[RFC/RFT PATCH] Staging: et131x: add suspend/resume support

[RFC/RFT PATCH] Staging: et131x: add suspend/resume support

[Herton Ronaldo Krzesinski]

Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>
---
 drivers/staging/et131x/README           |    2 +-
 drivers/staging/et131x/et131x.h         |    2 +
 drivers/staging/et131x/et131x_initpci.c |   61 ++++++++++++++++++++++++++++---
 drivers/staging/et131x/et131x_netdev.c  |    2 -
 4 files changed, 58 insertions(+), 9 deletions(-)

diff --git a/drivers/staging/et131x/README b/drivers/staging/et131x/README
index 28752a5..4dca010 100644
--- a/drivers/staging/et131x/README
+++ b/drivers/staging/et131x/README
@@ -14,7 +14,7 @@ TODO:
 	- kill useless typecasts (e.g. in et1310_phy.c)
 	- alloc_etherdev is initializing memory with zero?!?
 	- add_timer call in et131x_netdev.c is correct?
-	- Add power saving functionality (suspend, sleep, resume)
+	- Add power saving functionality (sleep)
 	- Implement a few more kernel Parameter (set mac )
 
 Please send patches to:
diff --git a/drivers/staging/et131x/et131x.h b/drivers/staging/et131x/et131x.h
index 8aa3365..1daf53c 100644
--- a/drivers/staging/et131x/et131x.h
+++ b/drivers/staging/et131x/et131x.h
@@ -84,6 +84,8 @@ void SetupDeviceForUnicast(struct et131x_adapter *adapter);
 
 /* et131x_netdev.c */
 struct net_device *et131x_device_alloc(void);
+int et131x_open(struct net_device *netdev);
+int et131x_close(struct net_device *netdev);
 
 /* et131x_pm.c */
 void EnablePhyComa(struct et131x_adapter *adapter);
diff --git a/drivers/staging/et131x/et131x_initpci.c b/drivers/staging/et131x/et131x_initpci.c
index 50237ac..1b39141 100644
--- a/drivers/staging/et131x/et131x_initpci.c
+++ b/drivers/staging/et131x/et131x_initpci.c
@@ -806,13 +806,62 @@ static struct pci_device_id et131x_pci_table[] __devinitdata = {
 
 MODULE_DEVICE_TABLE(pci, et131x_pci_table);
 
+static int et131x_stop(struct device *dev)
+{
+	struct pci_dev *pcidev = to_pci_dev(dev);
+	struct net_device *netdev = pci_get_drvdata(pcidev);
+	struct et131x_adapter *adapter = netdev_priv(netdev);
+
+	if (netif_running(netdev))
+		et131x_close(netdev);
+
+	et131x_adapter_memory_free(adapter);
+
+	return 0;
+}
+
+static int et131x_start(struct device *dev)
+{
+	int result;
+	struct pci_dev *pcidev = to_pci_dev(dev);
+	struct net_device *netdev = pci_get_drvdata(pcidev);
+	struct et131x_adapter *adapter = netdev_priv(netdev);
+
+	/* If Phy COMA mode was enabled when we went down, disable it here. */
+	writel(ET_PMCSR_INIT,  &adapter->regs->global.pm_csr);
+
+	et131x_soft_reset(adapter);
+
+	result = et131x_adapter_memory_alloc(adapter);
+	if (result != 0) {
+		dev_warn(&adapter->pdev->dev,
+			 "Resume failed; couldn't re-alloc DMA memory\n");
+		return result;
+	}
+
+	et131x_init_send(adapter);
+	et131x_adapter_setup(adapter);
+
+	if (netif_running(netdev))
+		result = et131x_open(netdev);
+	return result;
+}
+
+static const struct dev_pm_ops et131x_pm = {
+	.freeze = et131x_stop,
+	.poweroff = et131x_stop,
+	.suspend = et131x_stop,
+	.restore = et131x_start,
+	.resume = et131x_start,
+	.thaw = et131x_start,
+};
+
 static struct pci_driver et131x_driver = {
-      .name	= DRIVER_NAME,
-      .id_table	= et131x_pci_table,
-      .probe	= et131x_pci_setup,
-      .remove	= __devexit_p(et131x_pci_remove),
-      .suspend	= NULL,		/* et131x_pci_suspend */
-      .resume	= NULL,		/* et131x_pci_resume */
+	.name		= DRIVER_NAME,
+	.id_table	= et131x_pci_table,
+	.probe		= et131x_pci_setup,
+	.remove		= __devexit_p(et131x_pci_remove),
+	.driver.pm	= &et131x_pm,
 };
 
