Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-25 05:32:00 +00:00
Code Issues Packages Projects Releases Wiki Activity
d6c1b3599b
linux/fs/jfs/namei.c

1617 lines
37 KiB
C
Raw Normal View History

treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 156 Based on 1 normalized pattern(s): this program is free software you can redistribute it and or modify it under the terms of the gnu general public license as published by the free software foundation either version 2 of the license or at your option any later version this program is distributed in the hope that it will be useful but without any warranty without even the implied warranty of merchantability or fitness for a particular purpose see the gnu general public license for more details you should have received a copy of the gnu general public license along with this program if not write to the free software foundation inc 59 temple place suite 330 boston ma 02111 1307 usa extracted by the scancode license scanner the SPDX license identifier GPL-2.0-or-later has been chosen to replace the boilerplate/reference in 1334 file(s). Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Allison Randal <allison@lohutok.net> Reviewed-by: Richard Fontana <rfontana@redhat.com> Cc: linux-spdx@vger.kernel.org Link: https://lkml.kernel.org/r/20190527070033.113240726@linutronix.de Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-05-27 06:55:05 +00:00
// SPDX-License-Identifier: GPL-2.0-or-later
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* Copyright (C) International Business Machines Corp., 2000-2004
* Portions Copyright (C) Christoph Hellwig, 2001-2002
*/
#include <linux/fs.h>
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
#include <linux/namei.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#include <linux/ctype.h>
#include <linux/quotaops.h>
jfs: new export ops Trivial switch over to the new generic helpers. Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Neil Brown <neilb@suse.de> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Dave Kleikamp <shaggy@austin.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-21 23:42:09 +00:00
#include <linux/exportfs.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#include "jfs_incore.h"
#include "jfs_superblock.h"
#include "jfs_inode.h"
#include "jfs_dinode.h"
#include "jfs_dmap.h"
#include "jfs_unicode.h"
#include "jfs_metapage.h"
#include "jfs_xattr.h"
#include "jfs_acl.h"
#include "jfs_debug.h"
/*
* forward references
*/
constify dentry_operations: JFS Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2009-02-20 06:00:49 +00:00
const struct dentry_operations jfs_ci_dentry_operations;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static s64 commitZeroLink(tid_t, struct inode *);
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
/*
* NAME: free_ea_wmap(inode)
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* FUNCTION: free uncommitted extended attributes from working map
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
*
*/
static inline void free_ea_wmap(struct inode *inode)
{
dxd_t *ea = &JFS_IP(inode)->ea;
if (ea->flag & DXD_EXTENT) {
/* free EA pages from cache */
invalidate_dxd_metapages(inode, *ea);
dbFree(inode, addressDXD(ea), lengthDXD(ea));
}
ea->flag = 0;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* NAME: jfs_create(dip, dentry, mode)
*
* FUNCTION: create a regular file in the parent directory <dip>
* with name = <from dentry> and mode = <mode>
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* PARAMETER: dip - parent directory vnode
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* dentry - dentry of new file
* mode - create mode (rwxrwxrwx).
* nd- nd struct
*
* RETURN: Errors from subroutines
*
*/
fs: port ->create() to pass mnt_idmap Convert to struct mnt_idmap. Last cycle we merged the necessary infrastructure in 256c8aed2b42 ("fs: introduce dedicated idmap type for mounts"). This is just the conversion to struct mnt_idmap. Currently we still pass around the plain namespace that was attached to a mount. This is in general pretty convenient but it makes it easy to conflate namespaces that are relevant on the filesystem with namespaces that are relevent on the mount level. Especially for non-vfs developers without detailed knowledge in this area this can be a potential source for bugs. Once the conversion to struct mnt_idmap is done all helpers down to the really low-level helpers will take a struct mnt_idmap argument instead of two namespace arguments. This way it becomes impossible to conflate the two eliminating the possibility of any bugs. All of the vfs and all filesystems only operate on struct mnt_idmap. Acked-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2023-01-13 11:49:13 +00:00
static int jfs_create(struct mnt_idmap *idmap, struct inode *dip,
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
struct dentry *dentry, umode_t mode, bool excl)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int rc = 0;
tid_t tid; /* transaction id */
struct inode *ip = NULL; /* child directory inode */
ino_t ino;
struct component_name dname; /* child directory name */
struct btstack btstack;
struct inode *iplist[2];
struct tblock *tblk;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_create: dip:0x%p name:%pd", dip, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dip);
if (rc)
goto out1;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* search parent directory for entry/freespace
* (dtSearch() returns parent directory page pinned)
*/
if ((rc = get_UCSname(&dname, dentry)))
goto out1;
/*
* Either iAlloc() or txBegin() may block. Deadlock can occur if we
* block there while holding dtree page, so we allocate the inode &
* begin the transaction before we search the directory.
