Re: [PATCH 3/3] pci: gt96100eth use pci probing

[Jeff Garzik <jeff@garzik.org>]

Jiri Slaby wrote:
> Jiri Slaby napsal(a):
>> Jeff Garzik napsal(a):
>>> Jiri Slaby wrote:
>>>> +static int __init gt96100_probe1(struct pci_dev *pdev, int port_num)
>>>>  {
>>>>      struct gt96100_private *gp = NULL;
>>>> -    struct gt96100_if_t *gtif = &gt96100_iflist[port_num];
>>>> +    struct gt96100_if_t *gtifs = pci_get_drvdata(pdev);
>>>> +    struct gt96100_if_t *gtif = &gtifs[port_num];
>>>>      int phy_addr, phy_id1, phy_id2;
>>>>      u32 phyAD;
>>>> -    int retval;
>>>> +    int retval = -ENOMEM;
>>>>      unsigned char chip_rev;
>>>>      struct net_device *dev = NULL;
>>>>           if (gtif->irq < 0) {
>>>> -        printk(KERN_ERR "%s: irq unknown - probing not supported\n",
>>>> -              __FUNCTION__);
>>>> +        dev_err(&pdev->dev, "irq unknown - probing not supported\n");
>>>>          return -ENODEV;
>>>>      }
>>>> +
>>>> +    retval = pci_enable_device(pdev);
>>>> +    if (retval) {
>>>> +        dev_err(&pdev->dev, "cannot enable pci device\n");
>>>> +        return retval;
>>>> +    }
>>> bug #1:  please confirm pci_enable_device() is OK on this embedded hardware
>> I do not understand too much, did you mean something like this:
>> } else
>> 	dev_info(..., "pci device is enabled now\n");
>> or?
> Or..., aha, I got it now. How can I confirm?

Asking the maintainer would be a good first step.

	Jeff