Skip to contentSkip to conversation

sysv: Remove the filesystem

PATCH In review Jan Kara 8 participants 11 messages Jan 6, 2025 – Jan 9, 2025

Subsystems (7)
Documentation VFS PowerPC MIPS +3 more subsystems
lore
Share & tools
lore

Jan Kara

author reviewer
#0 lore raw

PATCH 1/1 sysv: Remove the filesystem

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

 Documentation/filesystems/index.rst         |   1 -
 Documentation/filesystems/sysv-fs.rst       | 264 ---------
 MAINTAINERS                                 |   6 -
 arch/loongarch/configs/loongson3_defconfig  |   1 -
 arch/m68k/configs/amiga_defconfig           |   1 -
 arch/m68k/configs/apollo_defconfig          |   1 -
 arch/m68k/configs/atari_defconfig           |   1 -
 arch/m68k/configs/bvme6000_defconfig        |   1 -
 arch/m68k/configs/hp300_defconfig           |   1 -
 arch/m68k/configs/mac_defconfig             |   1 -
 arch/m68k/configs/multi_defconfig           |   1 -
 arch/m68k/configs/mvme147_defconfig         |   1 -
 arch/m68k/configs/mvme16x_defconfig         |   1 -
 arch/m68k/configs/q40_defconfig             |   1 -
 arch/m68k/configs/sun3_defconfig            |   1 -
 arch/m68k/configs/sun3x_defconfig           |   1 -
 arch/mips/configs/malta_defconfig           |   1 -
 arch/mips/configs/malta_kvm_defconfig       |   1 -
 arch/mips/configs/maltaup_xpa_defconfig     |   1 -
 arch/mips/configs/rm200_defconfig           |   1 -
 arch/parisc/configs/generic-64bit_defconfig |   1 -
 arch/powerpc/configs/fsl-emb-nonhw.config   |   1 -
 arch/powerpc/configs/ppc6xx_defconfig       |   1 -
 fs/Kconfig                                  |   1 -
 fs/Makefile                                 |   1 -
 fs/sysv/Kconfig                             |  38 --
 fs/sysv/Makefile                            |   9 -
 fs/sysv/balloc.c                            | 240 --------
 fs/sysv/dir.c                               | 378 -------------
 fs/sysv/file.c                              |  59 --
 fs/sysv/ialloc.c                            | 235 --------
 fs/sysv/inode.c                             | 354 ------------
 fs/sysv/itree.c                             | 511 -----------------
 fs/sysv/namei.c                             | 280 ---------
 fs/sysv/super.c                             | 595 --------------------
 fs/sysv/sysv.h                              | 245 --------
 include/linux/sysv_fs.h                     | 214 -------
 37 files changed, 3451 deletions(-)
 delete mode 100644 Documentation/filesystems/sysv-fs.rst
 delete mode 100644 fs/sysv/Kconfig
 delete mode 100644 fs/sysv/Makefile
 delete mode 100644 fs/sysv/balloc.c
 delete mode 100644 fs/sysv/dir.c
 delete mode 100644 fs/sysv/file.c
 delete mode 100644 fs/sysv/ialloc.c
 delete mode 100644 fs/sysv/inode.c
 delete mode 100644 fs/sysv/itree.c
 delete mode 100644 fs/sysv/namei.c
 delete mode 100644 fs/sysv/super.c
 delete mode 100644 fs/sysv/sysv.h
 delete mode 100644 include/linux/sysv_fs.h
Documentation/filesystems/index.rst +0 -1
index 44e9e77ffe0d..db7a6419752b 100644
@@ -117,7 +117,6 @@ Documentation for filesystem implementations.
    spufs/index
    squashfs
    sysfs
-   sysv-fs
    tmpfs
    ubifs
    ubifs-authentication
Documentation/filesystems/sysv-fs.rst deleted +0 -264

264 lines not on this page · raw message

MAINTAINERS +0 -6
index 30cbc3d44cd5..ff8cc2663502 100644
@@ -22834,12 +22834,6 @@ L:	[email protected]
 S:	Maintained
 F:	drivers/platform/x86/system76_acpi.c
 