*/
ip = ialloc(dip, mode);
[PATCH] JFS: return correct error when i-node allocation failed I have seen confusing behavior on JFS when I injected many intentional slab allocation errors. The cp command failed with no disk space error with enough disk space. This patch makes: - change the return value in case slab allocation failures happen from -ENOSPC to -ENOMEM - ialloc() return error code so that the caller can know the reason of failures Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)
2006-09-14 14:22:38 +00:00
if (IS_ERR(ip)) {
rc = PTR_ERR(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out2;
}
tid = txBegin(dip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
rc = jfs_init_acl(tid, ip, dip);
if (rc)
goto out3;
fs/vfs/security: pass last path component to LSM on inode creation SELinux would like to implement a new labeling behavior of newly created inodes. We currently label new inodes based on the parent and the creating process. This new behavior would also take into account the name of the new object when deciding the new label. This is not the (supposed) full path, just the last component of the path. This is very useful because creating /etc/shadow is different than creating /etc/passwd but the kernel hooks are unable to differentiate these operations. We currently require that userspace realize it is doing some difficult operation like that and than userspace jumps through SELinux hoops to get things set up correctly. This patch does not implement new behavior, that is obviously contained in a seperate SELinux patch, but it does pass the needed name down to the correct LSM hook. If no such name exists it is fine to pass NULL. Signed-off-by: Eric Paris <eparis@redhat.com>
2011-02-01 16:05:39 +00:00
rc = jfs_init_security(tid, ip, dip, &dentry->d_name);
JFS: Implement jfs_init_security This atomically initializes the security xattr when an object is created Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:05:39 +00:00
if (rc) {
txAbort(tid, 0);
goto out3;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = dtSearch(dip, &dname, &ino, &btstack, JFS_CREATE))) {
jfs_err("jfs_create: dtSearch returned %d", rc);
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
txAbort(tid, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out3;
}
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_CREATE;
tblk->ino = ip->i_ino;
tblk->u.ixpxd = JFS_IP(ip)->ixpxd;
iplist[0] = dip;
iplist[1] = ip;
/*
* initialize the child XAD tree root in-line in inode
*/
xtInitRoot(tid, ip);
/*
* create entry in parent directory for child directory
* (dtInsert() releases parent directory page)
*/
ino = ip->i_ino;
if ((rc = dtInsert(tid, dip, &dname, &ino, &btstack))) {
if (rc == -EIO) {
jfs_err("jfs_create: dtInsert returned -EIO");
txAbort(tid, 1); /* Marks Filesystem dirty */
} else
txAbort(tid, 0); /* Filesystem full */
goto out3;
}
ip->i_op = &jfs_file_inode_operations;
ip->i_fop = &jfs_file_operations;
ip->i_mapping->a_ops = &jfs_aops;
mark_inode_dirty(ip);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dip, inode_set_ctime_current(dip));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(dip);
rc = txCommit(tid, 2, &iplist[0], 0);
out3:
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (rc) {
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
free_ea_wmap(ip);
filesystems: add missing nlink wrappers Replace direct i_nlink updates with the respective updater function (inc_nlink, drop_nlink, clear_nlink, inode_dec_link_count). Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-10-28 12:13:28 +00:00
clear_nlink(ip);
jfs: switch to discard_new_inode() we don't want open-by-handle to pick an in-core inode that has failed setup halfway through. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-06-29 15:59:37 +00:00
discard_new_inode(ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
} else {
do d_instantiate/unlock_new_inode combinations safely For anything NFS-exported we do _not_ want to unlock new inode before it has grown an alias; original set of fixes got the ordering right, but missed the nasty complication in case of lockdep being enabled - unlock_new_inode() does lockdep_annotate_inode_mutex_key(inode) which can only be done before anyone gets a chance to touch ->i_mutex. Unfortunately, flipping the order and doing unlock_new_inode() before d_instantiate() opens a window when mkdir can race with open-by-fhandle on a guessed fhandle, leading to multiple aliases for a directory inode and all the breakage that follows from that. Correct solution: a new primitive (d_instantiate_new()) combining these two in the right order - lockdep annotate, then d_instantiate(), then the rest of unlock_new_inode(). All combinations of d_instantiate() with unlock_new_inode() should be converted to that. Cc: stable@kernel.org # 2.6.29 and later Tested-by: Mike Marshall <hubcap@omnibond.com> Reviewed-by: Andreas Dilger <adilger@dilger.ca> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-05-04 12:23:01 +00:00
d_instantiate_new(dentry, ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out2:
free_UCSname(&dname);
out1:
jfs_info("jfs_create: rc:%d", rc);
return rc;
}
/*
* NAME: jfs_mkdir(dip, dentry, mode)
*
* FUNCTION: create a child directory in the parent directory <dip>
* with name = <from dentry> and mode = <mode>
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* PARAMETER: dip - parent directory vnode
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* dentry - dentry of child directory
* mode - create mode (rwxrwxrwx).
*
* RETURN: Errors from subroutines
*
* note:
jfs: fix spelling mistake, EACCESS -> EACCES Trivial fix to a spelling mistake of the error access name EACCESS, rename to EACCES Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2018-10-26 18:09:09 +00:00
* EACCES: user needs search+write permission on the parent directory
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
fs: port ->mkdir() to pass mnt_idmap Convert to struct mnt_idmap. Last cycle we merged the necessary infrastructure in 256c8aed2b42 ("fs: introduce dedicated idmap type for mounts"). This is just the conversion to struct mnt_idmap. Currently we still pass around the plain namespace that was attached to a mount. This is in general pretty convenient but it makes it easy to conflate namespaces that are relevant on the filesystem with namespaces that are relevent on the mount level. Especially for non-vfs developers without detailed knowledge in this area this can be a potential source for bugs. Once the conversion to struct mnt_idmap is done all helpers down to the really low-level helpers will take a struct mnt_idmap argument instead of two namespace arguments. This way it becomes impossible to conflate the two eliminating the possibility of any bugs. All of the vfs and all filesystems only operate on struct mnt_idmap. Acked-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2023-01-13 11:49:15 +00:00
static int jfs_mkdir(struct mnt_idmap *idmap, struct inode *dip,
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
struct dentry *dentry, umode_t mode)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int rc = 0;
tid_t tid; /* transaction id */
struct inode *ip = NULL; /* child directory inode */
ino_t ino;
struct component_name dname; /* child directory name */
struct btstack btstack;
struct inode *iplist[2];
struct tblock *tblk;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_mkdir: dip:0x%p name:%pd", dip, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dip);
if (rc)
goto out1;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* search parent directory for entry/freespace
* (dtSearch() returns parent directory page pinned)
*/
if ((rc = get_UCSname(&dname, dentry)))
goto out1;
/*
* Either iAlloc() or txBegin() may block. Deadlock can occur if we
* block there while holding dtree page, so we allocate the inode &
* begin the transaction before we search the directory.
