[netdrvr] tulip: oops in tulip_interrupt when hibernating with swsusp/suspend2
authorGrant Grundler <grundler@parisc-linux.org>
Mon, 12 May 2008 06:37:51 +0000 (00:37 -0600)
committerJeff Garzik <jgarzik@redhat.com>
Sat, 31 May 2008 02:10:11 +0000 (22:10 -0400)
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.

Original from:  kernelbugs@tap.homeip.net
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
drivers/net/tulip/tulip_core.c

index f9d13fa05d64dc010fcce435b237db7360df60af..55670b5eb611a0c5c6811f7b6cca3a1b85041963 100644 (file)
@@ -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;