[PATCH] romfs: only select BUFFER_HEAD for the block based path

[PATCH] romfs: only select BUFFER_HEAD for the block based path

[Christoph Hellwig]

selecting BUFFER_HEAD unconditionally does not work as romfs can also
be built with only the MTD backend and thus without CONFIG_BLOCK.

Fixes: 0f842210d97a ("fs/Kconfig: Fix compile error for romfs")
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 fs/romfs/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/romfs/Kconfig b/fs/romfs/Kconfig
index 3f6b550eee6192..f24a96a331af1b 100644
--- a/fs/romfs/Kconfig
+++ b/fs/romfs/Kconfig
@@ -2,7 +2,6 @@
 config ROMFS_FS
 	tristate "ROM file system support"
 	depends on BLOCK || MTD
-	select BUFFER_HEAD
 	help
 	  This is a very small read-only file system mainly intended for
 	  initial ram disks of installation disks, but it could be used for
@@ -58,6 +57,7 @@ endchoice
 config ROMFS_ON_BLOCK
 	bool
 	default y if ROMFS_BACKED_BY_BLOCK || ROMFS_BACKED_BY_BOTH
+	select BUFFER_HEAD
 
 config ROMFS_ON_MTD
 	bool
-- 
2.39.2

Re: [PATCH] romfs: only select BUFFER_HEAD for the block based path

[Christoph Hellwig]

Sorry, this should be:

From: Christoph Hellwig <hch@lst.de>

to match the signoff.  I managed to mess my mail setup, but it should
be fixed now.


On Fri, Aug 04, 2023 at 12:26:48PM +0200, Christoph Hellwig wrote:
> selecting BUFFER_HEAD unconditionally does not work as romfs can also
> be built with only the MTD backend and thus without CONFIG_BLOCK.
> 
> Fixes: 0f842210d97a ("fs/Kconfig: Fix compile error for romfs")
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  fs/romfs/Kconfig | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/romfs/Kconfig b/fs/romfs/Kconfig
> index 3f6b550eee6192..f24a96a331af1b 100644
> --- a/fs/romfs/Kconfig
> +++ b/fs/romfs/Kconfig
> @@ -2,7 +2,6 @@
>  config ROMFS_FS
>  	tristate "ROM file system support"
>  	depends on BLOCK || MTD
> -	select BUFFER_HEAD
>  	help
>  	  This is a very small read-only file system mainly intended for
>  	  initial ram disks of installation disks, but it could be used for
> @@ -58,6 +57,7 @@ endchoice
>  config ROMFS_ON_BLOCK
>  	bool
>  	default y if ROMFS_BACKED_BY_BLOCK || ROMFS_BACKED_BY_BOTH
> +	select BUFFER_HEAD
>  
>  config ROMFS_ON_MTD
>  	bool
> -- 
> 2.39.2
---end quoted text---

Re: [PATCH] romfs: only select BUFFER_HEAD for the block based path

[Li Zetao]

Hi Christoph,

On 2023/8/4 18:26, Christoph Hellwig wrote:
> selecting BUFFER_HEAD unconditionally does not work as romfs can also
> be built with only the MTD backend and thus without CONFIG_BLOCK.
> 
> Fixes: 0f842210d97a ("fs/Kconfig: Fix compile error for romfs")
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>   fs/romfs/Kconfig | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/romfs/Kconfig b/fs/romfs/Kconfig
> index 3f6b550eee6192..f24a96a331af1b 100644
> --- a/fs/romfs/Kconfig
> +++ b/fs/romfs/Kconfig
> @@ -2,7 +2,6 @@
>   config ROMFS_FS
>   	tristate "ROM file system support"
>   	depends on BLOCK || MTD
> -	select BUFFER_HEAD
>   	help
>   	  This is a very small read-only file system mainly intended for
>   	  initial ram disks of installation disks, but it could be used for
> @@ -58,6 +57,7 @@ endchoice
>   config ROMFS_ON_BLOCK
>   	bool
>   	default y if ROMFS_BACKED_BY_BLOCK || ROMFS_BACKED_BY_BOTH
> +	select BUFFER_HEAD
>   
>   config ROMFS_ON_MTD
>   	bool

I have tested the patch on my environment and it resolves the two 
compile errors reported by kernel test robot. The links are as follows:

Link1: 
https://lore.kernel.org/oe-kbuild-all/202308031810.pQzGmR1v-lkp@intel.com/
Link2: 
https://lore.kernel.org/oe-kbuild-all/202308041329.6SK97Asr-lkp@intel.com/

Tested-by: Li Zetao <lizetao1@huawei.com>

Thanks,
Li Zetao

Re: [PATCH] romfs: only select BUFFER_HEAD for the block based path

[Jens Axboe]

On 8/4/23 4:26?AM, Christoph Hellwig wrote:
> selecting BUFFER_HEAD unconditionally does not work as romfs can also
> be built with only the MTD backend and thus without CONFIG_BLOCK.

I folded this in with the other one and made a note about that.

-- 
Jens Axboe