* [PATCH 0/2] linux/kconfig.h: move IF_ENABLED() out of <linux/kconfig.h>
@ 2021-04-08 20:58 Masahiro Yamada
2021-04-08 20:58 ` [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h> Masahiro Yamada
0 siblings, 1 reply; 7+ messages in thread
From: Masahiro Yamada @ 2021-04-08 20:58 UTC (permalink / raw)
To: linux-kbuild
Cc: linux-gpio, Arnd Bergmann, Paul Cercueil, Masahiro Yamada,
Len Brown, Linus Walleij, Pavel Machek, Rafael J. Wysocki,
linux-kernel, linux-mips, linux-pm
I insist on <linux/kconfig.h> having only minimal set of macros
that are needed to evaluate CONFIG options.
Everytime somebody added an alien to <linux/kconfig.h>, I needed to
kick it out.
I did not notice 1b399bb04837183cecdc1b32ef1cfc7fcfa75d32 because
I was not addressed by [1].
[1]: https://lore.kernel.org/lkml/?q=kconfig.h%3A+Add+IF_ENABLED%28%29+macro
I like Paul's idea, but if I had noticed the patch in time, I would
have tried my best to persuade to implement it outside of <linux/kconfig.h>
(Paul's initial patch was adding it to a new header instead of <linux/kconfig.h>)
Before it is widely used, I want to fix it.
In 2/2, I converted pm.h to allow driver cleanups.
Masahiro Yamada (2):
linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in
<linux/kernel.h>
pm: allow drivers to drop #ifdef and __maybe_unused from pm callbacks
drivers/pinctrl/pinctrl-ingenic.c | 20 ++++-----
include/linux/kconfig.h | 6 ---
include/linux/kernel.h | 2 +
include/linux/pm.h | 67 +++++++++++--------------------
4 files changed, 36 insertions(+), 59 deletions(-)
--
2.27.0
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-08 20:58 [PATCH 0/2] linux/kconfig.h: move IF_ENABLED() out of <linux/kconfig.h> Masahiro Yamada
@ 2021-04-08 20:58 ` Masahiro Yamada
2021-04-09 8:15 ` Paul Cercueil
2021-04-09 9:24 ` Andy Shevchenko
0 siblings, 2 replies; 7+ messages in thread
From: Masahiro Yamada @ 2021-04-08 20:58 UTC (permalink / raw)
To: linux-kbuild
Cc: linux-gpio, Arnd Bergmann, Paul Cercueil, Masahiro Yamada,
Linus Walleij, linux-kernel, linux-mips
<linux/kconfig.h> is included from all the kernel-space source files,
including C, assembly, linker scripts. It is intended to contain minimal
set of macros to evaluate CONFIG options.
IF_ENABLED() is an intruder here because (x ? y : z) is C code, which
should not be included from assembly files or linker scripts.
Also, <linux/kconfig.h> is no longer self-contained because NULL is
defined in <linux/stddef.h>.
Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
sub-systems often define dedicated macros such as of_match_ptr(),
pm_ptr() etc. for common use-cases.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
---
drivers/pinctrl/pinctrl-ingenic.c | 20 ++++++++++----------
include/linux/kconfig.h | 6 ------
include/linux/kernel.h | 2 ++
3 files changed, 12 insertions(+), 16 deletions(-)
diff --git a/drivers/pinctrl/pinctrl-ingenic.c b/drivers/pinctrl/pinctrl-ingenic.c
index f2746125b077..b21e2ae4528d 100644
--- a/drivers/pinctrl/pinctrl-ingenic.c
+++ b/drivers/pinctrl/pinctrl-ingenic.c
@@ -2496,43 +2496,43 @@ static int __init ingenic_pinctrl_probe(struct platform_device *pdev)
static const struct of_device_id ingenic_pinctrl_of_match[] = {
{
.compatible = "ingenic,jz4740-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4740, &jz4740_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4740), &jz4740_chip_info)
},
{
.compatible = "ingenic,jz4725b-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4725B, &jz4725b_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4725B), &jz4725b_chip_info)
},
{
.compatible = "ingenic,jz4760-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
},
{
.compatible = "ingenic,jz4760b-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
},
{
.compatible = "ingenic,jz4770-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4770, &jz4770_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4770), &jz4770_chip_info)
},
{
.compatible = "ingenic,jz4780-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_JZ4780, &jz4780_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4780), &jz4780_chip_info)
},
{
.compatible = "ingenic,x1000-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
},
{
.compatible = "ingenic,x1000e-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
},
{
.compatible = "ingenic,x1500-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_X1500, &x1500_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1500), &x1500_chip_info)
},
{
.compatible = "ingenic,x1830-pinctrl",
- .data = IF_ENABLED(CONFIG_MACH_X1830, &x1830_chip_info)
+ .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1830), &x1830_chip_info)
},
{ /* sentinel */ },
};
diff --git a/include/linux/kconfig.h b/include/linux/kconfig.h
index 24a59cb06963..cc8fa109cfa3 100644
--- a/include/linux/kconfig.h
+++ b/include/linux/kconfig.h
@@ -70,10 +70,4 @@
*/
#define IS_ENABLED(option) __or(IS_BUILTIN(option), IS_MODULE(option))
-/*
- * IF_ENABLED(CONFIG_FOO, ptr) evaluates to (ptr) if CONFIG_FOO is set to 'y'
- * or 'm', NULL otherwise.