*/
ip = ialloc(dip, S_IFDIR | mode);
[PATCH] JFS: return correct error when i-node allocation failed I have seen confusing behavior on JFS when I injected many intentional slab allocation errors. The cp command failed with no disk space error with enough disk space. This patch makes: - change the return value in case slab allocation failures happen from -ENOSPC to -ENOMEM - ialloc() return error code so that the caller can know the reason of failures Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)
2006-09-14 14:22:38 +00:00
if (IS_ERR(ip)) {
rc = PTR_ERR(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out2;
}
tid = txBegin(dip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
rc = jfs_init_acl(tid, ip, dip);
if (rc)
goto out3;
fs/vfs/security: pass last path component to LSM on inode creation SELinux would like to implement a new labeling behavior of newly created inodes. We currently label new inodes based on the parent and the creating process. This new behavior would also take into account the name of the new object when deciding the new label. This is not the (supposed) full path, just the last component of the path. This is very useful because creating /etc/shadow is different than creating /etc/passwd but the kernel hooks are unable to differentiate these operations. We currently require that userspace realize it is doing some difficult operation like that and than userspace jumps through SELinux hoops to get things set up correctly. This patch does not implement new behavior, that is obviously contained in a seperate SELinux patch, but it does pass the needed name down to the correct LSM hook. If no such name exists it is fine to pass NULL. Signed-off-by: Eric Paris <eparis@redhat.com>
2011-02-01 16:05:39 +00:00
rc = jfs_init_security(tid, ip, dip, &dentry->d_name);
JFS: Implement jfs_init_security This atomically initializes the security xattr when an object is created Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:05:39 +00:00
if (rc) {
txAbort(tid, 0);
goto out3;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = dtSearch(dip, &dname, &ino, &btstack, JFS_CREATE))) {
jfs_err("jfs_mkdir: dtSearch returned %d", rc);
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
txAbort(tid, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out3;
}
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_CREATE;
tblk->ino = ip->i_ino;
tblk->u.ixpxd = JFS_IP(ip)->ixpxd;
iplist[0] = dip;
iplist[1] = ip;
/*
* initialize the child directory in-line in inode
*/
dtInitRoot(tid, ip, dip->i_ino);
/*
* create entry in parent directory for child directory
* (dtInsert() releases parent directory page)
*/
ino = ip->i_ino;
if ((rc = dtInsert(tid, dip, &dname, &ino, &btstack))) {
if (rc == -EIO) {
jfs_err("jfs_mkdir: dtInsert returned -EIO");
txAbort(tid, 1); /* Marks Filesystem dirty */
} else
txAbort(tid, 0); /* Filesystem full */
goto out3;
}
filesystems: add set_nlink() Replace remaining direct i_nlink updates with a new set_nlink() updater function. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Tested-by: Toshiyuki Okajima <toshi.okajima@jp.fujitsu.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
2011-10-28 12:13:29 +00:00
set_nlink(ip, 2); /* for '.' */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ip->i_op = &jfs_dir_inode_operations;
ip->i_fop = &jfs_dir_operations;
mark_inode_dirty(ip);
/* update parent directory inode */
[PATCH] r/o bind mount prepwork: inc_nlink() helper This is mostly included for parity with dec_nlink(), where we will have some more hooks. This one should stay pretty darn straightforward for now. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:04 +00:00
inc_nlink(dip); /* for '..' from child directory */
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dip, inode_set_ctime_current(dip));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(dip);
rc = txCommit(tid, 2, &iplist[0], 0);
out3:
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (rc) {
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
free_ea_wmap(ip);
filesystems: add missing nlink wrappers Replace direct i_nlink updates with the respective updater function (inc_nlink, drop_nlink, clear_nlink, inode_dec_link_count). Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-10-28 12:13:28 +00:00
clear_nlink(ip);
jfs: switch to discard_new_inode() we don't want open-by-handle to pick an in-core inode that has failed setup halfway through. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-06-29 15:59:37 +00:00
discard_new_inode(ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
} else {
do d_instantiate/unlock_new_inode combinations safely For anything NFS-exported we do _not_ want to unlock new inode before it has grown an alias; original set of fixes got the ordering right, but missed the nasty complication in case of lockdep being enabled - unlock_new_inode() does lockdep_annotate_inode_mutex_key(inode) which can only be done before anyone gets a chance to touch ->i_mutex. Unfortunately, flipping the order and doing unlock_new_inode() before d_instantiate() opens a window when mkdir can race with open-by-fhandle on a guessed fhandle, leading to multiple aliases for a directory inode and all the breakage that follows from that. Correct solution: a new primitive (d_instantiate_new()) combining these two in the right order - lockdep annotate, then d_instantiate(), then the rest of unlock_new_inode(). All combinations of d_instantiate() with unlock_new_inode() should be converted to that. Cc: stable@kernel.org # 2.6.29 and later Tested-by: Mike Marshall <hubcap@omnibond.com> Reviewed-by: Andreas Dilger <adilger@dilger.ca> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-05-04 12:23:01 +00:00
d_instantiate_new(dentry, ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out2:
free_UCSname(&dname);
out1:
jfs_info("jfs_mkdir: rc:%d", rc);
return rc;
}
/*
* NAME: jfs_rmdir(dip, dentry)
*
* FUNCTION: remove a link to child directory
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* PARAMETER: dip - parent inode
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* dentry - child directory dentry
*
* RETURN: -EINVAL - if name is . or ..
