[PATCH 1/2] tulip: explicity set to D0 power state during init

[PATCH 1/2] tulip: explicity set to D0 power state during init

[Steven Walter]

During the first suspend the chip would refuse to enter D3.  Subsequent
suspends worked okay.  During resume the chip is commanded into D0.
Doing so during initialization fixes the initial suspend.

Signed-off-by: Steven Walter <stevenrwalter@gmail.com>
---
 drivers/net/tulip/tulip_core.c |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)

diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
index 3810db9..bb8c0ee 100644
--- a/drivers/net/tulip/tulip_core.c
+++ b/drivers/net/tulip/tulip_core.c
@@ -1380,6 +1380,13 @@ static int __devinit tulip_init_one (struct pci_dev *pdev,
 		return i;
 	}
 
+	/* The chip will fail to enter a low-power state later unless
+	 * first explicitly commanded into D0 */
+	if (pci_set_power_state(pdev, PCI_D0)) {
+		printk (KERN_ERR PFX
+			"Failed to set power state to D0\n");
+	}
+
 	irq = pdev->irq;
 
 	/* alloc_etherdev ensures aligned and zeroed private structures */
-- 
1.6.3.3

Re: [PATCH 1/2] tulip: explicity set to D0 power state during init

[Grant Grundler]

On Mon, May 31, 2010 at 06:34:42PM -0400, Steven Walter wrote:
> During the first suspend the chip would refuse to enter D3.  Subsequent
> suspends worked okay.  During resume the chip is commanded into D0.
> Doing so during initialization fixes the initial suspend.
> 
> Signed-off-by: Steven Walter <stevenrwalter@gmail.com>

Signed-off-by: Grant Grundler <grundler@parisc-linux.org>

> ---
>  drivers/net/tulip/tulip_core.c |    7 +++++++
>  1 files changed, 7 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
> index 3810db9..bb8c0ee 100644
> --- a/drivers/net/tulip/tulip_core.c
> +++ b/drivers/net/tulip/tulip_core.c
> @@ -1380,6 +1380,13 @@ static int __devinit tulip_init_one (struct pci_dev *pdev,
>  		return i;
>  	}
>  
> +	/* The chip will fail to enter a low-power state later unless
> +	 * first explicitly commanded into D0 */
> +	if (pci_set_power_state(pdev, PCI_D0)) {
> +		printk (KERN_ERR PFX

My only quibble is this message really isn't "KERN_ERR" worthy.
Can you explain why you think this should be ERR and not say, KERN_NOTICE?

(I'm looking at the definitions in include/linux/kernel.h of 2.6 source tree.)

If you want to repost with KERN_NOTICE, please include my S-O-B: line above.

thanks,
grant

> +			"Failed to set power state to D0\n");
> +	}
> +
>  	irq = pdev->irq;
>  
>  	/* alloc_etherdev ensures aligned and zeroed private structures */
> -- 
> 1.6.3.3

[PATCH 1/2] tulip: explicity set to D0 power state during init

[Steven Walter]

During the first suspend the chip would refuse to enter D3.  Subsequent
suspends worked okay.  During resume the chip is commanded into D0.
Doing so during initialization fixes the initial suspend.

Signed-off-by: Steven Walter <stevenrwalter@gmail.com>
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
---
 drivers/net/tulip/tulip_core.c |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)

diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
index 3810db9..ec013c2 100644
--- a/drivers/net/tulip/tulip_core.c
+++ b/drivers/net/tulip/tulip_core.c
@@ -1380,6 +1380,13 @@ static int __devinit tulip_init_one (struct pci_dev *pdev,
 		return i;
 	}
 
+	/* The chip will fail to enter a low-power state later unless
+	 * first explicitly commanded into D0 */
+	if (pci_set_power_state(pdev, PCI_D0)) {
+		printk (KERN_NOTICE PFX
+			"Failed to set power state to D0\n");
+	}
+
 	irq = pdev->irq;
 
 	/* alloc_etherdev ensures aligned and zeroed private structures */
-- 
1.6.3.3

Re: [PATCH 1/2] tulip: explicity set to D0 power state during init

[David Miller]

From: Steven Walter <stevenrwalter@gmail.com>
Date: Mon, 31 May 2010 23:11:53 -0400

> During the first suspend the chip would refuse to enter D3.  Subsequent
> suspends worked okay.  During resume the chip is commanded into D0.
> Doing so during initialization fixes the initial suspend.
> 
> Signed-off-by: Steven Walter <stevenrwalter@gmail.com>
> Signed-off-by: Grant Grundler <grundler@parisc-linux.org>

Applied.