On 2014-09-09 12:26, Masahiro Yamada wrote:
> Kconfig never defines CONFIG_* as 'n'.
> Now obj-n is only used in firmware/Makefile and it can be
> replaced with obj-. No makefile uses lib-n.
BTW, there are a few more instances of *-n, none of which seem useful:
$ git --no-pager grep -e '[a-z]-n\>' -- '*Makefile*' '*Kbuild*'
Makefile: $(init-n) $(init-) \
Makefile: $(core-n) $(core-) $(drivers-n) $(drivers-) \
Makefile: $(net-n) $(net-) $(libs-n) $(libs-))))
arch/arm/mach-ixp4xx/Makefile:obj-pci-n :=
arch/powerpc/boot/Makefile:initrd- := $(patsubst zImage%,
zImage.initrd%, $(image-n) $(image-))
Michal
Hi Michal,
On Thu, 02 Oct 2014 15:04:44 +0200
Michal Marek <[email protected]> wrote:
> On 2014-09-09 12:26, Masahiro Yamada wrote:
> > Kconfig never defines CONFIG_* as 'n'.
> > Now obj-n is only used in firmware/Makefile and it can be
> > replaced with obj-. No makefile uses lib-n.
>
> BTW, there are a few more instances of *-n, none of which seem useful:
>
> $ git --no-pager grep -e '[a-z]-n\>' -- '*Makefile*' '*Kbuild*'
> Makefile: $(init-n) $(init-) \
> Makefile: $(core-n) $(core-) $(drivers-n) $(drivers-) \
> Makefile: $(net-n) $(net-) $(libs-n) $(libs-))))
> arch/arm/mach-ixp4xx/Makefile:obj-pci-n :=
> arch/powerpc/boot/Makefile:initrd- := $(patsubst zImage%,
> zImage.initrd%, $(image-n) $(image-))
Uh, I missed them.
Anyway, it should be done in a separate patch
even if we fix them.
Can you fix them, please?
(I do not want to steal what you found.)
Simpler makefiles seem better.
Best Regards
Masahiro Yamada
Kconfig options are either 'y', 'm', or undefined.
Signed-off-by: Michal Marek <[email protected]>
---
Makefile | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/Makefile b/Makefile
index efbd2cb..106f300 100644
--- a/Makefile
+++ b/Makefile
@@ -868,9 +868,7 @@ vmlinux-dirs := $(patsubst %/,%,$(filter %/, $(init-y) $(init-m) \
$(net-y) $(net-m) $(libs-y) $(libs-m)))
vmlinux-alldirs := $(sort $(vmlinux-dirs) $(patsubst %/,%,$(filter %/, \
- $(init-n) $(init-) \
- $(core-n) $(core-) $(drivers-n) $(drivers-) \
- $(net-n) $(net-) $(libs-n) $(libs-))))
+ $(init-) $(core-) $(drivers-) $(net-) $(libs-))))
init-y := $(patsubst %/, %/built-in.o, $(init-y))
core-y := $(patsubst %/, %/built-in.o, $(core-y))
--
1.8.4.5
On 2014-10-02 15:33, Michal Marek wrote:
> Kconfig options are either 'y', 'm', or undefined.
>
> Signed-off-by: Michal Marek <[email protected]>
> ---
> Makefile | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/Makefile b/Makefile
> index efbd2cb..106f300 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -868,9 +868,7 @@ vmlinux-dirs := $(patsubst %/,%,$(filter %/, $(init-y) $(init-m) \
> $(net-y) $(net-m) $(libs-y) $(libs-m)))
>
> vmlinux-alldirs := $(sort $(vmlinux-dirs) $(patsubst %/,%,$(filter %/, \
> - $(init-n) $(init-) \
> - $(core-n) $(core-) $(drivers-n) $(drivers-) \
> - $(net-n) $(net-) $(libs-n) $(libs-))))
> + $(init-) $(core-) $(drivers-) $(net-) $(libs-))))
Actually, all the directories are hardcoded in
{init,code,drivers,net,libs}-y; there are currently no Kconfig variables
involved. But let's keep this generic and handle *-m and *- just in case.
Michal
Hi Michal,
On Thu, 02 Oct 2014 15:35:35 +0200
Michal Marek <[email protected]> wrote:
> On 2014-10-02 15:33, Michal Marek wrote:
> > Kconfig options are either 'y', 'm', or undefined.
> >
> > Signed-off-by: Michal Marek <[email protected]>
> > ---
> > Makefile | 4 +---
> > 1 file changed, 1 insertion(+), 3 deletions(-)
> >
> > diff --git a/Makefile b/Makefile
> > index efbd2cb..106f300 100644
> > --- a/Makefile
> > +++ b/Makefile
> > @@ -868,9 +868,7 @@ vmlinux-dirs := $(patsubst %/,%,$(filter %/, $(init-y) $(init-m) \
> > $(net-y) $(net-m) $(libs-y) $(libs-m)))
> >
> > vmlinux-alldirs := $(sort $(vmlinux-dirs) $(patsubst %/,%,$(filter %/, \
> > - $(init-n) $(init-) \
> > - $(core-n) $(core-) $(drivers-n) $(drivers-) \
> > - $(net-n) $(net-) $(libs-n) $(libs-))))
> > + $(init-) $(core-) $(drivers-) $(net-) $(libs-))))
>
> Actually, all the directories are hardcoded in
> {init,code,drivers,net,libs}-y; there are currently no Kconfig variables
> involved. But let's keep this generic and handle *-m and *- just in case.
>
Agreed and thank you!
Best Regards
Masahiro Yamada