-SYSV FILESYSTEM
-S:	Orphan
-F:	Documentation/filesystems/sysv-fs.rst
-F:	fs/sysv/
-F:	include/linux/sysv_fs.h
-
 TASKSTATS STATISTICS INTERFACE
 M:	Balbir Singh <[email protected]>
 S:	Maintained
arch/loongarch/configs/loongson3_defconfig +0 -1
index 4dffc90192f7..d24be66176d7 100644
@@ -978,7 +978,6 @@ CONFIG_MINIX_FS=m
 CONFIG_ROMFS_FS=m
 CONFIG_PSTORE=m
 CONFIG_PSTORE_COMPRESS=y
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_EROFS_FS_ZIP_LZMA=y
arch/m68k/configs/amiga_defconfig +0 -1
index c705247e7b5b..70e2a3f2dffb 100644
@@ -486,7 +486,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/apollo_defconfig +0 -1
index 6d62b9187a58..4665d78feebc 100644
@@ -443,7 +443,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/atari_defconfig +0 -1
index c3c644df852d..d5e554dfbac5 100644
@@ -463,7 +463,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/bvme6000_defconfig +0 -1
index 20261f819691..269111675480 100644
@@ -435,7 +435,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/hp300_defconfig +0 -1
index ce4fe93a0f70..a0fabbd6513a 100644
@@ -445,7 +445,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/mac_defconfig +0 -1
index 040ae75f47c3..41db80850946 100644
@@ -462,7 +462,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/multi_defconfig +0 -1
index 20d877cb4e30..083009e3480d 100644
@@ -549,7 +549,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/mvme147_defconfig +0 -1
index 5e1c8d0d3da5..f33b59279aea 100644
@@ -435,7 +435,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/mvme16x_defconfig +0 -1
index 5d1409e6a137..ba74cd5047f2 100644
@@ -436,7 +436,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/q40_defconfig +0 -1
index e4c30e2b9bbb..3f79ecc95e78 100644
@@ -452,7 +452,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/sun3_defconfig +0 -1
index 980843a9ea1e..c5f3c30075bd 100644
@@ -433,7 +433,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/m68k/configs/sun3x_defconfig +0 -1
index 38681cc6b598..cc94736d4b3f 100644
@@ -433,7 +433,6 @@ CONFIG_OMFS_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_QNX6FS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_EROFS_FS=m
 CONFIG_NFS_FS=y
arch/mips/configs/malta_defconfig +0 -1
index 4390d30206d9..e308b82c094a 100644
@@ -347,7 +347,6 @@ CONFIG_CRAMFS=m
 CONFIG_VXFS_FS=m
 CONFIG_MINIX_FS=m
 CONFIG_ROMFS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_NFS_FS=y
 CONFIG_ROOT_NFS=y
arch/mips/configs/malta_kvm_defconfig +0 -1
index d63d8be8cb50..fa5b04063ddb 100644
@@ -354,7 +354,6 @@ CONFIG_CRAMFS=m
 CONFIG_VXFS_FS=m
 CONFIG_MINIX_FS=m
 CONFIG_ROMFS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_NFS_FS=y
 CONFIG_ROOT_NFS=y
arch/mips/configs/maltaup_xpa_defconfig +0 -1
index 338bb6544a93..40283171af68 100644
@@ -353,7 +353,6 @@ CONFIG_CRAMFS=m
 CONFIG_VXFS_FS=m
 CONFIG_MINIX_FS=m
 CONFIG_ROMFS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_NFS_FS=y
 CONFIG_ROOT_NFS=y
arch/mips/configs/rm200_defconfig +0 -1
index 08e1c1f2f4de..3a6d0384b774 100644
@@ -336,7 +336,6 @@ CONFIG_MINIX_FS=m
 CONFIG_HPFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_ROMFS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_NFS_FS=m
 CONFIG_NFSD=m
arch/parisc/configs/generic-64bit_defconfig +0 -1
index 19a804860ed5..ea623f910748 100644
@@ -268,7 +268,6 @@ CONFIG_PROC_KCORE=y
 CONFIG_TMPFS=y
 CONFIG_TMPFS_XATTR=y
 CONFIG_CONFIGFS_FS=y
-CONFIG_SYSV_FS=y
 CONFIG_NFS_FS=m
 CONFIG_NFS_V4=m
