_one(struct pci_dev *pdev)
 	}
 }
 
-static int niu_suspend(struct pci_dev *pdev, pm_message_t state)
+static int niu_suspend(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct niu *np = netdev_priv(dev);
 	unsigned long flags;
 
@@ -9897,14 +9897,12 @@ static int niu_suspend(struct pci_dev *pdev, pm_message_t state)
 	niu_stop_hw(np);
 	spin_unlock_irqrestore(&np->lock, flags);
 
-	pci_save_state(pdev);
-
 	return 0;
 }
 
-static int niu_resume(struct pci_dev *pdev)
+static int niu_resume(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct niu *np = netdev_priv(dev);
 	unsigned long flags;
 	int err;
@@ -9912,8 +9910,6 @@ static int niu_resume(struct pci_dev *pdev)
 	if (!netif_running(dev))
 		return 0;
 
-	pci_restore_state(pdev);
-
 	netif_device_attach(dev);
 
 	spin_lock_irqsave(&np->lock, flags);
@@ -9930,13 +9926,14 @@ static int niu_resume(struct pci_dev *pdev)
 	return err;
 }
 
+static SIMPLE_DEV_PM_OPS(niu_pm_ops, niu_suspend, niu_resume);
+
 static struct pci_driver niu_pci_driver = {
 	.name		= DRV_MODULE_NAME,
 	.id_table	= niu_pci_tbl,
 	.probe		= niu_pci_init_one,
 	.remove		= niu_pci_remove_one,
-	.suspend	= niu_suspend,
-	.resume		= niu_resume,
+	.driver.pm	= &niu_pm_ops,
 };
 
 #ifdef CONFIG_SPARC64
-- 
2.27.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* [PATCH v1 3/3] sun/cassini: use generic power management
  2020-07-06  8:57 [PATCH v1 0/3] ethernet: sun: use generic power management Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 1/3] sun/sungem: " Vaibhav Gupta
  2020-07-06  8:57 ` [PATCH v1 2/3] sun/niu: " Vaibhav Gupta
@ 2020-07-06  8:57 ` Vaibhav Gupta
  2020-07-06 19:24 ` [PATCH v1 0/3] ethernet: sun: " David Miller
  3 siblings, 0 replies; 5+ messages in thread
From: Vaibhav Gupta @ 2020-07-06  8:57 UTC (permalink / raw)
  To: Bjorn Helgaas, Bjorn Helgaas, bjorn, Vaibhav Gupta,
	David S. Miller, Jakub Kicinski
  Cc: Vaibhav Gupta, netdev, linux-kernel, linux-kernel-mentees, skhan

With legacy PM, drivers themselves were responsible for managing the
device's power states and takes care of register states.

After upgrading to the generic structure, PCI core will take care of
required tasks and drivers should do only device-specific operations.

Compile-tested only.

Signed-off-by: Vaibhav Gupta <vaibhavgupta40@gmail.com>
---
 drivers/net/ethernet/sun/cassini.c | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/sun/cassini.c b/drivers/net/ethernet/sun/cassini.c
index debd3c3fa6fb..5b11124450d6 100644
--- a/drivers/net/ethernet/sun/cassini.c
+++ b/drivers/net/ethernet/sun/cassini.c
@@ -5172,10 +5172,9 @@ static void cas_remove_one(struct pci_dev *pdev)
 	pci_disable_device(pdev);
 }
 
-#ifdef CONFIG_PM
-static int cas_suspend(struct pci_dev *pdev, pm_message_t state)
+static int __maybe_unused cas_suspend(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct cas *cp = netdev_priv(dev);
 	unsigned long flags;
 
@@ -5204,9 +5203,9 @@ static int cas_suspend(struct pci_dev *pdev, pm_message_t state)
 	return 0;
 }
 
-static int cas_resume(struct pci_dev *pdev)
+static int cas_resume(struct device *dev_d)
 {
-	struct net_device *dev = pci_get_drvdata(pdev);
+	struct net_device *dev = dev_get_drvdata(dev_d);
 	struct cas *cp = netdev_priv(dev);
 
 	netdev_info(dev, "resuming\n");
@@ -5227,17 +5226,15 @@ static int cas_resume(struct pci_dev *pdev)
 	mutex_unlock(&cp->pm_mutex);
 	return 0;
 }
-#endif /* CONFIG_PM */
+
+static SIMPLE_DEV_PM_OPS(cas_pm_ops, cas_suspend, cas_resume);
 
 static struct pci_driver cas_driver = {
 	.name		= DRV_MODULE_NAME,
 	.id_table	= cas_pci_tbl,
 	.probe		= cas_init_one,
 	.remove		= cas_remove_one,
-#ifdef CONFIG_PM
-	.suspend	= cas_suspend,
-	.resume		= cas_resume
-#endif
+	.driver.pm	= &cas_pm_ops,
 };
 
 static int __init cas_init(void)
-- 
2.27.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH v1 0/3] ethernet: sun: use generic power management
  2020-07-06  8:57 [PATCH v1 0/3] ethernet: sun: use generic power management Vaibhav Gupta
                   ` (2 preceding siblings ...)
  2020-07-06  8:57 ` [PATCH v1 3/3] sun/cassini: " Vaibhav Gupta
@ 2020-07-06 19:24 ` David Miller
  3 siblings, 0 replies; 5+ messages in thread
From: David Miller @ 2020-07-06 19:24 UTC (permalink / raw)
  To: vaibhavgupta40
  Cc: helgaas, bhelgaas, bjorn, vaibhav.varodek, kuba, netdev,
	linux-kernel, linux-kernel-mentees, skhan

From: Vaibhav Gupta <vaibhavgupta40@gmail.com>
Date: Mon,  6 Jul 2020 14:27:43 +0530

> Linux Kernel Mentee: Remove Legacy Power Management.
> 
> The purpose of this patch series is to remove legacy power management callbacks
> from sun ethernet drivers.
> 
> The callbacks performing suspend() and resume() operations are still calling
> pci_save_state(), pci_set_power_state(), etc. and handling the power management
> themselves, which is not recommended.
> 
> The conversion requires the removal of the those function calls and change the
> callback definition accordingly and make use of dev_pm_ops structure.
> 
> All patches are compile-tested only.

Series applied to net-next, thank you.

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2020-07-06 19:24 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-07-06  8:57 [PATCH v1 0/3] ethernet: sun: use generic power management Vaibhav Gupta
2020-07-06  8:57 ` [PATCH v1 1/3] sun/sungem: " Vaibhav Gupta
2020-07-06  8:57 ` [PATCH v1 2/3] sun/niu: " Vaibhav Gupta
2020-07-06  8:57 ` [PATCH v1 3/3] sun/cassini: " Vaibhav Gupta
2020-07-06 19:24 ` [PATCH v1 0/3] ethernet: sun: " David Miller

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).