| Message ID | 20230918125744.4342-1-peter@n8pjl.ca |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [RESEND] um: Remove CONFIG_REISERFS_FS from defconfig | expand |
On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > ReiserFS has been deprecated for a year and a half, yet is still built > as part of a defconfig UML kernel. > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > filesystem is slated to be removed in 2025. Remove it from the defconfig > profile now, as part of its deprecation process. > > Signed-off-by: Peter Lafreniere <peter@n8pjl.ca> Makes sense but I've noticed there are many more architectures that define CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, alpha). So perhaps we should do this for all of them when we are at it? Honza > --- > Fixed invalid DKIM settings causing the previous patch to be rejected. > > arch/um/configs/i386_defconfig | 1 - > arch/um/configs/x86_64_defconfig | 1 - > 2 files changed, 2 deletions(-) > > diff --git a/arch/um/configs/i386_defconfig b/arch/um/configs/i386_defconfig > index e543cbac8792..9c9c77f1255a 100644 > --- a/arch/um/configs/i386_defconfig > +++ b/arch/um/configs/i386_defconfig > @@ -61,7 +61,6 @@ CONFIG_UML_NET_DAEMON=y > CONFIG_UML_NET_MCAST=y > CONFIG_UML_NET_SLIRP=y > CONFIG_EXT4_FS=y > -CONFIG_REISERFS_FS=y > CONFIG_QUOTA=y > CONFIG_AUTOFS_FS=m > CONFIG_ISO9660_FS=m > diff --git a/arch/um/configs/x86_64_defconfig b/arch/um/configs/x86_64_defconfig > index 939cb12318ca..03b10d3f6816 100644 > --- a/arch/um/configs/x86_64_defconfig > +++ b/arch/um/configs/x86_64_defconfig > @@ -59,7 +59,6 @@ CONFIG_UML_NET_DAEMON=y > CONFIG_UML_NET_MCAST=y > CONFIG_UML_NET_SLIRP=y > CONFIG_EXT4_FS=y > -CONFIG_REISERFS_FS=y > CONFIG_QUOTA=y > CONFIG_AUTOFS_FS=m > CONFIG_ISO9660_FS=m > -- > 2.42.0 > >
----- Ursprüngliche Mail ----- > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: >> ReiserFS has been deprecated for a year and a half, yet is still built >> as part of a defconfig UML kernel. >> >> According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the >> filesystem is slated to be removed in 2025. Remove it from the defconfig >> profile now, as part of its deprecation process. >> >> Signed-off-by: Peter Lafreniere <peter@n8pjl.ca> > > Makes sense but I've noticed there are many more architectures that define > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > alpha). So perhaps we should do this for all of them when we are at it? Yes, a combined patch makes in this case more sense. >> >> arch/um/configs/i386_defconfig | 1 - >> arch/um/configs/x86_64_defconfig | 1 - >> 2 files changed, 2 deletions(-) For UML: Acked-by: Richard Weinberger <richard@nod.at> Thanks, //richard
On Monday, September 18th, 2023 at 09:26, Jan Kara <jack@suse.cz> wrote: > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > > ReiserFS has been deprecated for a year and a half, yet is still built > > as part of a defconfig UML kernel. > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > profile now, as part of its deprecation process. > > > > Signed-off-by: Peter Lafreniere peter@n8pjl.ca > > > Makes sense but I've noticed there are many more architectures that define > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > alpha). So perhaps we should do this for all of them when we are at it? I agree. I can send a patch series for those arches except for ia64, which has already been removed from linux-next. > > Honza > > > --- > > Fixed invalid DKIM settings causing the previous patch to be rejected. > > > > arch/um/configs/i386_defconfig | 1 - > > arch/um/configs/x86_64_defconfig | 1 - > > 2 files changed, 2 deletions(-) > > > > diff --git a/arch/um/configs/i386_defconfig b/arch/um/configs/i386_defconfig > > index e543cbac8792..9c9c77f1255a 100644 > > --- a/arch/um/configs/i386_defconfig > > +++ b/arch/um/configs/i386_defconfig > > @@ -61,7 +61,6 @@ CONFIG_UML_NET_DAEMON=y > > CONFIG_UML_NET_MCAST=y > > CONFIG_UML_NET_SLIRP=y > > CONFIG_EXT4_FS=y > > -CONFIG_REISERFS_FS=y > > CONFIG_QUOTA=y > > CONFIG_AUTOFS_FS=m > > CONFIG_ISO9660_FS=m > > diff --git a/arch/um/configs/x86_64_defconfig b/arch/um/configs/x86_64_defconfig > > index 939cb12318ca..03b10d3f6816 100644 > > --- a/arch/um/configs/x86_64_defconfig > > +++ b/arch/um/configs/x86_64_defconfig > > @@ -59,7 +59,6 @@ CONFIG_UML_NET_DAEMON=y > > CONFIG_UML_NET_MCAST=y > > CONFIG_UML_NET_SLIRP=y > > CONFIG_EXT4_FS=y > > -CONFIG_REISERFS_FS=y > > CONFIG_QUOTA=y > > CONFIG_AUTOFS_FS=m > > CONFIG_ISO9660_FS=m > > -- > > 2.42.0 > > -- > Jan Kara jack@suse.com > > SUSE Labs, CRi
On Mon 18-09-23 14:14:21, Peter Lafreniere wrote: > On Monday, September 18th, 2023 at 09:26, Jan Kara <jack@suse.cz> wrote: > > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > > > > ReiserFS has been deprecated for a year and a half, yet is still built > > > as part of a defconfig UML kernel. > > > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > > profile now, as part of its deprecation process. > > > > > > Signed-off-by: Peter Lafreniere peter@n8pjl.ca > > > > > > Makes sense but I've noticed there are many more architectures that define > > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > > alpha). So perhaps we should do this for all of them when we are at it? > > I agree. I can send a patch series for those arches except for ia64, which has > already been removed from linux-next. Than would be great. Thanks! Honza
Hi Jan, On Mon, Sep 18, 2023 at 3:26 PM Jan Kara <jack@suse.cz> wrote: > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > ReiserFS has been deprecated for a year and a half, yet is still built > > as part of a defconfig UML kernel. > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > profile now, as part of its deprecation process. > > > > Signed-off-by: Peter Lafreniere <peter@n8pjl.ca> > > Makes sense but I've noticed there are many more architectures that define > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > alpha). So perhaps we should do this for all of them when we are at it? The policy for the m68k defconfigs is to enable all available exotic file systems as modules, so support for them is available when needed. Please let me know why this should be changed, as long as ReiserFS is still available in the upstream kernel. Gr{oetje,eeting}s, Geert
On Mon 18-09-23 16:58:05, Geert Uytterhoeven wrote: > Hi Jan, > > On Mon, Sep 18, 2023 at 3:26 PM Jan Kara <jack@suse.cz> wrote: > > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > > ReiserFS has been deprecated for a year and a half, yet is still built > > > as part of a defconfig UML kernel. > > > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > > profile now, as part of its deprecation process. > > > > > > Signed-off-by: Peter Lafreniere <peter@n8pjl.ca> > > > > Makes sense but I've noticed there are many more architectures that define > > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > > alpha). So perhaps we should do this for all of them when we are at it? > > The policy for the m68k defconfigs is to enable all available exotic > file systems as modules, so support for them is available when needed. > > Please let me know why this should be changed, as long as ReiserFS is > still available in the upstream kernel. There's no need to change if you don't want to. I can just imagine how for other archs which seem to enable only "useful" filesystems in defconfig disabling a deprecated filesystem makes sense. Honza
