[RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Ezequiel Garcia]

Make this option a hidden one and get a cleaner configuration.
This option just selects a common infrastructure for MTD-based devices
to expose a block interface. There is no point in allowing a separate
enable/disable.

Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
---
 drivers/mtd/Kconfig | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
index 5fab4e6e..2a0ca2a 100644
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -158,9 +158,7 @@ config MTD_BCM47XX_PARTS
 comment "User Modules And Translation Layers"
 
 config MTD_BLKDEVS
-	tristate "Common interface to block layer for MTD 'translation layers'"
-	depends on BLOCK
-	default n
+	bool
 
 config MTD_BLOCK
 	tristate "Caching block device access to MTD devices"
-- 
1.8.1.5

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Ezequiel Garcia]

On Fri, Dec 13, 2013 at 10:58:44AM -0300, Ezequiel Garcia wrote:
> Make this option a hidden one and get a cleaner configuration.
> This option just selects a common infrastructure for MTD-based devices
> to expose a block interface. There is no point in allowing a separate
> enable/disable.
> 
> Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> ---
>  drivers/mtd/Kconfig | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> index 5fab4e6e..2a0ca2a 100644
> --- a/drivers/mtd/Kconfig
> +++ b/drivers/mtd/Kconfig
> @@ -158,9 +158,7 @@ config MTD_BCM47XX_PARTS
>  comment "User Modules And Translation Layers"
>  
>  config MTD_BLKDEVS
> -	tristate "Common interface to block layer for MTD 'translation layers'"
> -	depends on BLOCK
> -	default n
> +	bool
>  
>  config MTD_BLOCK
>  	tristate "Caching block device access to MTD devices"
> -- 
> 1.8.1.5
> 

Any comments on this one?
-- 
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Brian Norris]

On Thu, Dec 19, 2013 at 09:47:13PM -0300, Ezequiel Garcia wrote:
> On Fri, Dec 13, 2013 at 10:58:44AM -0300, Ezequiel Garcia wrote:
> > Make this option a hidden one and get a cleaner configuration.
> > This option just selects a common infrastructure for MTD-based devices
> > to expose a block interface. There is no point in allowing a separate
> > enable/disable.
> > 
> > Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> > ---
> >  drivers/mtd/Kconfig | 4 +---
> >  1 file changed, 1 insertion(+), 3 deletions(-)
> > 
> > diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> > index 5fab4e6e..2a0ca2a 100644
> > --- a/drivers/mtd/Kconfig
> > +++ b/drivers/mtd/Kconfig
> > @@ -158,9 +158,7 @@ config MTD_BCM47XX_PARTS
> >  comment "User Modules And Translation Layers"
> >  
> >  config MTD_BLKDEVS
> > -	tristate "Common interface to block layer for MTD 'translation layers'"
> > -	depends on BLOCK
> > -	default n
> > +	bool
> >  
> >  config MTD_BLOCK
> >  	tristate "Caching block device access to MTD devices"
> 
> Any comments on this one?

It looks OK to me. I might add a comment above MTD_BLKDEVS like I see in
some other similar Kconfig not-user-selectable options:

#
# MTD block device support is select'ed if needed
#

Brian

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Ezequiel Garcia]

On Thu, Dec 19, 2013 at 06:05:18PM -0800, Brian Norris wrote:
> On Thu, Dec 19, 2013 at 09:47:13PM -0300, Ezequiel Garcia wrote:
> > On Fri, Dec 13, 2013 at 10:58:44AM -0300, Ezequiel Garcia wrote:
> > > Make this option a hidden one and get a cleaner configuration.
> > > This option just selects a common infrastructure for MTD-based devices
> > > to expose a block interface. There is no point in allowing a separate
> > > enable/disable.
> > > 
> > > Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> > > ---
> > >  drivers/mtd/Kconfig | 4 +---
> > >  1 file changed, 1 insertion(+), 3 deletions(-)
> > > 
> > > diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> > > index 5fab4e6e..2a0ca2a 100644
> > > --- a/drivers/mtd/Kconfig
> > > +++ b/drivers/mtd/Kconfig
> > > @@ -158,9 +158,7 @@ config MTD_BCM47XX_PARTS
> > >  comment "User Modules And Translation Layers"
> > >  
> > >  config MTD_BLKDEVS
> > > -	tristate "Common interface to block layer for MTD 'translation layers'"
> > > -	depends on BLOCK
> > > -	default n
> > > +	bool
> > >  
> > >  config MTD_BLOCK
> > >  	tristate "Caching block device access to MTD devices"
> > 
> > Any comments on this one?
> 
> It looks OK to me. I might add a comment above MTD_BLKDEVS like I see in
> some other similar Kconfig not-user-selectable options:
> 
> #
> # MTD block device support is select'ed if needed
> #
> 

Fully agreed.
-- 
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Brian Norris]

On Fri, Dec 20, 2013 at 03:17:58PM -0300, Ezequiel Garcia wrote:
> On Thu, Dec 19, 2013 at 06:05:18PM -0800, Brian Norris wrote:
> > On Thu, Dec 19, 2013 at 09:47:13PM -0300, Ezequiel Garcia wrote:
> > > Any comments on this one?
> > 
> > It looks OK to me. I might add a comment above MTD_BLKDEVS like I see in
> > some other similar Kconfig not-user-selectable options:
> > 
> > #
> > # MTD block device support is select'ed if needed
> > #
> > 
> 
> Fully agreed.

Since I didn't receive a follow-up, I just added the comment and pushed
to l2-mtd.git. Diff pasted below.