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* -EINVAL - if . or .. exist but are invalid.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* errors from subroutines
*
* note:
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* if other threads have the directory open when the last link
* is removed, the "." and ".." entries, if present, are removed before
* rmdir() returns and no new entries may be created in the directory,
* but the directory is not removed until the last reference to
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* the directory is released (cf.unlink() of regular file).
*/
static int jfs_rmdir(struct inode *dip, struct dentry *dentry)
{
int rc;
tid_t tid; /* transaction id */
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
struct inode *ip = d_inode(dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ino_t ino;
struct component_name dname;
struct inode *iplist[2];
struct tblock *tblk;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_rmdir: dip:0x%p name:%pd", dip, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Init inode for quota operations. */
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dip);
if (rc)
goto out;
rc = dquot_initialize(ip);
if (rc)
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* directory must be empty to be removed */
if (!dtEmpty(ip)) {
rc = -ENOTEMPTY;
goto out;
}
if ((rc = get_UCSname(&dname, dentry))) {
goto out;
}
tid = txBegin(dip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
iplist[0] = dip;
iplist[1] = ip;
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_DELETE;
tblk->u.ip = ip;
/*
* delete the entry of target directory from parent directory
*/
ino = ip->i_ino;
if ((rc = dtDelete(tid, dip, &dname, &ino, JFS_REMOVE))) {
jfs_err("jfs_rmdir: dtDelete returned %d", rc);
if (rc == -EIO)
txAbort(tid, 1);
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out2;
}
/* update parent directory's link count corresponding
* to ".." entry of the target directory deleted
*/
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dip, inode_set_ctime_current(dip));
[PATCH] r/o bind mounts: unlink: monitor i_nlink When a filesystem decrements i_nlink to zero, it means that a write must be performed in order to drop the inode from the filesystem. We're shortly going to have keep filesystems from being remounted r/o between the time that this i_nlink decrement and that write occurs. So, add a little helper function to do the decrements. We'll tie into it in a bit to note when i_nlink hits zero. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:03 +00:00
inode_dec_link_count(dip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* OS/2 could have created EA and/or ACL
*/
/* free EA from both persistent and working map */
if (JFS_IP(ip)->ea.flag & DXD_EXTENT) {
/* free EA pages */
txEA(tid, ip, &JFS_IP(ip)->ea, NULL);
}
JFS_IP(ip)->ea.flag = 0;
/* free ACL from both persistent and working map */
if (JFS_IP(ip)->acl.flag & DXD_EXTENT) {
/* free ACL pages */
txEA(tid, ip, &JFS_IP(ip)->acl, NULL);
}
JFS_IP(ip)->acl.flag = 0;
/* mark the target directory as deleted */
[PATCH] r/o bind mounts: monitor zeroing of i_nlink Some filesystems, instead of simply decrementing i_nlink, simply zero it during an unlink operation. We need to catch these in addition to the decrement operations. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:06 +00:00
clear_nlink(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(ip);
rc = txCommit(tid, 2, &iplist[0], 0);
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* Truncating the directory index table is not guaranteed. It
* may need to be done iteratively
*/
if (test_cflag(COMMIT_Stale, dip)) {
if (dip->i_size > 1)
jfs_truncate_nolock(dip, 0);
clear_cflag(COMMIT_Stale, dip);
}
out2:
free_UCSname(&dname);
out:
jfs_info("jfs_rmdir: rc:%d", rc);
return rc;
}
/*
* NAME: jfs_unlink(dip, dentry)
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* FUNCTION: remove a link to object <vp> named by <name>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* from parent directory <dvp>
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* PARAMETER: dip - inode of parent directory
* dentry - dentry of object to be removed
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* RETURN: errors from subroutines
*
* note:
* temporary file: if one or more processes have the file open
* when the last link is removed, the link will be removed before
* unlink() returns, but the removal of the file contents will be
* postponed until all references to the files are closed.
*
* JFS does NOT support unlink() on directories.
*
*/
static int jfs_unlink(struct inode *dip, struct dentry *dentry)
{
int rc;
tid_t tid; /* transaction id */
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
struct inode *ip = d_inode(dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ino_t ino;
struct component_name dname; /* object name */
struct inode *iplist[2];
struct tblock *tblk;
s64 new_size = 0;
int commit_flag;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_unlink: dip:0x%p name:%pd", dip, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Init inode for quota operations. */
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dip);
if (rc)
goto out;
rc = dquot_initialize(ip);
if (rc)
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = get_UCSname(&dname, dentry)))
goto out;
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
IWRITE_LOCK(ip, RDWRLOCK_NORMAL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tid = txBegin(dip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
iplist[0] = dip;
iplist[1] = ip;
/*
* delete the entry of target file from parent directory
*/
ino = ip->i_ino;
if ((rc = dtDelete(tid, dip, &dname, &ino, JFS_REMOVE))) {
jfs_err("jfs_unlink: dtDelete returned %d", rc);
if (rc == -EIO)
txAbort(tid, 1); /* Marks FS Dirty */
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
IWRITE_UNLOCK(ip);
goto out1;
}
ASSERT(ip->i_nlink);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dip,
inode_set_ctime_to_ts(dip, inode_set_ctime_current(ip)));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(dip);
/* update target's inode */
[PATCH] r/o bind mounts: unlink: monitor i_nlink When a filesystem decrements i_nlink to zero, it means that a write must be performed in order to drop the inode from the filesystem. We're shortly going to have keep filesystems from being remounted r/o between the time that this i_nlink decrement and that write occurs. So, add a little helper function to do the decrements. We'll tie into it in a bit to note when i_nlink hits zero. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:03 +00:00
inode_dec_link_count(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* commit zero link count object
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
if (ip->i_nlink == 0) {
assert(!test_cflag(COMMIT_Nolink, ip));
/* free block resources */
if ((new_size = commitZeroLink(tid, ip)) < 0) {
txAbort(tid, 1); /* Marks FS Dirty */
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
IWRITE_UNLOCK(ip);
rc = new_size;
goto out1;
}
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_DELETE;
tblk->u.ip = ip;
}
/*
* Incomplete truncate of file data can
* result in timing problems unless we synchronously commit the
* transaction.
*/
if (new_size)
commit_flag = COMMIT_SYNC;
else
commit_flag = 0;
/*
* If xtTruncate was incomplete, commit synchronously to avoid
* timing complications
*/
rc = txCommit(tid, 2, &iplist[0], commit_flag);
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
while (new_size && (rc == 0)) {
tid = txBegin(dip->i_sb, 0);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_lock(&JFS_IP(ip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
new_size = xtTruncate_pmap(tid, ip, new_size);
if (new_size < 0) {
txAbort(tid, 1); /* Marks FS Dirty */
rc = new_size;
} else
rc = txCommit(tid, 2, &iplist[0], COMMIT_SYNC);
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
if (ip->i_nlink == 0)
set_cflag(COMMIT_Nolink, ip);
IWRITE_UNLOCK(ip);
/*
* Truncating the directory index table is not guaranteed. It
* may need to be done iteratively
*/
if (test_cflag(COMMIT_Stale, dip)) {
if (dip->i_size > 1)
jfs_truncate_nolock(dip, 0);
clear_cflag(COMMIT_Stale, dip);
}
out1:
free_UCSname(&dname);
out:
jfs_info("jfs_unlink: rc:%d", rc);
return rc;
}
/*
* NAME: commitZeroLink()
*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* FUNCTION: for non-directory, called by jfs_remove(),
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* truncate a regular file, directory or symbolic
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* link to zero length. return 0 if type is not
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* one of these.
*
* if the file is currently associated with a VM segment
* only permanent disk and inode map resources are freed,
* and neither the inode nor indirect blocks are modified
* so that the resources can be later freed in the work
* map by ctrunc1.
* if there is no VM segment on entry, the resources are
* freed in both work and permanent map.
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* (? for temporary file - memory object is cached even
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* after no reference:
* reference count > 0 - )
*
* PARAMETERS: cd - pointer to commit data structure.
* current inode is the one to truncate.
*
* RETURN: Errors from subroutines
*/
static s64 commitZeroLink(tid_t tid, struct inode *ip)
{
int filetype;
struct tblock *tblk;
jfs_info("commitZeroLink: tid = %d, ip = 0x%p", tid, ip);
filetype = ip->i_mode & S_IFMT;
switch (filetype) {
case S_IFREG:
break;
case S_IFLNK:
/* fast symbolic link */
if (ip->i_size < IDATASIZE) {
ip->i_size = 0;
return 0;
}
break;
default:
assert(filetype != S_IFDIR);
return 0;
}
set_cflag(COMMIT_Freewmap, ip);
/* mark transaction of block map update type */
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_PMAP;
/*
* free EA
*/
if (JFS_IP(ip)->ea.flag & DXD_EXTENT)
/* acquire maplock on EA to be freed from block map */
txEA(tid, ip, &JFS_IP(ip)->ea, NULL);
/*
* free ACL
*/
if (JFS_IP(ip)->acl.flag & DXD_EXTENT)
/* acquire maplock on EA to be freed from block map */
txEA(tid, ip, &JFS_IP(ip)->acl, NULL);
/*
* free xtree/data (truncate to zero length):
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* free xtree/data pages from cache if COMMIT_PWMAP,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* free xtree/data blocks from persistent block map, and
* free xtree/data blocks from working block map if COMMIT_PWMAP;
*/
if (ip->i_size)
return xtTruncate_pmap(tid, ip, 0);
return 0;
}
/*
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
* NAME: jfs_free_zero_link()
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* FUNCTION: for non-directory, called by iClose(),
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* free resources of a file from cache and WORKING map
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* for a file previously committed with zero link count
* while associated with a pager object,
*
* PARAMETER: ip - pointer to inode of file.
*/
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
void jfs_free_zero_link(struct inode *ip)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int type;
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
jfs_info("jfs_free_zero_link: ip = 0x%p", ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* return if not reg or symbolic link or if size is
* already ok.
*/
type = ip->i_mode & S_IFMT;
switch (type) {
case S_IFREG:
break;
case S_IFLNK:
/* if its contained in inode nothing to do */
if (ip->i_size < IDATASIZE)
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
default:
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
return;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/*
* free EA
*/
if (JFS_IP(ip)->ea.flag & DXD_EXTENT) {
s64 xaddr = addressDXD(&JFS_IP(ip)->ea);
int xlen = lengthDXD(&JFS_IP(ip)->ea);
struct maplock maplock; /* maplock for COMMIT_WMAP */
struct pxd_lock *pxdlock; /* maplock for COMMIT_WMAP */
/* free EA pages from cache */
invalidate_dxd_metapages(ip, JFS_IP(ip)->ea);
/* free EA extent from working block map */
maplock.index = 1;
pxdlock = (struct pxd_lock *) & maplock;
pxdlock->flag = mlckFREEPXD;
PXDaddress(&pxdlock->pxd, xaddr);
PXDlength(&pxdlock->pxd, xlen);
txFreeMap(ip, pxdlock, NULL, COMMIT_WMAP);
}
/*
* free ACL
*/
if (JFS_IP(ip)->acl.flag & DXD_EXTENT) {
s64 xaddr = addressDXD(&JFS_IP(ip)->acl);
int xlen = lengthDXD(&JFS_IP(ip)->acl);
struct maplock maplock; /* maplock for COMMIT_WMAP */
struct pxd_lock *pxdlock; /* maplock for COMMIT_WMAP */
invalidate_dxd_metapages(ip, JFS_IP(ip)->acl);
/* free ACL extent from working block map */
maplock.index = 1;
pxdlock = (struct pxd_lock *) & maplock;
pxdlock->flag = mlckFREEPXD;
PXDaddress(&pxdlock->pxd, xaddr);
PXDlength(&pxdlock->pxd, xlen);
txFreeMap(ip, pxdlock, NULL, COMMIT_WMAP);
}
/*
* free xtree/data (truncate to zero length):
* free xtree/data pages from cache, and
* free xtree/data blocks from working block map;
*/
if (ip->i_size)
JFS: fix sparse warnings by moving extern declarations to headers Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-05-04 20:29:35 +00:00
xtTruncate(0, ip, 0, COMMIT_WMAP);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/*
* NAME: jfs_link(vp, dvp, name, crp)
*
* FUNCTION: create a link to <vp> by the name = <name>
* in the parent directory <dvp>
*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* PARAMETER: vp - target object
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* dvp - parent directory of new link
* name - name of new link to target object
* crp - credential
*
* RETURN: Errors from subroutines
*
* note:
* JFS does NOT support link() on directories (to prevent circular
* path in the directory hierarchy);
* EPERM: the target object is a directory, and either the caller
* does not have appropriate privileges or the implementation prohibits
* using link() on directories [XPG4.2].
