Re: [PATCH 1/4] driver: core: add subsys_driver() macro

[Brian Masney <bmasney@redhat.com>]

On Tue, Sep 08, 2026 at 05:58:45PM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Low] Missing semicolons after macro invocations in `subsys_driver`.
> --
> 
> commit cdf3179a9f2205efd1b552f70b7522e1db5ee888
> Author: Brian Masney <bmasney@redhat.com>
> 
> driver: core: add subsys_driver() macro
> 
> Add a new macro subsys_driver() to eliminate some boilerplate code in
> drivers that need to register earlier in the boot at the subsys_initcall
> level.
> 
> > diff --git a/include/linux/device/driver.h b/include/linux/device/driver.h
> > index 768a1334c0a1d..29fbc01ef06fc 100644
> > --- a/include/linux/device/driver.h
> > +++ b/include/linux/device/driver.h
> > @@ -297,4 +297,35 @@ static int __init __driver##_init(void) \
> >  } \
> >  device_initcall(__driver##_init);
> >  
> > +/**
> > + * subsys_driver() - Helper macro for drivers that don't do anything special
> 
> [ ... ]
> 
> > + */
> > +#define subsys_driver(__driver, __register, __unregister, ...) \
> > +static int __init __driver##_init(void) \
> > +{ \
> > +	return __register(&(__driver), ##__VA_ARGS__); \
> > +} \
> > +subsys_initcall(__driver##_init) \
> 
> [Severity: Low]
> Are we missing a semicolon at the end of this line? This relies on underlying
> macros providing a trailing semicolon in their expansions. While it compiles
> today, it makes the macro brittle. If the initcall infrastructure is ever
> refactored to not include an internal trailing semicolon, drivers using
> subsys_driver could fail to compile.

The other code in this file (like module_driver + friends) has a semicolon
however checkpatch.pl complains if the semicolon is there. I didn't know
if the semicolon was legacy. I ended up going with the approach to
satisfy checkpatch.pl.

I have no opinion either way about this.

> 
> > +static void __exit __driver##_exit(void) \
> > +{ \
> > +	__unregister(&(__driver), ##__VA_ARGS__); \
> > +} \
> > +module_exit(__driver##_exit)
> 
> [Severity: Low]
> Could we also add a semicolon at the end of the macro definition? Omitting it
> is inconsistent with similar kernel macros like module_driver and relies entirely
> on how the user invokes the macro.

Same comment as above applies here as well.

Brian