 CONFIG_NFS_V4_1=y
arch/powerpc/configs/fsl-emb-nonhw.config +0 -1
index 3009b0efaf34..d6d2a458847b 100644
@@ -112,7 +112,6 @@ CONFIG_QNX4FS_FS=m
 CONFIG_RCU_TRACE=y
 CONFIG_RESET_CONTROLLER=y
 CONFIG_ROOT_NFS=y
-CONFIG_SYSV_FS=m
 CONFIG_SYSVIPC=y
 CONFIG_TMPFS=y
 CONFIG_UBIFS_FS=y
arch/powerpc/configs/ppc6xx_defconfig +0 -1
index ca0c90e95837..364d1a78bc12 100644
@@ -986,7 +986,6 @@ CONFIG_MINIX_FS=m
 CONFIG_OMFS_FS=m
 CONFIG_QNX4FS_FS=m
 CONFIG_ROMFS_FS=m
-CONFIG_SYSV_FS=m
 CONFIG_UFS_FS=m
 CONFIG_NFS_FS=m
 CONFIG_NFS_V3_ACL=y
fs/Kconfig +0 -1
index 64d420e3c475..afe21866d6b4 100644
@@ -336,7 +336,6 @@ source "fs/qnx4/Kconfig"
 source "fs/qnx6/Kconfig"
 source "fs/romfs/Kconfig"
 source "fs/pstore/Kconfig"
-source "fs/sysv/Kconfig"
 source "fs/ufs/Kconfig"
 source "fs/erofs/Kconfig"
 source "fs/vboxsf/Kconfig"
fs/Makefile +0 -1
index 15df0a923d3a..77fd7f7b5d02 100644
@@ -87,7 +87,6 @@ obj-$(CONFIG_NFSD)		+= nfsd/
 obj-$(CONFIG_LOCKD)		+= lockd/
 obj-$(CONFIG_NLS)		+= nls/
 obj-y				+= unicode/
-obj-$(CONFIG_SYSV_FS)		+= sysv/
 obj-$(CONFIG_SMBFS)		+= smb/
 obj-$(CONFIG_HPFS_FS)		+= hpfs/
 obj-$(CONFIG_NTFS3_FS)		+= ntfs3/
fs/sysv/Kconfig deleted +0 -38
deleted file mode 100644
index 67b3f90afbfd..000000000000
@@ -1,38 +0,0 @@
-# SPDX-License-Identifier: GPL-2.0-only
-config SYSV_FS
-	tristate "System V/Xenix/V7/Coherent file system support"
-	depends on BLOCK
-	select BUFFER_HEAD
-	help
-	  SCO, Xenix and Coherent are commercial Unix systems for Intel
-	  machines, and Version 7 was used on the DEC PDP-11. Saying Y
-	  here would allow you to read from their floppies and hard disk
-	  partitions.
-
-	  If you have floppies or hard disk partitions like that, it is likely
-	  that they contain binaries from those other Unix systems; in order
-	  to run these binaries, you will want to install linux-abi which is
-	  a set of kernel modules that lets you run SCO, Xenix, Wyse,
-	  UnixWare, Dell Unix and System V programs under Linux.  It is
-	  available via FTP (user: ftp) from
-	  <ftp://ftp.openlinux.org/pub/people/hch/linux-abi/>).
-	  NOTE: that will work only for binaries from Intel-based systems;
-	  PDP ones will have to wait until somebody ports Linux to -11 ;-)
-
-	  If you only intend to mount files from some other Unix over the
-	  network using NFS, you don't need the System V file system support
-	  (but you need NFS file system support obviously).
-
-	  Note that this option is generally not needed for floppies, since a
-	  good portable way to transport files and directories between unixes
-	  (and even other operating systems) is given by the tar program ("man
-	  tar" or preferably "info tar").  Note also that this option has
-	  nothing whatsoever to do with the option "System V IPC". Read about
-	  the System V file system in
-	  <file:Documentation/filesystems/sysv-fs.rst>.
-	  Saying Y here will enlarge your kernel by about 27 KB.
-
-	  To compile this as a module, choose M here: the module will be called
-	  sysv.
-
-	  If you haven't heard about all of this before, it's safe to say N.
fs/sysv/Makefile deleted +0 -9
deleted file mode 100644
index 17d12ba04b18..000000000000
@@ -1,9 +0,0 @@
-# SPDX-License-Identifier: GPL-2.0-only
-#
-# Makefile for the Linux SystemV/Coherent filesystem routines.
-#
-
-obj-$(CONFIG_SYSV_FS) += sysv.o
-
-sysv-objs := ialloc.o balloc.o inode.o itree.o file.o dir.o \
-	     namei.o super.o
fs/sysv/balloc.c deleted +0 -240

240 lines not on this page · raw message

fs/sysv/dir.c deleted +0 -378

378 lines not on this page · raw message

fs/sysv/file.c deleted +0 -59

59 lines not on this page · raw message

fs/sysv/ialloc.c deleted +0 -235

235 lines not on this page · raw message

fs/sysv/inode.c deleted +0 -354

354 lines not on this page · raw message

fs/sysv/itree.c deleted +0 -511

511 lines not on this page · raw message

fs/sysv/namei.c deleted +0 -280

280 lines not on this page · raw message

fs/sysv/super.c deleted +0 -595

459 lines not on this page · raw message

fs/sysv/sysv.h deleted +0 -245

245 lines not on this page · raw message

include/linux/sysv_fs.h deleted +0 -214

214 lines not on this page · raw message

The rest of this diff is not shown. View the raw message.

Signature · 1 line
2.43.0

Part of the diff is not shown here. View the raw message.

Reviewed-by on v1 1/1
#1 lore raw
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 7 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

FWIW, it was orphaned in 2023:

commit a8cd2990b694ed2c0ef0e8fc80686c664b4ebbe5
Author: Christoph Hellwig <[email protected]>
Date: Thu Feb 16 07:29:22 2023 +0100

orphan sysvfs

This code has been stale for years and I have no way to test it.

Given how long this was broken with no one noticing, and since it's not being adequately tested, I vote we remove it.

Reviewed-by
#2 lore raw
Quoted 22 lines · Jeff Layton
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 7 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

FWIW, it was orphaned in 2023:

commit a8cd2990b694ed2c0ef0e8fc80686c664b4ebbe5
Author: Christoph Hellwig <[email protected]>
Date: Thu Feb 16 07:29:22 2023 +0100

Quoted 7 lines

orphan sysvfs

This code has been stale for years and I have no way to test it.

Given how long this was broken with no one noticing, and since it's not being adequately tested, I vote we remove it.

I concur, if someone really wants this we can always add it back (after making them deal with the bugs):

--D

Quoted 3 lines
#3 lore raw

I disagree with the removal.

This is still being used, but people running Debian will notice such bugs only with the next stable release. Imagine their nasty xmas present when SYSVFS support is gone for no reason.

Better add a test to CI

Ced

Quoted 43 lines · Darrick J. Wong
Quoted 22 lines · Jeff Layton
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 7 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

FWIW, it was orphaned in 2023:

commit a8cd2990b694ed2c0ef0e8fc80686c664b4ebbe5
Author: Christoph Hellwig <[email protected]>
Date: Thu Feb 16 07:29:22 2023 +0100

Quoted 7 lines

orphan sysvfs

This code has been stale for years and I have no way to test it.

Given how long this was broken with no one noticing, and since it's not being adequately tested, I vote we remove it.

I concur, if someone really wants this we can always add it back (after making them deal with the bugs):

--D

Quoted 3 lines
Signature · 3 lines
Cedric Blancher <[email protected]>
[https://plus.google.com/u/0/+CedricBlancher/]
Institute Pasteur
author reviewer
#4 lore raw
Quoted 5 lines · Cedric Blancher

I disagree with the removal.

This is still being used, but people running Debian will notice such bugs only with the next stable release. Imagine their nasty xmas present when SYSVFS support is gone for no reason.

Hum, do you *know* any user of sysv filesystem driver? Because the kernel should be spewing warnings for sleeping in atomic context left and right when you use that driver for 20 years and nobody complained.

Quoted 1 line

Better add a test to CI

That's easier said than done. AFAIK we don't have tools to create the filesystem or verify its integrity. And perhaps most importantly I don't know of anyone wishing to invest their time into keeping this filesystem alive. Are you volunteering to become a SYSV maintainer?

Honza

Quoted 50 lines · Darrick J. Wong
Quoted 43 lines · Jeff Layton
Quoted 22 lines · Jan Kara
Quoted 7 lines

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 7 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

FWIW, it was orphaned in 2023:

commit a8cd2990b694ed2c0ef0e8fc80686c664b4ebbe5
Author: Christoph Hellwig <[email protected]>
Date: Thu Feb 16 07:29:22 2023 +0100

Quoted 7 lines

orphan sysvfs

This code has been stale for years and I have no way to test it.

Given how long this was broken with no one noticing, and since it's not being adequately tested, I vote we remove it.

I concur, if someone really wants this we can always add it back (after making them deal with the bugs):

--D

Quoted 3 lines
Signature · 3 lines
Cedric Blancher <[email protected]>
[https://plus.google.com/u/0/+CedricBlancher/]
Institute Pasteur
Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR
maintainer
#5 lore raw
Quoted 6 lines · Jan Kara
Quoted 1 line

Better add a test to CI

That's easier said than done. AFAIK we don't have tools to create the filesystem or verify its integrity. And perhaps most importantly I don't know of anyone wishing to invest their time into keeping this filesystem alive. Are you volunteering to become a SYSV maintainer?

To be fair, it's very close to minixfs, so the burden wouldn't be all that awful - just need to keep changes in sync, mostly.

That, however, presumes an actual mind - AI and CI won't cut it.

#6 lore raw
In reply to Jan Kara
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 6 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

The sooner we delete filesystems, the fewer syzbot bugs we get about them. Let's delete as many as we can.

#7 lore raw
In reply to Jan Kara
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 6 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

Let's try and kill it. We can always put it back if we have to.

#8 lore raw
Quoted 8 lines · Christian Brauner
Quoted 7 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

Quoted 9 lines
Quoted 6 lines

[1] https://lore.kernel.org/all/[email protected]/

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

Let's try and kill it. We can always put it back if we have to.

So should any work toward converting sysv to the new mount API stop? ;) Thanks-
Bill

#9 lore raw
Quoted 11 lines · Bill O'Donnell
Quoted 10 lines · Christian Brauner
Quoted 9 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

[1] https://lore.kernel.org/all/[email protected]/

Quoted 8 lines
Quoted 7 lines
Quoted 4 lines

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

Let's try and kill it. We can always put it back if we have to.

So should any work toward converting sysv to the new mount API stop? ;)

So I would suggest we try and remove it for v6.15. If the removal survives the release of v6.16 I would call it a (qualified) success.

If during this time we find out that we have to keep it and have to reintroduce it then we may as well spend the time to port it to the new mount api.

Thoughts?

#10 lore raw
Quoted 21 lines · Christian Brauner
Quoted 11 lines · Bill O'Donnell
Quoted 10 lines · Christian Brauner
Quoted 9 lines · Jan Kara

Since 2002 (change "Replace BKL for chain locking with sysvfs-private rwlock") the sysv filesystem was doing IO under a rwlock in its get_block() function (yes, a non-sleepable lock hold over a function used to read inode metadata for all reads and writes). Nobody noticed until syzbot in 2023 [1]. This shows nobody is using the filesystem. Just drop it.

[1] https://lore.kernel.org/all/[email protected]/

Quoted 8 lines
Quoted 7 lines
Quoted 4 lines

---
What do people think about this? Or should we perhaps go through a (short) deprecation period where we warn about removal?

Let's try and kill it. We can always put it back if we have to.

So should any work toward converting sysv to the new mount API stop? ;)

Quoted 8 lines

So I would suggest we try and remove it for v6.15. If the removal survives the release of v6.16 I would call it a (qualified) success.

If during this time we find out that we have to keep it and have to reintroduce it then we may as well spend the time to port it to the new mount api.

Thoughts?

Makes sense to me.