 
diff --git a/drivers/staging/et131x/et131x_netdev.c b/drivers/staging/et131x/et131x_netdev.c
index b25bae2..8643cd6 100644
--- a/drivers/staging/et131x/et131x_netdev.c
+++ b/drivers/staging/et131x/et131x_netdev.c
@@ -89,8 +89,6 @@
 #include "et131x.h"
 
 struct net_device_stats *et131x_stats(struct net_device *netdev);
-int et131x_open(struct net_device *netdev);
-int et131x_close(struct net_device *netdev);
 int et131x_ioctl(struct net_device *netdev, struct ifreq *reqbuf, int cmd);
 void et131x_multicast(struct net_device *netdev);
 int et131x_tx(struct sk_buff *skb, struct net_device *netdev);
-- 
1.7.4.1

Re: [RFC/RFT PATCH] Staging: et131x: add suspend/resume support

[Greg KH]

On Wed, May 18, 2011 at 10:17:47AM -0300, Herton Ronaldo Krzesinski wrote:
> Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>

I think we need a bit more comments here.

Like have you tested this to verify that it works?  Anything else?

thanks,

greg k-h

Re: [RFC/RFT PATCH] Staging: et131x: add suspend/resume support

[Herton Ronaldo Krzesinski]

On Wed, May 18, 2011 at 08:24:28AM -0700, Greg KH wrote:
> On Wed, May 18, 2011 at 10:17:47AM -0300, Herton Ronaldo Krzesinski wrote:
> > Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>
> 
> I think we need a bit more comments here.
> 
> Like have you tested this to verify that it works?  Anything else?

An user complained about lack of suspend/resume support, so I did the
patch, it's working accordingly to him from what I could get
(unfortunately I can't test myself, don't have the hardware). I want
others to test too, before this can be pushed (RFT on subject). The bug
report I got: http://bugs.launchpad.net/bugs/766549
But good luck in parsing that... I also want others with same
hardware testing this, to confirm it is ok.

May be this isn't important at this stage, the driver may need another
important things done before...

> 
> thanks,
> 
> greg k-h
> 

-- 
[]'s
Herton

Re: [RFC/RFT PATCH] Staging: et131x: add suspend/resume support

[Greg KH]

On Wed, May 18, 2011 at 01:03:17PM -0300, Herton Ronaldo Krzesinski wrote:
> On Wed, May 18, 2011 at 08:24:28AM -0700, Greg KH wrote:
> > On Wed, May 18, 2011 at 10:17:47AM -0300, Herton Ronaldo Krzesinski wrote:
> > > Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>
> > 
> > I think we need a bit more comments here.
> > 
> > Like have you tested this to verify that it works?  Anything else?
> 
> An user complained about lack of suspend/resume support, so I did the
> patch, it's working accordingly to him from what I could get
> (unfortunately I can't test myself, don't have the hardware). I want
> others to test too, before this can be pushed (RFT on subject). The bug
> report I got: http://bugs.launchpad.net/bugs/766549
> But good luck in parsing that... I also want others with same
> hardware testing this, to confirm it is ok.

Ok, then please say so in the patch body itself next time.

I'll wait for you to get some testing, and then resend the patch, before
applying it.

thanks,

greg k-h