*
* JFS does NOT support links between file systems:
* EXDEV: target object and new link are on different file systems and
* implementation does not support links between file systems [XPG4.2].
*/
static int jfs_link(struct dentry *old_dentry,
struct inode *dir, struct dentry *dentry)
{
int rc;
tid_t tid;
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
struct inode *ip = d_inode(old_dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ino_t ino;
struct component_name dname;
struct btstack btstack;
struct inode *iplist[2];
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_link: %pd %pd", old_dentry, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dir);
if (rc)
goto out;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
FS: JFS: Fix null-ptr-deref Read in txBegin Syzkaller reported an issue where txBegin may be called on a superblock in a read-only mounted filesystem which leads to NULL pointer deref. This could be solved by checking if the filesystem is read-only before calling txBegin, and returning with appropiate error code. Reported-By: syzbot+f1faa20eec55e0c8644c@syzkaller.appspotmail.com Link: https://syzkaller.appspot.com/bug?id=be7e52c50c5182cc09a09ea6fc456446b2039de3 Signed-off-by: Immad Mir <mirimmad17@gmail.com> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2023-06-23 13:44:01 +00:00
if (isReadOnly(ip)) {
jfs_error(ip->i_sb, "read-only filesystem\n");
return -EROFS;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tid = txBegin(ip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dir)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* scan parent directory for entry/freespace
*/
if ((rc = get_UCSname(&dname, dentry)))
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
goto out_tx;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = dtSearch(dir, &dname, &ino, &btstack, JFS_CREATE)))
goto free_dname;
/*
* create entry for new link in parent directory
*/
ino = ip->i_ino;
if ((rc = dtInsert(tid, dir, &dname, &ino, &btstack)))
goto free_dname;
/* update object inode */
[PATCH] r/o bind mount prepwork: inc_nlink() helper This is mostly included for parity with dec_nlink(), where we will have some more hooks. This one should stay pretty darn straightforward for now. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:04 +00:00
inc_nlink(ip); /* for new link */
jfs: convert to ctime accessor functions In later patches, we're going to change how the inode's ctime field is used. Switch to using accessor functions instead of raw accesses of inode->i_ctime. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Jan Kara <jack@suse.cz> Message-Id: <20230705190309.579783-53-jlayton@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-07-05 19:01:20 +00:00
inode_set_ctime_current(ip);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dir, inode_set_ctime_current(dir));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(dir);
new helper: ihold() Clones an existing reference to inode; caller must already hold one. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2010-10-23 15:11:40 +00:00
ihold(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
iplist[0] = ip;
iplist[1] = dir;
rc = txCommit(tid, 2, &iplist[0], 0);
if (rc) {
filesystems: add missing nlink wrappers Replace direct i_nlink updates with the respective updater function (inc_nlink, drop_nlink, clear_nlink, inode_dec_link_count). Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-10-28 12:13:28 +00:00
drop_nlink(ip); /* never instantiated */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
iput(ip);
} else
d_instantiate(dentry, ip);
free_dname:
free_UCSname(&dname);
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
out_tx:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dir)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
out:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs_info("jfs_link: rc:%d", rc);
return rc;
}
/*
* NAME: jfs_symlink(dip, dentry, name)
*
* FUNCTION: creates a symbolic link to <symlink> by name <name>
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* in directory <dip>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* PARAMETER: dip - parent directory vnode
* dentry - dentry of symbolic link
* name - the path name of the existing object
* that will be the source of the link
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
* RETURN: errors from subroutines
*
* note:
* ENAMETOOLONG: pathname resolution of a symbolic link produced
* an intermediate result whose length exceeds PATH_MAX [XPG4.2]
*/
fs: port ->symlink() to pass mnt_idmap Convert to struct mnt_idmap. Last cycle we merged the necessary infrastructure in 256c8aed2b42 ("fs: introduce dedicated idmap type for mounts"). This is just the conversion to struct mnt_idmap. Currently we still pass around the plain namespace that was attached to a mount. This is in general pretty convenient but it makes it easy to conflate namespaces that are relevant on the filesystem with namespaces that are relevent on the mount level. Especially for non-vfs developers without detailed knowledge in this area this can be a potential source for bugs. Once the conversion to struct mnt_idmap is done all helpers down to the really low-level helpers will take a struct mnt_idmap argument instead of two namespace arguments. This way it becomes impossible to conflate the two eliminating the possibility of any bugs. All of the vfs and all filesystems only operate on struct mnt_idmap. Acked-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2023-01-13 11:49:14 +00:00
static int jfs_symlink(struct mnt_idmap *idmap, struct inode *dip,
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
struct dentry *dentry, const char *name)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int rc;
tid_t tid;
ino_t ino = 0;
struct component_name dname;