On Monday, September 18th, 2023 at 09:26, Jan Kara <jack@suse.cz> wrote: > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > > ReiserFS has been deprecated for a year and a half, yet is still built > > as part of a defconfig UML kernel. > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > profile now, as part of its deprecation process. > > > > Signed-off-by: Peter Lafreniere peter@n8pjl.ca > > > Makes sense but I've noticed there are many more architectures that define > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > alpha). So perhaps we should do this for all of them when we are at it? I agree. I can send a patch series for those arches except for ia64, which has already been removed from linux-next. > > Honza > > > --- > > Fixed invalid DKIM settings causing the previous patch to be rejected. > > > > arch/um/configs/i386_defconfig | 1 - > > arch/um/configs/x86_64_defconfig | 1 - > > 2 files changed, 2 deletions(-) > > > > diff --git a/arch/um/configs/i386_defconfig b/arch/um/configs/i386_defconfig > > index e543cbac8792..9c9c77f1255a 100644 > > --- a/arch/um/configs/i386_defconfig > > +++ b/arch/um/configs/i386_defconfig > > @@ -61,7 +61,6 @@ CONFIG_UML_NET_DAEMON=y > > CONFIG_UML_NET_MCAST=y > > CONFIG_UML_NET_SLIRP=y > > CONFIG_EXT4_FS=y > > -CONFIG_REISERFS_FS=y > > CONFIG_QUOTA=y > > CONFIG_AUTOFS_FS=m > > CONFIG_ISO9660_FS=m > > diff --git a/arch/um/configs/x86_64_defconfig b/arch/um/configs/x86_64_defconfig > > index 939cb12318ca..03b10d3f6816 100644 > > --- a/arch/um/configs/x86_64_defconfig > > +++ b/arch/um/configs/x86_64_defconfig > > @@ -59,7 +59,6 @@ CONFIG_UML_NET_DAEMON=y > > CONFIG_UML_NET_MCAST=y > > CONFIG_UML_NET_SLIRP=y > > CONFIG_EXT4_FS=y > > -CONFIG_REISERFS_FS=y > > CONFIG_QUOTA=y > > CONFIG_AUTOFS_FS=m > > CONFIG_ISO9660_FS=m > > -- > > 2.42.0 > > -- > Jan Kara jack@suse.com > > SUSE Labs, CRi
Hi Geert, On Monday, September 18th, 2023 at 10:58, Geert Uytterhoeven <geert@linux-m68k.org> wrote: > Hi Jan, > > On Mon, Sep 18, 2023 at 3:26 PM Jan Kara jack@suse.cz wrote: > > > On Mon 18-09-23 12:58:26, Peter Lafreniere wrote: > > > > > ReiserFS has been deprecated for a year and a half, yet is still built > > > as part of a defconfig UML kernel. > > > > > > According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the > > > filesystem is slated to be removed in 2025. Remove it from the defconfig > > > profile now, as part of its deprecation process. > > > > > > Signed-off-by: Peter Lafreniere peter@n8pjl.ca > > > > Makes sense but I've noticed there are many more architectures that define > > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > > alpha). So perhaps we should do this for all of them when we are at it? > > > The policy for the m68k defconfigs is to enable all available exotic > file systems as modules, so support for them is available when needed. > > Please let me know why this should be changed, as long as ReiserFS is > still available in the upstream kernel. I understand that policy and it makes particular sense for older/hobbyist hardware like the m68k classic where ReiserFS is included in make defconfig. However, ReiserFS is deprecated and should be phased out as best as possible before its removal, and pruning config options for a filesystem that is considered obsolete and should not be used on the newest kernels is a step towards that. ReiserFS is still availiable to be configured by anyone who builds their own kernel and is still availiable in all(mod|yes)config. As such I can justify the removal of some old config options that doesn't actually remove support for the filesystem when it simplifies the deprecation and removal processes for an ancient, largely unused filesystem. Please remember that users who build and install their own kernel on hobbyist hardware should be capable enough to select one Kconfig option if they happen to need ReiserFS. If not, then removing ReiserFS from the defconfig on a kernel update is an easier issue to work around than waiting to remove the filesystem from the tree entirely. > > Gr{oetje,eeting}s, > > Geert > > -- > Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org > > In personal conversations with technical people, I call myself a hacker. But > when I'm talking to journalists I just say "programmer" or something like that. > -- Linus Torvalds Cheers, Peter Lafreniere <peter@n8pjl.ca> And I sent the wrong email the first time. So sorry about that.