- */
-#define IF_ENABLED(option, ptr) (IS_ENABLED(option) ? (ptr) : NULL)
-
#endif /* __LINUX_KCONFIG_H */
diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index 5b7ed6dc99ac..8685ca4cf287 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -38,6 +38,8 @@
#define PTR_ALIGN_DOWN(p, a) ((typeof(p))ALIGN_DOWN((unsigned long)(p), (a)))
#define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) - 1)) == 0)
+#define PTR_IF(cond, ptr) ((cond) ? (ptr) : NULL)
+
/* generic data direction definitions */
#define READ 0
#define WRITE 1
--
2.27.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-08 20:58 ` [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h> Masahiro Yamada
@ 2021-04-09 8:15 ` Paul Cercueil
2021-04-09 8:30 ` Masahiro Yamada
2021-04-09 9:24 ` Andy Shevchenko
1 sibling, 1 reply; 7+ messages in thread
From: Paul Cercueil @ 2021-04-09 8:15 UTC (permalink / raw)
To: Masahiro Yamada
Cc: linux-kbuild, linux-gpio, Arnd Bergmann, Linus Walleij,
linux-kernel, linux-mips
Hi Masahiro,
Le ven. 9 avril 2021 à 5:58, Masahiro Yamada <masahiroy@kernel.org> a
écrit :
> <linux/kconfig.h> is included from all the kernel-space source files,
> including C, assembly, linker scripts. It is intended to contain
> minimal
> set of macros to evaluate CONFIG options.
>
> IF_ENABLED() is an intruder here because (x ? y : z) is C code, which
> should not be included from assembly files or linker scripts.
>
> Also, <linux/kconfig.h> is no longer self-contained because NULL is
> defined in <linux/stddef.h>.
>
> Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
>
> PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
> IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
> sub-systems often define dedicated macros such as of_match_ptr(),
> pm_ptr() etc. for common use-cases.
What's the idea behind changing IF_ENABLED() to PTR_IF()? You didn't
explain that. What's wrong with IF_ENABLED()?
Cheers,
-Paul
> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> ---
>
> drivers/pinctrl/pinctrl-ingenic.c | 20 ++++++++++----------
> include/linux/kconfig.h | 6 ------
> include/linux/kernel.h | 2 ++
> 3 files changed, 12 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/pinctrl/pinctrl-ingenic.c
> b/drivers/pinctrl/pinctrl-ingenic.c
> index f2746125b077..b21e2ae4528d 100644
> --- a/drivers/pinctrl/pinctrl-ingenic.c
> +++ b/drivers/pinctrl/pinctrl-ingenic.c
> @@ -2496,43 +2496,43 @@ static int __init
> ingenic_pinctrl_probe(struct platform_device *pdev)
> static const struct of_device_id ingenic_pinctrl_of_match[] = {
> {
> .compatible = "ingenic,jz4740-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4740, &jz4740_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4740), &jz4740_chip_info)
> },
> {
> .compatible = "ingenic,jz4725b-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4725B, &jz4725b_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4725B), &jz4725b_chip_info)
> },
> {
> .compatible = "ingenic,jz4760-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
> },
> {
> .compatible = "ingenic,jz4760b-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
> },
> {
> .compatible = "ingenic,jz4770-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4770, &jz4770_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4770), &jz4770_chip_info)
> },
> {
> .compatible = "ingenic,jz4780-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_JZ4780, &jz4780_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4780), &jz4780_chip_info)
> },
> {
> .compatible = "ingenic,x1000-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
> },
> {
> .compatible = "ingenic,x1000e-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
> },
> {
> .compatible = "ingenic,x1500-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_X1500, &x1500_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1500), &x1500_chip_info)
> },
> {
> .compatible = "ingenic,x1830-pinctrl",
> - .data = IF_ENABLED(CONFIG_MACH_X1830, &x1830_chip_info)
> + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1830), &x1830_chip_info)
> },
> { /* sentinel */ },
> };
> diff --git a/include/linux/kconfig.h b/include/linux/kconfig.h
> index 24a59cb06963..cc8fa109cfa3 100644
> --- a/include/linux/kconfig.h
> +++ b/include/linux/kconfig.h
> @@ -70,10 +70,4 @@
> */
> #define IS_ENABLED(option) __or(IS_BUILTIN(option),
> IS_MODULE(option))
>
> -/*
> - * IF_ENABLED(CONFIG_FOO, ptr) evaluates to (ptr) if CONFIG_FOO is
> set to 'y'
> - * or 'm', NULL otherwise.
