Re: [PATCH 3/5] staging: tidspbridge: Lindent to drv_interface.c

[Joe Perches <joe@perches.com>]

On Mon, 2012-01-30 at 22:29 +0300, Dan Carpenter wrote:
> On Mon, Jan 30, 2012 at 11:25:34AM -0600, Ramirez Luna, Omar wrote:
> > > +                       pr_info("%s:%d handle(s) still opened\n", __func__,
> > > +                               atomic_read(&bridge_cref));
> > I remember the rule was to break lines as far to the right as
> > possible, no? Chapter 2 CodingStyle, same for the other similar
> > changes.
> It doesn't mean you have to right justify things, it just means
> indented.  The original code is fine here and the new code is fine
> here.  It's up to whoever writes the code to decide.

I concur.

My personal preference is to use a new line after the format
string if necessary.

ie:
	pr_<level>("fmt\n"[, args to 80 columns if all fit])
or
	pr_<level>("fmt\n",
		   args when single line exceeds 80 columns);

So for this case:
			pr_info("%s:%d handle(s) still opened\n",
				__func__, atomic_read(&bridge_cref));

I've done a patch here to tidspbridge that standardizes
printk output.

Basically, the patch adds
#define pr_fmt(fmt) KBUILD_MODNAME "%s: ", __func__
to prefix "tidspbridge:%s:", removes the leading
"%s:...",  __func__ from the uses, coalesces
formats and does argument alignment.

It cleans up the DBC_ASSERT, DBC_REQUIRE and DBC_ENSURE
macros too.

I'm waiting for the Makefile change and whatever
patches Víctor produces to be applied.  I'll then
redo my patch and submit it.