Re: [PATCH 02/11] mfd: da9055-core: make it explicitly non-modular

From: Lee Jones
Date: Tue Nov 27 2018 - 08:07:17 EST


On Fri, 23 Nov 2018, Paul Gortmaker wrote:

> [Re: [PATCH 02/11] mfd: da9055-core: make it explicitly non-modular] On 22/11/2018 (Thu 22:14) Paul Gortmaker wrote:
>
> > [Re: [PATCH 02/11] mfd: da9055-core: make it explicitly non-modular] On 23/11/2018 (Fri 10:21) kbuild test robot wrote:
> >
>
> [...]
>
> > >
> > > All errors (new ones prefixed by >>):
> > >
> > > drivers/mfd/da9055-i2c.o: In function `da9055_i2c_remove':
> > > >> drivers/mfd/da9055-i2c.c:53: undefined reference to `da9055_device_exit'
> >
> > Thanks for the report -- I'll look into what causes it, why my testing
> > didn't see it, and get an update to Lee as soon as possible.
>
> OK, mystery solved. I chose this smaller subset of MFD "simple" patches
> from my pending queue of MFD patches - to create a reasonable sized
> maintainer-friendly send, based on patches with zero runtime changes.
>
> My other pending MFD patches have a trivial runtime behavior change;
> deleting a ".remove" field/function - that will never be used for a
> non-module case, but in theory could be (pointlessly) triggered by
> forcing a driver unbind. (see mainline 98b72b94def9 as an example)
> Patches like this were left behind for a future send batch.

What about when .remove() is invoked during shutdown?

> Unfortunately that allowed me to overlook the fact that patch #2 link
> depended on the below ".remove" patch (not sent) to be applied 1st.
>
> Lee, what would you like to have happen? I can resend the queue with
> this patch, or I can resend with #2 being temporarily deferred until
> a future patch batch that has the below da9055-i2c in it, or ...
>
> Whatever is easiest for you - let me know.
>
> Paul.

--
Lee Jones [æçæ]
Linaro Services Technical Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog