two more oddities with the fs/Kconfig file

two more oddities with the fs/Kconfig file

[Robert P. J. Day]

  there are a few options that are categorized as simply
"bool", with no following label -- examples being UMSDOS,
QUOTACTL, and a couple of others.  without a label on that
line, the option is not displayed for selection anywhere
on the menu.  is this deliberate?

  also, right near the bottom of the file, there are three
options (ZISOFS_FS, FS_MBCACHE and FS_POSIX_ACL), that are
defined after the first submenu, and they don't appear to
show up anywhere on the screen for selection either.  where
are these supposed to be displayed?

rday

Re: two more oddities with the fs/Kconfig file

[Adrian Bunk]

On Sun, Jan 12, 2003 at 02:07:13AM -0500, Robert P. J. Day wrote:
> 
>   there are a few options that are categorized as simply
> "bool", with no following label -- examples being UMSDOS,
> QUOTACTL, and a couple of others.  without a label on that
> line, the option is not displayed for selection anywhere
> on the menu.  is this deliberate?
>...

Yes, this is what was called define_bool in the old kconfig.

E.g.

config QUOTACTL
        bool
        depends on XFS_QUOTA || QUOTA
        default y

says that QUOTACTL is automatically selected if XFS_QUOTA or QUOTA is 
selected. This is a config option that is never visible to the user 
configuring the kernel.

> rday

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: two more oddities with the fs/Kconfig file

[Robert P. J. Day]

On Sun, 12 Jan 2003, Adrian Bunk wrote:

> On Sun, Jan 12, 2003 at 02:07:13AM -0500, Robert P. J. Day wrote:
> > 
> >   there are a few options that are categorized as simply
> > "bool", with no following label -- examples being UMSDOS,
> > QUOTACTL, and a couple of others.  without a label on that
> > line, the option is not displayed for selection anywhere
> > on the menu.  is this deliberate?
> >...
> 
> Yes, this is what was called define_bool in the old kconfig.
> 
> E.g.
> 
> config QUOTACTL
>         bool
>         depends on XFS_QUOTA || QUOTA
>         default y
> 
> says that QUOTACTL is automatically selected if XFS_QUOTA or QUOTA is 
> selected. This is a config option that is never visible to the user 
> configuring the kernel.

ah, and the same would apply to those options categorized as
"tristate", with no label then?  thanks.

rday

Re: two more oddities with the fs/Kconfig file

[Robert P. J. Day]

On Sun, 12 Jan 2003, Adrian Bunk wrote:

> On Sun, Jan 12, 2003 at 02:07:13AM -0500, Robert P. J. Day wrote:
> > 
> >   there are a few options that are categorized as simply
> > "bool", with no following label -- examples being UMSDOS,
> > QUOTACTL, and a couple of others.  without a label on that
> > line, the option is not displayed for selection anywhere
> > on the menu.  is this deliberate?
> >...
> 
> Yes, this is what was called define_bool in the old kconfig.
> 
> E.g.
> 
> config QUOTACTL
>         bool
>         depends on XFS_QUOTA || QUOTA
>         default y
> 
> says that QUOTACTL is automatically selected if XFS_QUOTA or QUOTA is 
> selected. This is a config option that is never visible to the user 
> configuring the kernel.

just a picky point -- since "JBD" is one of these bool options
with no label, why does it need a help screen as it has in the
current Kconfig file?  particularly with info about how and
why to *select* this module.  

or is that cruft leftover from before?

rday

Re: two more oddities with the fs/Kconfig file

[Robert P. J. Day]

On Sun, 12 Jan 2003, Adrian Bunk wrote:

> On Sun, Jan 12, 2003 at 02:07:13AM -0500, Robert P. J. Day wrote:
> > 
> >   there are a few options that are categorized as simply
> > "bool", with no following label -- examples being UMSDOS,
> > QUOTACTL, and a couple of others.  without a label on that
> > line, the option is not displayed for selection anywhere
> > on the menu.  is this deliberate?
> >...
> 
> Yes, this is what was called define_bool in the old kconfig.
> 
> E.g.
> 
> config QUOTACTL
>         bool
>         depends on XFS_QUOTA || QUOTA
>         default y
> 
> says that QUOTACTL is automatically selected if XFS_QUOTA or QUOTA is 
> selected. This is a config option that is never visible to the user 
> configuring the kernel.

hold everything.  i *thought* i understood this, now i'm not so sure.
i zipped thru fs/Kconfig, and extracted all of the options i could
find that were either "bool" or "tristate" with no labels.  supposedly,
these are never visible to the user.  examples:

config QUOTACTL
	bool
	depends on XFS_QUOTA || QUOTA
	default y

config JBD
	bool
	default EXT3_FS
	---help---

config UMSDOS_FS
	bool
	---help---

config RAMFS
	bool
	default y
	---help---

config SUNRPC
	tristate
	default m if NFS_FS!=y && NFSD!=y && (NFS_FS=m || NFSD=m)
	default y if NFS_FS=y || NFSD=y

config LOCKD
	tristate
	default m if NFS_FS!=y && NFSD!=y && (NFS_FS=m || NFSD=m)
	default y if NFS_FS=y || NFSD=y

config LOCKD_V4
	bool
	depends on NFSD_V3 || NFS_V3
	default y

config EXPORTFS
	tristate
	default NFSD

config RXRPC
	tristate
	default m if AFS_FS=m
	default y if AFS_FS=y

config ZISOFS_FS
	tristate
	depends on ZISOFS
	default ISO9660_FS

config FS_MBCACHE
	tristate
	depends on EXT2_FS_XATTR || EXT3_FS_XATTR
	default y if EXT2_FS=y || EXT3_FS=y
	default m if EXT2_FS=m || EXT3_FS=m

config FS_POSIX_ACL
	bool
	depends on EXT2_FS_POSIX_ACL || EXT3_FS_POSIX_ACL || JFS_POSIX_ACL
	default y


  so, according to the above, i would *always* get RAMFS, no matter
what?  there's no way to turn this off?

  but what about UMSDOS_FS?  it's a non-displayed boolean, but
it has no dependencies, and more weirdly, no default value.
so what will it be set to?  and why *shouldn't* i be able to
select UMSDOS support or not?

  and if they're not displayed, why do some of them have help
screens?  and why can't i get to sleep?  argh.

rday

Re: two more oddities with the fs/Kconfig file

[Robert P. J. Day]

On Sun, 12 Jan 2003, Robert P. J. Day wrote:

> config UMSDOS_FS
> 	bool
> 	---help---

oh, never mind ... i was confused by what the above meant,
not realizing that i had inadvertantly removed the comment
above it that stated that it was broken.  argh.

rday

Re: two more oddities with the fs/Kconfig file

[Russell King]

On Sun, Jan 12, 2003 at 04:54:54AM -0500, Robert P. J. Day wrote:
>   so, according to the above, i would *always* get RAMFS, no matter
> what?  there's no way to turn this off?

iirc, RAMFS is used for initramfs (which currently provides the basis for
mounting the root filesystem) and therefore must always be enabled.

-- 
Russell King (rmk@arm.linux.org.uk)                The developer of ARM Linux
             http://www.arm.linux.org.uk/personal/aboutme.html