Config Items appearing twice in same Kconfig file?

Config Items appearing twice in same Kconfig file?

[Christoph Egger]

Hi all!

   First of all sorry for the huge CC, I couldn't come up with a way
to create a smaller list while reaching all the relevant people.

   We, the vamos project[1] at the Friedrich Alexander University are
analything the linux configuration model trying to detect
errors. During a run with our tools we noticed  the fact that the
option ALPHA_LEGACY_START_ADDRESS in arch/alpha/Kconfig.debug is
listed twice with contradictory dependencies on ALPHA_GENERIC.

   Taking a closer look[2] we noticed 22 occurences where the same
config option is defined more than once in the same Kconfig file, not
all of them are, of course, contradictory as the ALPHA sample.

   We"d be interested in reasons to keep this configuration split up
in the same file or if this is just a error introduced someeewhere in
linux' development that should be fixed.

Regards

   Chritoph Egger

[ Please keep vamos@ Cc-ed -- we're not following all the lkml traffic
  ;) no need to keep me though ]

[1] http://vamos1.informatik.uni-erlangen.de
[2] grep  ^config **/Kconfig* | sort | uniq -d -c

Re: Config Items appearing twice in same Kconfig file?

[Christoph Egger]

Just noticed it might have been worth adding the output so everyone
out there doesn't ned to grep through his tree for reference:

% grep  ^config **/Kconfig* | sort | uniq -dc                               
      2 arch/alpha/Kconfig.debug:config ALPHA_LEGACY_START_ADDRESS
      2 arch/alpha/Kconfig:config ALPHA_EV4
      3 arch/alpha/Kconfig:config ALPHA_EV56
      2 arch/alpha/Kconfig:config ALPHA_GAMMA
      2 arch/blackfin/Kconfig:config MEM_MT48LC32M8A2_75
      2 arch/cris/arch-v32/drivers/Kconfig:config ETRAX_RS485
      2 arch/cris/arch-v32/drivers/Kconfig:config ETRAX_RS485_DISABLE_RECEIVER
      2 arch/cris/arch-v32/drivers/Kconfig:config ETRAX_SER3_CD_BIT
      2 arch/h8300/Kconfig.ide:config H8300_IDE_ALT
      2 arch/h8300/Kconfig.ide:config H8300_IDE_BASE
      2 arch/h8300/Kconfig.ide:config H8300_IDE_IRQ
      2 arch/ia64/Kconfig:config IOMMU_HELPER
      2 arch/powerpc/Kconfig:config KERNEL_START
      2 arch/powerpc/Kconfig:config PAGE_OFFSET
      2 arch/powerpc/Kconfig:config PHYSICAL_START
      2 arch/powerpc/Kconfig:config RELOCATABLE
      2 arch/score/Kconfig:config RWSEM_GENERIC_SPINLOCK
      4 arch/x86/Kconfig.debug:config DEFAULT_IO_DELAY_TYPE
      2 arch/x86/Kconfig:config PCI_MMCONFIG
      2 arch/x86/Kconfig:config X86_EXTENDED_PLATFORM
      2 drivers/isdn/hisax/Kconfig:config HISAX_AVM_A1_PCMCIA
      2 net/ipv4/Kconfig:config TCP_CONG_CUBIC

Re: Config Items appearing twice in same Kconfig file?

[Michael Ellerman]

[-- Attachment #1: Type: text/plain, Size: 994 bytes --]

On Wed, 2010-05-12 at 16:40 +0200, Christoph Egger wrote:
> Just noticed it might have been worth adding the output so everyone
> out there doesn't ned to grep through his tree for reference:
> 
> % grep  ^config **/Kconfig* | sort | uniq -dc                               

That didn't actually work for me?

>       2 arch/powerpc/Kconfig:config KERNEL_START
>       2 arch/powerpc/Kconfig:config PAGE_OFFSET
>       2 arch/powerpc/Kconfig:config PHYSICAL_START
>       2 arch/powerpc/Kconfig:config RELOCATABLE

The duplication of these is caused by the second definitions being
inside an "if PPC64" block.

For PAGE_OFFSET and KERNEL_START we want the 2nd definitions, the
default value is different between 32 & 64 bit.

For RELOCATABLE it's a little weird, but we want the 2nd definition
because the depends condition is different between 32 & 64bit - though
they could probably be merged.

I don't really see why we need two versions of PHYSICAL_START.

cheers

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [Uclinux-dist-devel] Config Items appearing twice in same Kconfig file?

[Mike Frysinger]

On Wed, May 12, 2010 at 10:40, Christoph Egger wrote:
> =C2=A0 =C2=A0 =C2=A02 arch/blackfin/Kconfig:config MEM_MT48LC32M8A2_75

thanks, i'll get this simple bug fixed
-mike