Hi, newbie here.
am writing mdio driver for mpc8560 for a new phy. I am referring to gianfar driver codes which is for gigabyte ethernet and some freescale fcc ethernet driver for mpc8540 cpm (which i need to port it to mpc8560 cpm, not sure if they are reusable), but here's my problem. A modular driver is reference and modified as below:

static int __init fs_init(void)
{
...
/* let's insert mii stuff */
r = fs_enet_mdio_bb_init()

if (r != 0) {
printk(KERN_ERR DRV_MODULE_NAME
"BB PHY init failed.\n");
return r;
}
r = driver_register(&fs_enet_fcc_driver);
if (r != 0)
goto err;

...
}

static struct device_driver fs_enet_bb_mdio_driver = {
.name = "fsl-bb-mdio",
.bus = &platform_bus_type,
.probe = fs_enet_mdio_probe,
.remove = fs_enet_mdio_remove,
};

int fs_enet_mdio_bb_init(void)
{
return driver_register(&fs_enet_bb_mdio_driver);
}


my understanding is that once the driver registered, the probe will be called though im not sure what is happening in the background. However, in my testing, eventhough the above fs_enet_mdio_bb_init() is called, the fs_enet_mdio_probe() is not called.

Can anybody explains why is this happening? would be helpful if anybody could explain how the registration triggers the probe??

I strongly recommend you report your thread, and have it moved to either Kernel or Programming, otherwise I very much doubt anyone will give you an answer, because nobody who knows about this stuff will see it.

Thanks texmex, copied the thread into kernel section.

The suggestion was to "report to be moved", not create a dupe
and ask for closure. The post in Linux-Kernel was closed (as
it was the duplicate), and I'm moving this one to PROGRAMMING
because I feel it's more likely to get a coding related response
there. Best of luck.


Cheers,
Tink