Re: linux-next: build failure after merge of the devicetree tree

[Rob Herring]

On Thu, Jan 18, 2024 at 6:58 PM Stephen Rothwell <sfr@xxxxxxxxx.orgau> wrote:
>
> Hi all,
>
> On Mon, 11 Dec 2023 16:05:10 +1100 Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
> >
> > After merging the devicetree tree, today's linux-next build (x86_64
> > allmodconfig) failed like this:
> >
> > drivers/clk/qcom/gcc-x1e80100.c:6786:15: error: variable 'gcc_x1e80100_driver' has initializer but incomplete type
> >  6786 | static struct platform_driver gcc_x1e80100_driver = {
> >       |               ^~~~~~~~~~~~~~~
> > drivers/clk/qcom/gcc-x1e80100.c:6787:10: error: 'struct platform_driver' has no member named 'probe'
> >  6787 |         .probe = gcc_x1e80100_probe,
> >       |          ^~~~~
> > drivers/clk/qcom/gcc-x1e80100.c:6787:18: warning: excess elements in struct initializer
> >  6787 |         .probe = gcc_x1e80100_probe,
> >       |                  ^~~~~~~~~~~~~~~~~~
> > drivers/clk/qcom/gcc-x1e80100.c:6787:18: note: (near initialization for 'gcc_x1e80100_driver')
> > drivers/clk/qcom/gcc-x1e80100.c:6788:10: error: 'struct platform_driver' has no member named 'driver'
> >  6788 |         .driver = {
> >       |          ^~~~~~
> > drivers/clk/qcom/gcc-x1e80100.c:6788:19: error: extra brace group at end of initializer
> >  6788 |         .driver = {
> >       |                   ^
> > drivers/clk/qcom/gcc-x1e80100.c:6788:19: note: (near initialization for 'gcc_x1e80100_driver')
> > drivers/clk/qcom/gcc-x1e80100.c:6788:19: warning: excess elements in struct initializer
> > drivers/clk/qcom/gcc-x1e80100.c:6788:19: note: (near initialization for 'gcc_x1e80100_driver')
> > drivers/clk/qcom/gcc-x1e80100.c: In function 'gcc_x1e80100_init':
> > drivers/clk/qcom/gcc-x1e80100.c:6796:16: error: implicit declaration of function 'platform_driver_register' [-Werror=implicit-function-declaration]
> >  6796 |         return platform_driver_register(&gcc_x1e80100_driver);
> >       |                ^~~~~~~~~~~~~~~~~~~~~~~~
> > drivers/clk/qcom/gcc-x1e80100.c: In function 'gcc_x1e80100_exit':
> > drivers/clk/qcom/gcc-x1e80100.c:6802:9: error: implicit declaration of function 'platform_driver_unregister'; did you mean 'driver_unregister'? [-Werror=implicit-function-declaration]
> >  6802 |         platform_driver_unregister(&gcc_x1e80100_driver);
> >       |         ^~~~~~~~~~~~~~~~~~~~~~~~~~
> >       |         driver_unregister
> > drivers/clk/qcom/gcc-x1e80100.c: At top level:
> > drivers/clk/qcom/gcc-x1e80100.c:6786:31: error: storage size of 'gcc_x1e80100_driver' isn't known
> >  6786 | static struct platform_driver gcc_x1e80100_driver = {
> >       |                               ^~~~~~~~~~~~~~~~~~~
> >
> > Caused by commit
> >
> >   0d18bcdebb2f ("of: Stop circularly including of_device.h and of_platform.h")
> >
> > interacting with commit
> >
> >   161b7c401f4b ("clk: qcom: Add Global Clock controller (GCC) driver for X1E80100")
> >
> > from the qcom tree.
> >
> > I have applied the following merge resolution patch.  This patch could
> > be applied to the gcom tree.
> >
> > From: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> > Date: Mon, 11 Dec 2023 15:47:55 +1100
> > Subject: [PATCH] fix up for "of: Stop circularly including of_device.h and of_platform.h"
> >
> > interacting with
> > "clk: qcom: Add Global Clock controller (GCC) driver for X1E80100"
> >
> > Signed-off-by: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> > ---
> >  drivers/clk/qcom/gcc-x1e80100.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/clk/qcom/gcc-x1e80100.c b/drivers/clk/qcom/gcc-x1e80100.c
> > index 74db7fef237b..d7182d6e9783 100644
> > --- a/drivers/clk/qcom/gcc-x1e80100.c
> > +++ b/drivers/clk/qcom/gcc-x1e80100.c
> > @@ -4,8 +4,9 @@
> >   */
> >
> >  #include <linux/clk-provider.h>
> > +#include <linux/mod_devicetable.h>
> >  #include <linux/module.h>
> > -#include <linux/of_device.h>
> > +#include <linux/platform_device.h>
> >  #include <linux/regmap.h>
> >
> >  #include <dt-bindings/clock/qcom,x1e80100-gcc.h>
> > --
> > 2.40.1
>
> Did this get lost somewhere among the merges?  I am still applying the patch
> to linux-next.

Thanks for pointing that out. I guess the QCom folks never applied it.
I'm going to send the final patch to Linus today and will add this one
in.

Rob