Messages in this thread | | | Date | Mon, 4 Jun 2012 18:45:31 +0200 | From | Alessandro Rubini <> | Subject | Re: [PATCH RFC] c_can_pci: generic module for c_can on PCI |
| |
> Anythign wrong with > > bool aligned32;
I personally think booleans are evil. But both this and the other thing:
>> +static u16 c_can_pci_read_reg_aligned_to_16bit(struct c_can_priv *priv, >> + void *reg) > > I'm a bit worried this function name might be too short ;)
come from the platform driver this is based on (I already blamed federico offlist for not preserving authorship of the original file).
So, this file is mostly copied from the platform driver, which is a duplication of code. A mandated duplication, given how the thing is currently laid out: the c_can core driver exports functions that the other two files are using (the platform and the new pci driver).
In my opinion, it would be much better to have one less layer and no exports at all. The core driver should be a platform driver, and the pci driver would just build platform data and register the platform device.
Sure this isn't up to federico, who has the pci device but cannot access any boards where the previous driver is used. What do the maintainers think? I (or federico :) may propose a reshaping, if the idea makes sense.
/alessandro, another user of the sta2x11 where c_can_pci lives
| |