> - */
> -#define IF_ENABLED(option, ptr) (IS_ENABLED(option) ? (ptr) : NULL)
> -
> #endif /* __LINUX_KCONFIG_H */
> diff --git a/include/linux/kernel.h b/include/linux/kernel.h
> index 5b7ed6dc99ac..8685ca4cf287 100644
> --- a/include/linux/kernel.h
> +++ b/include/linux/kernel.h
> @@ -38,6 +38,8 @@
> #define PTR_ALIGN_DOWN(p, a) ((typeof(p))ALIGN_DOWN((unsigned
> long)(p), (a)))
> #define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) - 1)) == 0)
>
> +#define PTR_IF(cond, ptr) ((cond) ? (ptr) : NULL)
> +
> /* generic data direction definitions */
> #define READ 0
> #define WRITE 1
> --
> 2.27.0
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-09 8:15 ` Paul Cercueil
@ 2021-04-09 8:30 ` Masahiro Yamada
2021-04-09 9:38 ` Paul Cercueil
0 siblings, 1 reply; 7+ messages in thread
From: Masahiro Yamada @ 2021-04-09 8:30 UTC (permalink / raw)
To: Paul Cercueil
Cc: Linux Kbuild mailing list, open list:GPIO SUBSYSTEM,
Arnd Bergmann, Linus Walleij, Linux Kernel Mailing List,
linux-mips
On Fri, Apr 9, 2021 at 5:15 PM Paul Cercueil <paul@crapouillou.net> wrote:
>
> Hi Masahiro,
>
> Le ven. 9 avril 2021 à 5:58, Masahiro Yamada <masahiroy@kernel.org> a
> écrit :
> > <linux/kconfig.h> is included from all the kernel-space source files,
> > including C, assembly, linker scripts. It is intended to contain
> > minimal
> > set of macros to evaluate CONFIG options.
> >
> > IF_ENABLED() is an intruder here because (x ? y : z) is C code, which
> > should not be included from assembly files or linker scripts.
> >
> > Also, <linux/kconfig.h> is no longer self-contained because NULL is
> > defined in <linux/stddef.h>.
> >
> > Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
> >
> > PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
> > IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
> > sub-systems often define dedicated macros such as of_match_ptr(),
> > pm_ptr() etc. for common use-cases.
>
> What's the idea behind changing IF_ENABLED() to PTR_IF()? You didn't
> explain that. What's wrong with IF_ENABLED()?
PTR_IF() is a more generalized variant, which I believe is
a better fit in <linux/kernel.h>
The first parameter does not need to be a CONFIG option,
but any expression.
