PATCH for [Bug 8952] tulip driver oops in tulip_interrupt when hibernating with swsusp/suspend2

PATCH for [Bug 8952] tulip driver oops in tulip_interrupt when hibernating with swsusp/suspend2

[Grant Grundler]

Jeff,
The following patch is seems to fix the tulip suspend/resume panic:
	http://bugzilla.kernel.org/show_bug.cgi?id=8952#c46

My attempts at a cleaner patch failed and Pavel thinks this is OK.
Since suspend/resume is getting an overhaul in 2.6.27 (per comment
#49 by  Rafael J. Wysocki), it makes sense to invest more time as
part of that rework and apply the known fix to 2.6.26.

hth,
grant

Original from:  kernelbugs@tap.homeip.net
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>

diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
index f9d13fa..088d3bf 100644
--- a/drivers/net/tulip/tulip_core.c
+++ b/drivers/net/tulip/tulip_core.c
@@ -1729,12 +1729,15 @@ static int tulip_suspend (struct pci_dev *pdev, pm_message_t state)
 	if (!dev)
 		return -EINVAL;
 
-	if (netif_running(dev))
-		tulip_down(dev);
+	if (!netif_running(dev))
+		goto save_state:
+
+	tulip_down(dev);
 
 	netif_device_detach(dev);
 	free_irq(dev->irq, dev);
 
+save_state:
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 	pci_set_power_state(pdev, pci_choose_state(pdev, state));
@@ -1754,6 +1757,9 @@ static int tulip_resume(struct pci_dev *pdev)
 	pci_set_power_state(pdev, PCI_D0);
 	pci_restore_state(pdev);
 
+	if (!netif_running(dev))
+		return 0;
+
 	if ((retval = pci_enable_device(pdev))) {
 		printk (KERN_ERR "tulip: pci_enable_device failed in resume\n");
 		return retval;

Re: PATCH for [Bug 8952] tulip driver oops in tulip_interrupt when hibernating with swsusp/suspend2

[Jeff Garzik]

Grant Grundler wrote:
> Jeff,
> The following patch is seems to fix the tulip suspend/resume panic:
> 	http://bugzilla.kernel.org/show_bug.cgi?id=8952#c46
> 
> My attempts at a cleaner patch failed and Pavel thinks this is OK.
> Since suspend/resume is getting an overhaul in 2.6.27 (per comment
> #49 by  Rafael J. Wysocki), it makes sense to invest more time as
> part of that rework and apply the known fix to 2.6.26.
> 
> hth,
> grant
> 
> Original from:  kernelbugs@tap.homeip.net
> Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
> 
> diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
> index f9d13fa..088d3bf 100644
> --- a/drivers/net/tulip/tulip_core.c
> +++ b/drivers/net/tulip/tulip_core.c
> @@ -1729,12 +1729,15 @@ static int tulip_suspend (struct pci_dev *pdev, pm_message_t state)
>  	if (!dev)
>  		return -EINVAL;
>  
> -	if (netif_running(dev))
> -		tulip_down(dev);
> +	if (!netif_running(dev))
> +		goto save_state:
> +
> +	tulip_down(dev);

how could this be tested if it doesn't even compile?

Oh well, seems sane, let's see...

Re: PATCH for [Bug 8952] tulip driver oops in tulip_interrupt when hibernating with swsusp/suspend2

[Jeff Garzik]

Grant Grundler wrote:
> Jeff,
> The following patch is seems to fix the tulip suspend/resume panic:
> 	http://bugzilla.kernel.org/show_bug.cgi?id=8952#c46
> 
> My attempts at a cleaner patch failed and Pavel thinks this is OK.
> Since suspend/resume is getting an overhaul in 2.6.27 (per comment
> #49 by  Rafael J. Wysocki), it makes sense to invest more time as
> part of that rework and apply the known fix to 2.6.26.
> 
> hth,
> grant
> 
> Original from:  kernelbugs@tap.homeip.net
> Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
> 
> diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c

applied

Re: PATCH for [Bug 8952] tulip driver oops in tulip_interrupt when hibernating with swsusp/suspend2

[Grant Grundler]

On Fri, May 30, 2008 at 10:09:35PM -0400, Jeff Garzik wrote:
> Grant Grundler wrote:
>> Jeff,
>> The following patch is seems to fix the tulip suspend/resume panic:
>> 	http://bugzilla.kernel.org/show_bug.cgi?id=8952#c46
>> My attempts at a cleaner patch failed and Pavel thinks this is OK.
>> Since suspend/resume is getting an overhaul in 2.6.27 (per comment
>> #49 by  Rafael J. Wysocki), it makes sense to invest more time as
>> part of that rework and apply the known fix to 2.6.26.
>> hth,
>> grant
>> Original from:  kernelbugs@tap.homeip.net
>> Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
>> diff --git a/drivers/net/tulip/tulip_core.c 
>> b/drivers/net/tulip/tulip_core.c
>> index f9d13fa..088d3bf 100644
>> --- a/drivers/net/tulip/tulip_core.c
>> +++ b/drivers/net/tulip/tulip_core.c
>> @@ -1729,12 +1729,15 @@ static int tulip_suspend (struct pci_dev *pdev, 
>> pm_message_t state)
>>  	if (!dev)
>>  		return -EINVAL;
>>  -	if (netif_running(dev))
>> -		tulip_down(dev);
>> +	if (!netif_running(dev))
>> +		goto save_state:
>> +
>> +	tulip_down(dev);
>
> how could this be tested if it doesn't even compile?
> Oh well, seems sane, let's see...

Thanks for fixing up the : vs ;  and applying.
Is there a canonical description of what suspend/resume has to do?

As noted in the bug report, the NIC drivers are very inconsistent
in what they do in these routines.

thanks,
grant