Re: Ironic visualization in menuconfig for NTFS_FS vs NTFS3_FS

Hi Xi,

On 6/16/26 10:24, Xi Ruoyao wrote:
> Hi,
> 
> In the Kconfig for NTFS3 we now have:
> 
> config NTFS3_FS
>     tristate "NTFS Read-Write file system support"
>     depends on !NTFS_FS || m
> 
> It basically means some sort of conflict between NTFS_FS and NTFS3_FS. 
> But then menuconfig is apparently puzzled to believe NTFS3_FS "depends
> on" NTFS_FS, and put NTFS3_FS as a subentry of NTFS_FS:
> 
>   │ │    <M> NTFS file system support                                     │ │  
>   │ │    [ ]   NTFS debugging support (NEW)                               │ │  
>   │ │    [ ]   NTFS POSIX Access Control Lists (NEW)                      │ │  
>   │ │    <M>   NTFS Read-Write file system support                        │ │  
>   │ │    [ ]     64 bits per NTFS clusters                                │ │  
>   │ │    [*]     activate support of external compressions lzx/xpress     │ │  
>   │ │    [ ]     NTFS POSIX Access Control Lists                          │  

This looks like a bug in menuconfig's behavior, as from
Documentation/kbuild/kconfig-language.rst:
"If a menu entry somehow depends on the previous entry, it can be made a
submenu of it."

And since Kconfig treats 'source' similarly to C's '#include' (the text
of the file effectively just gets copy-pasted in), it's like the config
options in fs/ntfs3/Kconfig immediately follow the options in
fs/ntfs/Kconfig since the sources are back-to-back.

You can test this by simply adding some other, unrelated config option
between the two 'source' entries that doesn't depend on NTFS_FS or
!NTFS_FS.

So, anyway, it seems to be simply ignoring the '!', and just treating it
as a dependency.
And you can test this by removing that dependency on !NTFS_FS, which
causes menuconfig to stop rendering it as part of the menu.

I think your workaround is okay (there are other similar 'NO_' config
options in the kernel), but it would be better to fix this menuconfig
bug.

- Julian Braha