> Cheers,
> -Paul
>
> > Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> > ---
> >
> > drivers/pinctrl/pinctrl-ingenic.c | 20 ++++++++++----------
> > include/linux/kconfig.h | 6 ------
> > include/linux/kernel.h | 2 ++
> > 3 files changed, 12 insertions(+), 16 deletions(-)
> >
> > diff --git a/drivers/pinctrl/pinctrl-ingenic.c
> > b/drivers/pinctrl/pinctrl-ingenic.c
> > index f2746125b077..b21e2ae4528d 100644
> > --- a/drivers/pinctrl/pinctrl-ingenic.c
> > +++ b/drivers/pinctrl/pinctrl-ingenic.c
> > @@ -2496,43 +2496,43 @@ static int __init
> > ingenic_pinctrl_probe(struct platform_device *pdev)
> > static const struct of_device_id ingenic_pinctrl_of_match[] = {
> > {
> > .compatible = "ingenic,jz4740-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4740, &jz4740_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4740), &jz4740_chip_info)
> > },
> > {
> > .compatible = "ingenic,jz4725b-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4725B, &jz4725b_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4725B), &jz4725b_chip_info)
> > },
> > {
> > .compatible = "ingenic,jz4760-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
> > },
> > {
> > .compatible = "ingenic,jz4760b-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4760, &jz4760_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760), &jz4760_chip_info)
> > },
> > {
> > .compatible = "ingenic,jz4770-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4770, &jz4770_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4770), &jz4770_chip_info)
> > },
> > {
> > .compatible = "ingenic,jz4780-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_JZ4780, &jz4780_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4780), &jz4780_chip_info)
> > },
> > {
> > .compatible = "ingenic,x1000-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
> > },
> > {
> > .compatible = "ingenic,x1000e-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_X1000, &x1000_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000), &x1000_chip_info)
> > },
> > {
> > .compatible = "ingenic,x1500-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_X1500, &x1500_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1500), &x1500_chip_info)
> > },
> > {
> > .compatible = "ingenic,x1830-pinctrl",
> > - .data = IF_ENABLED(CONFIG_MACH_X1830, &x1830_chip_info)
> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1830), &x1830_chip_info)
> > },
> > { /* sentinel */ },
> > };
> > diff --git a/include/linux/kconfig.h b/include/linux/kconfig.h
> > index 24a59cb06963..cc8fa109cfa3 100644
> > --- a/include/linux/kconfig.h
> > +++ b/include/linux/kconfig.h
> > @@ -70,10 +70,4 @@
> > */
> > #define IS_ENABLED(option) __or(IS_BUILTIN(option),
> > IS_MODULE(option))
> >
> > -/*
> > - * IF_ENABLED(CONFIG_FOO, ptr) evaluates to (ptr) if CONFIG_FOO is
> > set to 'y'
> > - * or 'm', NULL otherwise.
> > - */
> > -#define IF_ENABLED(option, ptr) (IS_ENABLED(option) ? (ptr) : NULL)
> > -
> > #endif /* __LINUX_KCONFIG_H */
> > diff --git a/include/linux/kernel.h b/include/linux/kernel.h
> > index 5b7ed6dc99ac..8685ca4cf287 100644
> > --- a/include/linux/kernel.h
> > +++ b/include/linux/kernel.h
> > @@ -38,6 +38,8 @@
> > #define PTR_ALIGN_DOWN(p, a) ((typeof(p))ALIGN_DOWN((unsigned
> > long)(p), (a)))
> > #define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) - 1)) == 0)
> >
> > +#define PTR_IF(cond, ptr) ((cond) ? (ptr) : NULL)
> > +
> > /* generic data direction definitions */
> > #define READ 0
> > #define WRITE 1
> > --
> > 2.27.0
> >
>
>
--
Best Regards
Masahiro Yamada
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-08 20:58 ` [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h> Masahiro Yamada
2021-04-09 8:15 ` Paul Cercueil
@ 2021-04-09 9:24 ` Andy Shevchenko
2021-04-10 3:52 ` Masahiro Yamada
1 sibling, 1 reply; 7+ messages in thread
From: Andy Shevchenko @ 2021-04-09 9:24 UTC (permalink / raw)
To: Masahiro Yamada
Cc: Linux Kbuild mailing list, open list:GPIO SUBSYSTEM,
Arnd Bergmann, Paul Cercueil, Linus Walleij,
Linux Kernel Mailing List, linux-mips
On Fri, Apr 9, 2021 at 12:00 AM Masahiro Yamada <masahiroy@kernel.org> wrote:
>
> <linux/kconfig.h> is included from all the kernel-space source files,
> including C, assembly, linker scripts. It is intended to contain minimal
a minimal
> set of macros to evaluate CONFIG options.
>
> IF_ENABLED() is an intruder here because (x ? y : z) is C code, which
> should not be included from assembly files or linker scripts.
>
> Also, <linux/kconfig.h> is no longer self-contained because NULL is
> defined in <linux/stddef.h>.
>
> Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
>
> PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
> IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
> sub-systems often define dedicated macros such as of_match_ptr(),
> pm_ptr() etc. for common use-cases.
> include/linux/kernel.h | 2 ++
Why kernel.h? Shouldn't it belong to a particular domain with a
respective header file?
Really what we have in the kernel.h right now is a complete train
wreck of something.
We have to define what exactly is kernel.h for?
