Re: [2.6 patch] parisc: "extern inline" -> "static inline" (fwd)

On Fri, Dec 01, 2006 at 05:54:27PM +0100, Adrian Bunk wrote:
> If you read John David Anglin's email, you'll note that if you take the 
> address you need this function provided somewhere.

Let me turn that around.
Which of the "extern inline" functions are we taking the address?
The parisc kernel wouldn't (shouldn't) link if that were true.

> Which of the functions my patch changes also have a global function 
> provided within the kernel?
> 
> If none, "extern inline" didn't make any sense.

I expect none.

...
> Currently, "inline" is defined to be always_inline, and 
> __always_inline is for cases that produce non-compiling or non-working 
> (opposed to only suboptimal) code.

Ok.  Sounds like "extern inline" is the same as __always_inline.

Has gcc community confirmed "gcc -Wmissing-prototypes" behavior
is really correct with respect to "extern inline"?

If so, I'm ok with changing "extern inline" to __always_inline.

thanks,
grant