jfs: Use unsigned variable for length calculations To avoid confusing the compiler about possible negative sizes, switch "ssize" which can never be negative from int to u32. Seen with GCC 13: ../fs/jfs/namei.c: In function 'jfs_symlink': ../include/linux/fortify-string.h:57:33: warning: '__builtin_memcpy' pointer overflow between offset 0 and size [-2147483648, -1] [-Warray-bounds=] 57 | #define __underlying_memcpy __builtin_memcpy | ^ ... ../fs/jfs/namei.c:950:17: note: in expansion of macro 'memcpy' 950 | memcpy(ip->i_link, name, ssize); | ^~~~~~ Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Chinner <dchinner@redhat.com> Cc: jfs-discussion@lists.sourceforge.net Signed-off-by: Kees Cook <keescook@chromium.org> Acked-by: Jeff Xu <jeffxu@chromium.org> Message-Id: <20230204183355.never.877-kees@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-02-04 18:33:56 +00:00
u32 ssize; /* source pathname size */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct btstack btstack;
jfs: remove redundant initialization to pointer ip The pointer ip is being initialized with a value that is never read, it is being re-assigned later on. The assignment is redundant and can be removed. Cleans up clang scan warning: fs/jfs/namei.c:886:16: warning: Value stored to 'ip' during its initialization is never read [deadcode.DeadStores] Signed-off-by: Colin Ian King <colin.i.king@gmail.com> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2023-08-18 15:02:10 +00:00
struct inode *ip;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
s64 xlen = 0;
int bmask = 0, xsize;
jfs: clean up some compiler warnings jfs has a few variables being set but never used. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2011-06-20 18:00:27 +00:00
s64 xaddr;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct metapage *mp;
struct super_block *sb;
struct tblock *tblk;
struct inode *iplist[2];
jfs_info("jfs_symlink: dip:0x%p name:%s", dip, name);
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dip);
if (rc)
goto out1;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ssize = strlen(name) + 1;
/*
* search parent directory for entry/freespace
* (dtSearch() returns parent directory page pinned)
*/
if ((rc = get_UCSname(&dname, dentry)))
goto out1;
/*
* allocate on-disk/in-memory inode for symbolic link:
* (iAlloc() returns new, locked inode)
*/
ip = ialloc(dip, S_IFLNK | 0777);
[PATCH] JFS: return correct error when i-node allocation failed I have seen confusing behavior on JFS when I injected many intentional slab allocation errors. The cp command failed with no disk space error with enough disk space. This patch makes: - change the return value in case slab allocation failures happen from -ENOSPC to -ENOMEM - ialloc() return error code so that the caller can know the reason of failures Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)
2006-09-14 14:22:38 +00:00
if (IS_ERR(ip)) {
rc = PTR_ERR(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out2;
}
tid = txBegin(dip->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
fs/vfs/security: pass last path component to LSM on inode creation SELinux would like to implement a new labeling behavior of newly created inodes. We currently label new inodes based on the parent and the creating process. This new behavior would also take into account the name of the new object when deciding the new label. This is not the (supposed) full path, just the last component of the path. This is very useful because creating /etc/shadow is different than creating /etc/passwd but the kernel hooks are unable to differentiate these operations. We currently require that userspace realize it is doing some difficult operation like that and than userspace jumps through SELinux hoops to get things set up correctly. This patch does not implement new behavior, that is obviously contained in a seperate SELinux patch, but it does pass the needed name down to the correct LSM hook. If no such name exists it is fine to pass NULL. Signed-off-by: Eric Paris <eparis@redhat.com>
2011-02-01 16:05:39 +00:00
rc = jfs_init_security(tid, ip, dip, &dentry->d_name);
JFS: Implement jfs_init_security This atomically initializes the security xattr when an object is created Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:05:39 +00:00
if (rc)
goto out3;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_CREATE;
tblk->ino = ip->i_ino;
tblk->u.ixpxd = JFS_IP(ip)->ixpxd;
/* fix symlink access permission
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* (dir_create() ANDs in the u.u_cmask,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* but symlinks really need to be 777 access)
*/
ip->i_mode |= 0777;
/*
* write symbolic link target path name
*/
xtInitRoot(tid, ip);
/*
* write source path name inline in on-disk inode (fast symbolic link)
*/
if (ssize <= IDATASIZE) {
jfs: add jfs specific ->setattr call generic setattr not longer responsible for quota transfer. use jfs_setattr for all jfs's inodes. Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org> Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2010-04-16 13:05:50 +00:00
ip->i_op = &jfs_fast_symlink_inode_operations;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Fix fortify moan in symlink JFS has in jfs_incore.h: /* _inline may overflow into _inline_ea when needed */ /* _inline_ea may overlay the last part of * file._xtroot if maxentry = XTROOTINITSLOT */ union { struct { /* 128: inline symlink */ unchar _inline[128]; /* 128: inline extended attr */ unchar _inline_ea[128]; }; unchar _inline_all[256]; and currently the symlink code copies into _inline; if this is larger than 128 bytes it triggers a fortify warning of the form: memcpy: detected field-spanning write (size 132) of single field "ip->i_link" at fs/jfs/namei.c:950 (size 18446744073709551615) when it's actually OK. Copy it into _inline_all instead. Reported-by: syzbot+5fc38b2ddbbca7f5c680@syzkaller.appspotmail.com Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org> Reviewed-by: Kees Cook <keescook@chromium.org> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2022-10-22 20:39:14 +00:00