Arnd? Others? Shall we start a wider discussion on the topic?
--
With Best Regards,
Andy Shevchenko
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-09 8:30 ` Masahiro Yamada
@ 2021-04-09 9:38 ` Paul Cercueil
0 siblings, 0 replies; 7+ messages in thread
From: Paul Cercueil @ 2021-04-09 9:38 UTC (permalink / raw)
To: Masahiro Yamada
Cc: Linux Kbuild mailing list, GPIO SUBSYSTEM, Arnd Bergmann,
Linus Walleij, Linux Kernel Mailing List, linux-mips
Le ven. 9 avril 2021 à 17:30, Masahiro Yamada <masahiroy@kernel.org> a
écrit :
> On Fri, Apr 9, 2021 at 5:15 PM Paul Cercueil <paul@crapouillou.net>
> wrote:
>>
>> Hi Masahiro,
>>
>> Le ven. 9 avril 2021 à 5:58, Masahiro Yamada
>> <masahiroy@kernel.org> a
>> écrit :
>> > <linux/kconfig.h> is included from all the kernel-space source
>> files,
>> > including C, assembly, linker scripts. It is intended to contain
>> > minimal
>> > set of macros to evaluate CONFIG options.
>> >
>> > IF_ENABLED() is an intruder here because (x ? y : z) is C code,
>> which
>> > should not be included from assembly files or linker scripts.
>> >
>> > Also, <linux/kconfig.h> is no longer self-contained because NULL
>> is
>> > defined in <linux/stddef.h>.
>> >
>> > Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
>> >
>> > PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
>> > IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
>> > sub-systems often define dedicated macros such as of_match_ptr(),
>> > pm_ptr() etc. for common use-cases.
>>
>> What's the idea behind changing IF_ENABLED() to PTR_IF()? You didn't
>> explain that. What's wrong with IF_ENABLED()?
>
>
> PTR_IF() is a more generalized variant, which I believe is
> a better fit in <linux/kernel.h>
> The first parameter does not need to be a CONFIG option,
> but any expression.
Fair enough, but we could still have IF_ENABLED as a specialized
variant of PTR_IF:
#define IF_ENABLED(cfg, ptr) PTR_IF(IS_ENABLED(cfg), (ptr))
-Paul
>
>
>> Cheers,
>> -Paul
>>
>> > Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
>> > ---
>> >
>> > drivers/pinctrl/pinctrl-ingenic.c | 20 ++++++++++----------
>> > include/linux/kconfig.h | 6 ------
>> > include/linux/kernel.h | 2 ++
>> > 3 files changed, 12 insertions(+), 16 deletions(-)
>> >
>> > diff --git a/drivers/pinctrl/pinctrl-ingenic.c
>> > b/drivers/pinctrl/pinctrl-ingenic.c
>> > index f2746125b077..b21e2ae4528d 100644
>> > --- a/drivers/pinctrl/pinctrl-ingenic.c
>> > +++ b/drivers/pinctrl/pinctrl-ingenic.c
>> > @@ -2496,43 +2496,43 @@ static int __init
>> > ingenic_pinctrl_probe(struct platform_device *pdev)
>> > static const struct of_device_id ingenic_pinctrl_of_match[] = {
>> > {
>> > .compatible = "ingenic,jz4740-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4740,
>> &jz4740_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4740),
>> &jz4740_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,jz4725b-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4725B,
>> &jz4725b_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4725B),
>> &jz4725b_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,jz4760-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4760,
>> &jz4760_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760),
>> &jz4760_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,jz4760b-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4760,
>> &jz4760_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4760),
>> &jz4760_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,jz4770-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4770,
>> &jz4770_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4770),
>> &jz4770_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,jz4780-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_JZ4780,
>> &jz4780_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_JZ4780),
>> &jz4780_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,x1000-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_X1000,
>> &x1000_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000),
>> &x1000_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,x1000e-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_X1000,
>> &x1000_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1000),
>> &x1000_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,x1500-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_X1500,
>> &x1500_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1500),
>> &x1500_chip_info)
>> > },
>> > {
>> > .compatible = "ingenic,x1830-pinctrl",
>> > - .data = IF_ENABLED(CONFIG_MACH_X1830,
>> &x1830_chip_info)
>> > + .data = PTR_IF(IS_ENABLED(CONFIG_MACH_X1830),
>> &x1830_chip_info)
>> > },
>> > { /* sentinel */ },
>> > };
>> > diff --git a/include/linux/kconfig.h b/include/linux/kconfig.h
>> > index 24a59cb06963..cc8fa109cfa3 100644
>> > --- a/include/linux/kconfig.h
>> > +++ b/include/linux/kconfig.h
>> > @@ -70,10 +70,4 @@
>> > */
>> > #define IS_ENABLED(option) __or(IS_BUILTIN(option),
>> > IS_MODULE(option))
>> >
>> > -/*
>> > - * IF_ENABLED(CONFIG_FOO, ptr) evaluates to (ptr) if CONFIG_FOO
>> is
>> > set to 'y'
>> > - * or 'm', NULL otherwise.
