[PATCH] ext4: remove obsolete EXT3 config options

[PATCH] ext4: remove obsolete EXT3 config options

[Lukas Bulwahn]

From: Lukas Bulwahn <lukas.bulwahn@redhat.com>

In June 2015, commit c290ea01abb7 ("fs: Remove ext3 filesystem driver")
removed the historic ext3 filesystem support as ext3 partitions are fully
supported with the ext4 filesystem support. To simplify updating the kernel
build configuration, which had only EXT3 support but not EXT4 support
enabled, the three config options EXT3_{FS,FS_POSIX_ACL,FS_SECURITY} were
kept, instead of immediately removing them. The three options just enable
the corresponding EXT4 counterparts when configs from older kernel versions
are used to build on later kernel versions. This ensures that the kernels
from those kernel build configurations would then continue to have EXT4
enabled for supporting booting from ext3 and ext4 file systems, to avoid
potential unexpected surprises.

Given that the kernel build configuration has no backwards-compatibility
guarantee and this transition phase for such build configurations has been
in place for a decade, we can reasonably expect all such users to have
transitioned to use the EXT4 config options in their config files at this
point in time. With that in mind, the three EXT3 config options are
obsolete by now.

Remove the obsolete EXT3 config options.

Signed-off-by: Lukas Bulwahn <lukas.bulwahn@redhat.com>
---
 fs/ext4/Kconfig | 27 ---------------------------
 1 file changed, 27 deletions(-)

diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig
index c9ca41d91a6c..01873c2a34ad 100644
--- a/fs/ext4/Kconfig
+++ b/fs/ext4/Kconfig
@@ -1,31 +1,4 @@
 # SPDX-License-Identifier: GPL-2.0-only
-# Ext3 configs are here for backward compatibility with old configs which may
-# have EXT3_FS set but not EXT4_FS set and thus would result in non-bootable
-# kernels after the removal of ext3 driver.
-config EXT3_FS
-	tristate "The Extended 3 (ext3) filesystem"
-	select EXT4_FS
-	help
-	  This config option is here only for backward compatibility. ext3
-	  filesystem is now handled by the ext4 driver.
-
-config EXT3_FS_POSIX_ACL
-	bool "Ext3 POSIX Access Control Lists"
-	depends on EXT3_FS
-	select EXT4_FS_POSIX_ACL
-	select FS_POSIX_ACL
-	help
-	  This config option is here only for backward compatibility. ext3
-	  filesystem is now handled by the ext4 driver.
-
-config EXT3_FS_SECURITY
-	bool "Ext3 Security Labels"
-	depends on EXT3_FS
-	select EXT4_FS_SECURITY
-	help
-	  This config option is here only for backward compatibility. ext3
-	  filesystem is now handled by the ext4 driver.
-
 config EXT4_FS
 	tristate "The Extended 4 (ext4) filesystem"
 	select BUFFER_HEAD
-- 
2.50.1

Re: [PATCH] ext4: remove obsolete EXT3 config options

[Theodore Ts'o]

On Wed, Aug 27, 2025 at 11:08:08AM +0200, Lukas Bulwahn wrote:
> From: Lukas Bulwahn <lukas.bulwahn@redhat.com>
> 
> In June 2015, commit c290ea01abb7 ("fs: Remove ext3 filesystem driver")
> removed the historic ext3 filesystem support as ext3 partitions are fully
> supported with the ext4 filesystem support. To simplify updating the kernel
> build configuration, which had only EXT3 support but not EXT4 support
> enabled, the three config options EXT3_{FS,FS_POSIX_ACL,FS_SECURITY} were
> kept, instead of immediately removing them. The three options just enable
> the corresponding EXT4 counterparts when configs from older kernel versions
> are used to build on later kernel versions. This ensures that the kernels
> from those kernel build configurations would then continue to have EXT4
> enabled for supporting booting from ext3 and ext4 file systems, to avoid
> potential unexpected surprises.
> 
> Given that the kernel build configuration has no backwards-compatibility
> guarantee and this transition phase for such build configurations has been
> in place for a decade, we can reasonably expect all such users to have
> transitioned to use the EXT4 config options in their config files at this
> point in time. With that in mind, the three EXT3 config options are
> obsolete by now.

Do we need to worry about what happens if someone uses an oldconfig
from an RHEL 7/8/9 kernel (or equivalent SuSE kernel) to try to build
the latest upstream kernel?

I don't really care because I don't use enterprise distros, and I
haven't worked for a company that supports enterprise customers for
over a decade, but I thought I should just check.

Cheers,

					- Ted