Thanks,
Brian

diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
index 5fab4e6e8301..cdaa4e512c97 100644
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -157,10 +157,11 @@ config MTD_BCM47XX_PARTS
 
 comment "User Modules And Translation Layers"
 
+#
+# MTD block device support is select'ed if needed
+#
 config MTD_BLKDEVS
-	tristate "Common interface to block layer for MTD 'translation layers'"
-	depends on BLOCK
-	default n
+	bool
 
 config MTD_BLOCK
 	tristate "Caching block device access to MTD devices"

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Brian Norris]

Hi Ezequiel,

On Fri, Dec 13, 2013 at 10:58:44AM -0300, Ezequiel Garcia wrote:
> Make this option a hidden one and get a cleaner configuration.
> This option just selects a common infrastructure for MTD-based devices
> to expose a block interface. There is no point in allowing a separate
> enable/disable.
> 
> Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
> ---
>  drivers/mtd/Kconfig | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> index 5fab4e6e..2a0ca2a 100644
> --- a/drivers/mtd/Kconfig
> +++ b/drivers/mtd/Kconfig
> @@ -158,9 +158,7 @@ config MTD_BCM47XX_PARTS
>  comment "User Modules And Translation Layers"
>  
>  config MTD_BLKDEVS
> -	tristate "Common interface to block layer for MTD 'translation layers'"
> -	depends on BLOCK
> -	default n
> +	bool
>  
>  config MTD_BLOCK
>  	tristate "Caching block device access to MTD devices"

Hmm, I'm seeing some issues with this patch. I think it's problematic
that you are changing MTD_BLKDEVS from tristate to bool. We can't have
MTD_BLKDEVS built into the kernel while MTD itself is built as a module.
This manifests in linker errors in my all-module builds, with .config
like this (note that MTD_BLKDEVS gets "select"ed by MTD_SWAP, which
forces it to be built-in, not built as a module):

CONFIG_MTD=m
CONFIG_MTD_BLKDEVS=y
CONFIG_MTD_SWAP=m

[...]
  Building modules, stage 2.
    MODPOST 14 modules
    ERROR: "deregister_mtd_blktrans" [drivers/mtd/mtdswap.ko] undefined!
    ERROR: "register_mtd_blktrans" [drivers/mtd/mtdswap.ko] undefined!
    ERROR: "add_mtd_blktrans_dev" [drivers/mtd/mtdswap.ko] undefined!
    ERROR: "mtd_blktrans_cease_background" [drivers/mtd/mtdswap.ko] undefined!
    ERROR: "del_mtd_blktrans_dev" [drivers/mtd/mtdswap.ko] undefined!
    make[1]: *** [__modpost] Error 1
    make: *** [modules] Error 2

I believe the correct solution would be the following additional diff:

diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
index cdaa4e512c97..5ebcda39f554 100644
--- a/drivers/mtd/Kconfig
+++ b/drivers/mtd/Kconfig
@@ -161,7 +161,7 @@ comment "User Modules And Translation Layers"
 # MTD block device support is select'ed if needed
 #
 config MTD_BLKDEVS
-	bool
+	tristate
 
 config MTD_BLOCK
 	tristate "Caching block device access to MTD devices"
--

I'll squash this into your patch if I don't hear any objections.

Brian

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Ezequiel Garcia]

On Sun, Jan 05, 2014 at 12:46:10AM -0800, Brian Norris wrote:
> 
> Hmm, I'm seeing some issues with this patch. I think it's problematic
> that you are changing MTD_BLKDEVS from tristate to bool. We can't have
[..]
> 
> I believe the correct solution would be the following additional diff:
> 
> diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> index cdaa4e512c97..5ebcda39f554 100644
> --- a/drivers/mtd/Kconfig
> +++ b/drivers/mtd/Kconfig
> @@ -161,7 +161,7 @@ comment "User Modules And Translation Layers"
>  # MTD block device support is select'ed if needed
>  #
>  config MTD_BLKDEVS
> -	bool
> +	tristate
>  
>  config MTD_BLOCK
>  	tristate "Caching block device access to MTD devices"
> --
> 
> I'll squash this into your patch if I don't hear any objections.
> 

The solution looks fine. Thanks for taking care of it.
-- 
Ezequiel García, Free Electrons
Embedded Linux, Kernel and Android Engineering
http://free-electrons.com

Re: [RFC/PATCH] mtd: Hide CONFIG_MTD_BLKDEVS from the menu

[Brian Norris]

On Sun, Jan 05, 2014 at 02:37:57PM -0300, Ezequiel Garcia wrote:
> On Sun, Jan 05, 2014 at 12:46:10AM -0800, Brian Norris wrote:
> > 
> > Hmm, I'm seeing some issues with this patch. I think it's problematic
> > that you are changing MTD_BLKDEVS from tristate to bool. We can't have
> [..]
> > 
> > I believe the correct solution would be the following additional diff:
> > 
> > diff --git a/drivers/mtd/Kconfig b/drivers/mtd/Kconfig
> > index cdaa4e512c97..5ebcda39f554 100644
> > --- a/drivers/mtd/Kconfig
> > +++ b/drivers/mtd/Kconfig
> > @@ -161,7 +161,7 @@ comment "User Modules And Translation Layers"
> >  # MTD block device support is select'ed if needed
> >  #
> >  config MTD_BLKDEVS
> > -	bool
> > +	tristate
> >  
> >  config MTD_BLOCK
> >  	tristate "Caching block device access to MTD devices"
> > --
> > 
> > I'll squash this into your patch if I don't hear any objections.
> > 
> 
> The solution looks fine. Thanks for taking care of it.

OK, squahsed and pushed to l2-mtd.git.

Brian