>> > - */
>> > -#define IF_ENABLED(option, ptr) (IS_ENABLED(option) ? (ptr) :
>> NULL)
>> > -
>> > #endif /* __LINUX_KCONFIG_H */
>> > diff --git a/include/linux/kernel.h b/include/linux/kernel.h
>> > index 5b7ed6dc99ac..8685ca4cf287 100644
>> > --- a/include/linux/kernel.h
>> > +++ b/include/linux/kernel.h
>> > @@ -38,6 +38,8 @@
>> > #define PTR_ALIGN_DOWN(p, a) ((typeof(p))ALIGN_DOWN((unsigned
>> > long)(p), (a)))
>> > #define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) -
>> 1)) == 0)
>> >
>> > +#define PTR_IF(cond, ptr) ((cond) ? (ptr) : NULL)
>> > +
>> > /* generic data direction definitions */
>> > #define READ 0
>> > #define WRITE 1
>> > --
>> > 2.27.0
>> >
>>
>>
>
>
> --
> Best Regards
> Masahiro Yamada
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h>
2021-04-09 9:24 ` Andy Shevchenko
@ 2021-04-10 3:52 ` Masahiro Yamada
0 siblings, 0 replies; 7+ messages in thread
From: Masahiro Yamada @ 2021-04-10 3:52 UTC (permalink / raw)
To: Andy Shevchenko
Cc: Linux Kbuild mailing list, open list:GPIO SUBSYSTEM,
Arnd Bergmann, Paul Cercueil, Linus Walleij,
Linux Kernel Mailing List, linux-mips
On Fri, Apr 9, 2021 at 6:24 PM Andy Shevchenko
<andy.shevchenko@gmail.com> wrote:
>
> On Fri, Apr 9, 2021 at 12:00 AM Masahiro Yamada <masahiroy@kernel.org> wrote:
> >
> > <linux/kconfig.h> is included from all the kernel-space source files,
> > including C, assembly, linker scripts. It is intended to contain minimal
>
> a minimal
>
> > set of macros to evaluate CONFIG options.
> >
> > IF_ENABLED() is an intruder here because (x ? y : z) is C code, which
> > should not be included from assembly files or linker scripts.
> >
> > Also, <linux/kconfig.h> is no longer self-contained because NULL is
> > defined in <linux/stddef.h>.
> >
> > Move IF_ENABLED() out to <linux/kernel.h> as PTR_IF().
> >
> > PTR_IF(IS_ENABLED(CONFIG_FOO), ...) is slightly longer than
> > IF_ENABLED(CONFIG_FOO, ...), but it is not a big deal because
> > sub-systems often define dedicated macros such as of_match_ptr(),
> > pm_ptr() etc. for common use-cases.
>
> > include/linux/kernel.h | 2 ++
>
> Why kernel.h? Shouldn't it belong to a particular domain with a
> respective header file?
>
> Really what we have in the kernel.h right now is a complete train
> wreck of something.
> We have to define what exactly is kernel.h for?
<linux/kernel.h> contains random utility macros.
I did not find a good header to put it in otherwise.
>
> Arnd? Others? Shall we start a wider discussion on the topic?
>
> --
> With Best Regards,
> Andy Shevchenko
--
Best Regards
Masahiro Yamada
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2021-04-10 3:53 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2021-04-08 20:58 [PATCH 0/2] linux/kconfig.h: move IF_ENABLED() out of <linux/kconfig.h> Masahiro Yamada
2021-04-08 20:58 ` [PATCH 1/2] linux/kconfig.h: replace IF_ENABLED() with PTR_IF() in <linux/kernel.h> Masahiro Yamada
2021-04-09 8:15 ` Paul Cercueil
2021-04-09 8:30 ` Masahiro Yamada
2021-04-09 9:38 ` Paul Cercueil
2021-04-09 9:24 ` Andy Shevchenko
2021-04-10 3:52 ` Masahiro Yamada
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox