[PATCH] pinctrl: fix build failure on PPC

[PATCH] pinctrl: fix build failure on PPC

[Linus Walleij]

From: Linus Walleij <linus.walleij@linaro.org>

commit 23289e6ecc6d ("pinctrl: core device tree mapping table
parsing support") caused a build failure on PPC:

drivers/pinctrl/devicetree.c: In function 'dt_free_map':
drivers/pinctrl/devicetree.c:44:42: error: dereferencing pointer to incomplete type
drivers/pinctrl/devicetree.c:45:6: error: dereferencing pointer to incomplete type
drivers/pinctrl/devicetree.c: In function 'dt_to_map_one_config':
drivers/pinctrl/devicetree.c:140:21: error: dereferencing pointer to incomplete type
drivers/pinctrl/devicetree.c:141:10: error: dereferencing pointer to incomplete type
drivers/pinctrl/devicetree.c:146:11: error: dereferencing pointer to incomplete type

This is probably because the Makefile for pinctrl contains this,
that makes a small piece of the implementation build and link
in isolation:

obj-$(CONFIG_OF)                += devicetree.o

So let's use some local Kconfig option that depends on both OF
and PINCTRL so that this file is only compiled if you have
both enabled.

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Stephen Warren <swarren@wwwdotorg.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
---
 drivers/pinctrl/Kconfig  |    5 +++++
 drivers/pinctrl/Makefile |    2 +-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/pinctrl/Kconfig b/drivers/pinctrl/Kconfig
index abfb964..d68c878 100644
--- a/drivers/pinctrl/Kconfig
+++ b/drivers/pinctrl/Kconfig
@@ -17,6 +17,11 @@ config PINMUX
 config PINCONF
 	bool "Support pin configuration controllers"
 
+config PINCTRL_OF
+	bool
+	depends on OF
+	default y if OF
+
 config GENERIC_PINCONF
 	bool
 	select PINCONF
diff --git a/drivers/pinctrl/Makefile b/drivers/pinctrl/Makefile
index 049c9fb..5412f60 100644
--- a/drivers/pinctrl/Makefile
+++ b/drivers/pinctrl/Makefile
@@ -5,7 +5,7 @@ ccflags-$(CONFIG_DEBUG_PINCTRL)	+= -DDEBUG
 obj-$(CONFIG_PINCTRL)		+= core.o
 obj-$(CONFIG_PINMUX)		+= pinmux.o
 obj-$(CONFIG_PINCONF)		+= pinconf.o
-obj-$(CONFIG_OF)		+= devicetree.o
+obj-$(CONFIG_PINCTRL_OF)	+= devicetree.o
 obj-$(CONFIG_GENERIC_PINCONF)	+= pinconf-generic.o
 obj-$(CONFIG_PINCTRL_PXA3xx)	+= pinctrl-pxa3xx.o
 obj-$(CONFIG_PINCTRL_MMP2)	+= pinctrl-mmp2.o
-- 
1.7.9.2

Re: [PATCH] pinctrl: fix build failure on PPC

[Stephen Warren]

On 04/10/2012 01:52 AM, Linus Walleij wrote:
> From: Linus Walleij <linus.walleij@linaro.org>
> 
> commit 23289e6ecc6d ("pinctrl: core device tree mapping table
> parsing support") caused a build failure on PPC:
> 
> drivers/pinctrl/devicetree.c: In function 'dt_free_map':
> drivers/pinctrl/devicetree.c:44:42: error: dereferencing pointer to incomplete type
> drivers/pinctrl/devicetree.c:45:6: error: dereferencing pointer to incomplete type
> drivers/pinctrl/devicetree.c: In function 'dt_to_map_one_config':
> drivers/pinctrl/devicetree.c:140:21: error: dereferencing pointer to incomplete type
> drivers/pinctrl/devicetree.c:141:10: error: dereferencing pointer to incomplete type
> drivers/pinctrl/devicetree.c:146:11: error: dereferencing pointer to incomplete type
> 
> This is probably because the Makefile for pinctrl contains this,
> that makes a small piece of the implementation build and link
> in isolation:
> 
> obj-$(CONFIG_OF)                += devicetree.o
> 
> So let's use some local Kconfig option that depends on both OF
> and PINCTRL so that this file is only compiled if you have
> both enabled.

It seems simpler to just fix the Makefile to only compile the file when
its needed directly, without introducing a new Kconfig option. See
https://lkml.org/lkml/2012/4/3/445.

Re: [PATCH] pinctrl: fix build failure on PPC

[Linus Walleij]

On Tue, Apr 10, 2012 at 5:14 PM, Stephen Warren <swarren@wwwdotorg.org> wrote:
> On 04/10/2012 01:52 AM, Linus Walleij wrote:

>> So let's use some local Kconfig option that depends on both OF
>> and PINCTRL so that this file is only compiled if you have
>> both enabled.
>
> It seems simpler to just fix the Makefile to only compile the file when
> its needed directly, without introducing a new Kconfig option. See
> https://lkml.org/lkml/2012/4/3/445.

But I think it is basically a config problem, not a compilation problem,
that is why I did it this way.

But I won't be stubborn about it ... I'll see if I can just switch that one
patch out then.

Yours,
Linus Walleij