Re: [cocci] Adjustments for first statements from selected function implementations?

On Wed, 25 Mar 2026, Markus Elfring wrote:

> Hello!
>
> A contribution like the following shows how a statement can occasionally be added
> to the beginning of a function implementation.
>
> PM: runtime: Mark last busy stamp in pm_request_autosuspend()
> 2025-06-18
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=18c1fe53d186867243f4cf17f4eef60737a16c4c
>
> Such an adjustment triggers further opportunities for corresponding collateral evolution.
> The semantic patch language can help in the determination to some degree
> if the mentioned statement is still specified before a call of the known function
> at other source code places.
> A desire can evolve then to omit repeated statements before more selected function calls.
>
> Usual case distinctions trigger further development challenges.
>
> * How would you prefer to manage function names (for SmPL scripts)?
>
> * Several inline function implementations are also stored in header files.
>   Some header files tend to provide contents which are not really supported
>   by the Coccinelle software so far.
>
>
> Would you be looking then for further possibilities with improving affected
> data processing?

I looked at the link, but I really have no idea what any of this means.

As long as inline functions are parseable, it doesn't matter about the
other contents of header files.

julia