> > > Makes sense but I've noticed there are many more architectures that define > > > CONFIG_REISERFS_FS in their defconfig (sh, powerpc, mips, m68k, ia64, arm, > > > alpha). So perhaps we should do this for all of them when we are at it? > > > > The policy for the m68k defconfigs is to enable all available exotic > > file systems as modules, so support for them is available when needed. > > > > Please let me know why this should be changed, as long as ReiserFS is > > still available in the upstream kernel. > > > I understand that policy and it makes particular sense for older/hobbyist > hardware like the m68k classic where ReiserFS is included in make defconfig. > > However, ReiserFS is deprecated and should be phased out as best as possible > before its removal, and pruning config options for a filesystem that is > considered obsolete and should not be used on the newest kernels is a step > towards that. > > ReiserFS is still availiable to be configured by anyone who builds their own > kernel and is still availiable in all(mod|yes)config. As such I can justify > the removal of some old config options that doesn't actually remove support for > the filesystem when it simplifies the deprecation and removal processes for an > ancient, largely unused filesystem. > > Please remember that users who build and install their own kernel on hobbyist > hardware should be capable enough to select one Kconfig option if they happen to > need ReiserFS. If not, then removing ReiserFS from the defconfig on a kernel > update is an easier issue to work around than waiting to remove the filesystem > from the tree entirely. Though as Jan said, you're free to not accept a patch removing the config option until the filesystem is gone for good. I just feel like it's easier to remove ReiserFS from all defconfigs in one fell swoop. Cheers, Peter Lafreniere <peter@n8pjl.ca>
diff --git a/arch/um/configs/i386_defconfig b/arch/um/configs/i386_defconfig index e543cbac8792..9c9c77f1255a 100644 --- a/arch/um/configs/i386_defconfig +++ b/arch/um/configs/i386_defconfig @@ -61,7 +61,6 @@ CONFIG_UML_NET_DAEMON=y CONFIG_UML_NET_MCAST=y CONFIG_UML_NET_SLIRP=y CONFIG_EXT4_FS=y -CONFIG_REISERFS_FS=y CONFIG_QUOTA=y CONFIG_AUTOFS_FS=m CONFIG_ISO9660_FS=m diff --git a/arch/um/configs/x86_64_defconfig b/arch/um/configs/x86_64_defconfig index 939cb12318ca..03b10d3f6816 100644 --- a/arch/um/configs/x86_64_defconfig +++ b/arch/um/configs/x86_64_defconfig @@ -59,7 +59,6 @@ CONFIG_UML_NET_DAEMON=y CONFIG_UML_NET_MCAST=y CONFIG_UML_NET_SLIRP=y CONFIG_EXT4_FS=y -CONFIG_REISERFS_FS=y CONFIG_QUOTA=y CONFIG_AUTOFS_FS=m CONFIG_ISO9660_FS=m
ReiserFS has been deprecated for a year and a half, yet is still built as part of a defconfig UML kernel. According to commit eb103a51640e ("reiserfs: Deprecate reiserfs"), the filesystem is slated to be removed in 2025. Remove it from the defconfig profile now, as part of its deprecation process. Signed-off-by: Peter Lafreniere <peter@n8pjl.ca> --- Fixed invalid DKIM settings causing the previous patch to be rejected. arch/um/configs/i386_defconfig | 1 - arch/um/configs/x86_64_defconfig | 1 - 2 files changed, 2 deletions(-)