On Sun, Jun 21, 2009 at 8:52 PM, Pavel Roskin<proski at gnu.org> wrote: > On Sat, 2009-06-20 at 18:59 -0400, J.R. Mauro wrote: >> > -void hfa384x_create(hfa384x_t *hw, struct usb_device *usb); >> > -void hfa384x_destroy(hfa384x_t *hw); >> > +void hfa384x_create(hfa384x_t * hw, struct usb_device *usb); >> > +void hfa384x_destroy(hfa384x_t * hw); >> > >> > Please revert. :( >> >> Lindent is stupid like this. I think I had some changes for it to fix >> it but I can't remember. I will look. > > Unfortunately, GNU indent cannot figure out that hfa384x_t is a type, > even though it's used in a function declaration. ?It's possible to add > "-T hfa384x_t" to the indent arguments to tell it that hfa384x_t is a > type. > > But the best fix would be to replace hfa384x_t with "struct hfa384x" as > the Linux coding standard requires. ?Then GNU indent would do the right > thing. And now I remember what my fix was: Lindent does not understand the usual typedefs that we use, such as u8, and will mung those as well. My fix simply added a bunch of the standard typedefs to Lindent so that it wouldn't change ``u8 *up'' to ``u8 * up'' As I recall, there was no interest in it on LKML. If anyone thinks it has value, I can dig up the patch if you want to ACK it. > > I would not bother with reverting a patch over such details. > > -- > Regards, > Pavel Roskin >
