2011-02-06 15:19:40

by Marco Stornelli

Subject: [PATCH] Kconfig: XIP doesn't depend on block

From: Marco Stornelli <[email protected]>

XIP doesn't depend on block symbol, then we can reorder the Kconfig.
For ext2 doesn't change the Kconfig behavior but if other fs will use
FS_XIP it won't need to include block support if not needed.

Signed-off-by: Marco Stornelli <[email protected]>
---

--- Kconfig.orig 2011-01-19 00:14:02.000000000 +0100
+++ Kconfig 2011-02-06 16:04:51.000000000 +0100
@@ -9,13 +9,6 @@ if BLOCK
source "fs/ext2/Kconfig"
source "fs/ext3/Kconfig"
source "fs/ext4/Kconfig"
-
-config FS_XIP
-# execute in place
- bool
- depends on EXT2_FS_XIP
- default y
-
source "fs/jbd/Kconfig"
source "fs/jbd2/Kconfig"

@@ -38,6 +31,12 @@ source "fs/nilfs2/Kconfig"

endif # BLOCK

+config FS_XIP
+# execute in place
+ bool
+ depends on EXT2_FS_XIP
+ default y
+
# Posix ACL utility routines
#
# Note: Posix ACLs can be implemented without these helpers. Never use

2011-02-06 16:44:45

by Randy Dunlap

Subject: Re: [PATCH] Kconfig: XIP doesn't depend on block

On Sun, 06 Feb 2011 16:15:00 +0100 Marco Stornelli wrote:

> From: Marco Stornelli <[email protected]>
>
> XIP doesn't depend on block symbol, then we can reorder the Kconfig.
> For ext2 doesn't change the Kconfig behavior but if other fs will use
> FS_XIP it won't need to include block support if not needed.

Hi Marco,

Do you know of a filesystem where this matters?

> Signed-off-by: Marco Stornelli <[email protected]>
> ---
>
> --- Kconfig.orig 2011-01-19 00:14:02.000000000 +0100
> +++ Kconfig 2011-02-06 16:04:51.000000000 +0100

This filename should include path, like
--- fs/Kconfig.orig
+++ fs/Kconfig

> @@ -9,13 +9,6 @@ if BLOCK
> source "fs/ext2/Kconfig"
> source "fs/ext3/Kconfig"
> source "fs/ext4/Kconfig"
> -

The 3 filesystems above are immediately under:

if BLOCK

so ext[234] depend on BLOCK. Why would it matter about FS_XIP?

I don't object to the patch if FS_XIP builds/works without BLOCK being
enabled.


> -config FS_XIP
> -# execute in place
> - bool
> - depends on EXT2_FS_XIP
> - default y
> -
> source "fs/jbd/Kconfig"
> source "fs/jbd2/Kconfig"
>
> @@ -38,6 +31,12 @@ source "fs/nilfs2/Kconfig"
>
> endif # BLOCK
>
> +config FS_XIP
> +# execute in place
> + bool
> + depends on EXT2_FS_XIP
> + default y
> +
> # Posix ACL utility routines
> #
> # Note: Posix ACLs can be implemented without these helpers. Never use
>


---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

2011-02-07 07:52:44

by Marco Stornelli

Subject: Re: [PATCH] Kconfig: XIP doesn't depend on block

2011/2/6 Randy Dunlap <[email protected]>:
> On Sun, 06 Feb 2011 16:15:00 +0100 Marco Stornelli wrote:
>
>> From: Marco Stornelli <[email protected]>
>>
>> XIP doesn't depend on block symbol, then we can reorder the Kconfig.
>> For ext2 doesn't change the Kconfig behavior but if other fs will use
>> FS_XIP it won't need to include block support if not needed.
>
> Hi Marco,
>
> Do you know of a filesystem where this matters?
>

Yes, for example mine (pramfs) :)

>> Signed-off-by: Marco Stornelli <[email protected]>
>> ---
>>
>> --- Kconfig.orig ? ? ?2011-01-19 00:14:02.000000000 +0100
>> +++ Kconfig ? 2011-02-06 16:04:51.000000000 +0100
>
> This filename should include path, like
> --- fs/Kconfig.orig
> +++ fs/Kconfig
>
>> @@ -9,13 +9,6 @@ if BLOCK
>> ?source "fs/ext2/Kconfig"
>> ?source "fs/ext3/Kconfig"
>> ?source "fs/ext4/Kconfig"
>> -
>
> The 3 filesystems above are immediately under:
>
> if BLOCK
>
> so ext[234] depend on BLOCK. ?Why would it matter about FS_XIP?
>

I don't know but for example even POSIX ACL was under "if BLOCK",
maybe there is only an historical refuse.

Marco