ip->i_link = JFS_IP(ip)->i_inline_all;
jfs: switch to simple_follow_link() Reviewed-by: Jan Kara <jack@suse.cz> Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-05-02 14:41:20 +00:00
memcpy(ip->i_link, name, ssize);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ip->i_size = ssize - 1;
/*
* if symlink is > 128 bytes, we don't have the space to
* store inline extended attributes
*/
if (ssize > sizeof (JFS_IP(ip)->i_inline))
JFS_IP(ip)->mode2 &= ~INLINEEA;
jfs: Use unsigned variable for length calculations To avoid confusing the compiler about possible negative sizes, switch "ssize" which can never be negative from int to u32. Seen with GCC 13: ../fs/jfs/namei.c: In function 'jfs_symlink': ../include/linux/fortify-string.h:57:33: warning: '__builtin_memcpy' pointer overflow between offset 0 and size [-2147483648, -1] [-Warray-bounds=] 57 | #define __underlying_memcpy __builtin_memcpy | ^ ... ../fs/jfs/namei.c:950:17: note: in expansion of macro 'memcpy' 950 | memcpy(ip->i_link, name, ssize); | ^~~~~~ Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Chinner <dchinner@redhat.com> Cc: jfs-discussion@lists.sourceforge.net Signed-off-by: Kees Cook <keescook@chromium.org> Acked-by: Jeff Xu <jeffxu@chromium.org> Message-Id: <20230204183355.never.877-kees@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-02-04 18:33:56 +00:00
jfs_info("jfs_symlink: fast symlink added ssize:%u name:%s ",
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ssize, name);
}
/*
* write source path name in a single extent
*/
else {
jfs_info("jfs_symlink: allocate extent ip:0x%p", ip);
jfs: add jfs specific ->setattr call generic setattr not longer responsible for quota transfer. use jfs_setattr for all jfs's inodes. Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org> Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2010-04-16 13:05:50 +00:00
ip->i_op = &jfs_symlink_inode_operations;
don't put symlink bodies in pagecache into highmem kmap() in page_follow_link_light() needed to go - allowing to hold an arbitrary number of kmaps for long is a great way to deadlocking the system. new helper (inode_nohighmem(inode)) needs to be used for pagecache symlinks inodes; done for all in-tree cases. page_follow_link_light() instrumented to yell about anything missed. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-11-17 06:07:57 +00:00
inode_nohighmem(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ip->i_mapping->a_ops = &jfs_aops;
/*
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
* even though the data of symlink object (source
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* path name) is treated as non-journaled user data,
* it is read/written thru buffer cache for performance.
*/
sb = ip->i_sb;
bmask = JFS_SBI(sb)->bsize - 1;
xsize = (ssize + bmask) & ~bmask;
xaddr = 0;
xlen = xsize >> JFS_SBI(sb)->l2bsize;
if ((rc = xtInsert(tid, ip, 0, 0, xlen, &xaddr, 0))) {
txAbort(tid, 0);
goto out3;
}
ip->i_size = ssize - 1;
while (ssize) {
/* This is kind of silly since PATH_MAX == 4K */
jfs: Use unsigned variable for length calculations To avoid confusing the compiler about possible negative sizes, switch "ssize" which can never be negative from int to u32. Seen with GCC 13: ../fs/jfs/namei.c: In function 'jfs_symlink': ../include/linux/fortify-string.h:57:33: warning: '__builtin_memcpy' pointer overflow between offset 0 and size [-2147483648, -1] [-Warray-bounds=] 57 | #define __underlying_memcpy __builtin_memcpy | ^ ... ../fs/jfs/namei.c:950:17: note: in expansion of macro 'memcpy' 950 | memcpy(ip->i_link, name, ssize); | ^~~~~~ Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Christian Brauner <brauner@kernel.org> Cc: Dave Chinner <dchinner@redhat.com> Cc: jfs-discussion@lists.sourceforge.net Signed-off-by: Kees Cook <keescook@chromium.org> Acked-by: Jeff Xu <jeffxu@chromium.org> Message-Id: <20230204183355.never.877-kees@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-02-04 18:33:56 +00:00
u32 copy_size = min_t(u32, ssize, PSIZE);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mp = get_metapage(ip, xaddr, PSIZE, 1);
if (mp == NULL) {
xtTruncate(tid, ip, 0, COMMIT_PWMAP);
rc = -EIO;
txAbort(tid, 0);
goto out3;
}
memcpy(mp->data, name, copy_size);
flush_metapage(mp);
ssize -= copy_size;
name += copy_size;
xaddr += JFS_SBI(sb)->nbperpage;
}
}
/*
* create entry for symbolic link in parent directory
*/
rc = dtSearch(dip, &dname, &ino, &btstack, JFS_CREATE);
if (rc == 0) {
ino = ip->i_ino;
rc = dtInsert(tid, dip, &dname, &ino, &btstack);
}
if (rc) {
if (xlen)
xtTruncate(tid, ip, 0, COMMIT_PWMAP);
txAbort(tid, 0);
/* discard new inode */
goto out3;
}
mark_inode_dirty(ip);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dip, inode_set_ctime_current(dip));
JFS: set i_ctime & i_mtime on target directory when creating links jfs has never been setting i_ctime or i_mtime when creating either hard or symbolic links. I'm surprised nobody had noticed until now. Thanks to Chris Spiegel for reporting the problem. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-10-31 22:53:04 +00:00
mark_inode_dirty(dip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* commit update of parent directory and link object
*/
iplist[0] = dip;
iplist[1] = ip;
rc = txCommit(tid, 2, &iplist[0], 0);
out3:
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(dip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (rc) {
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
free_ea_wmap(ip);
filesystems: add missing nlink wrappers Replace direct i_nlink updates with the respective updater function (inc_nlink, drop_nlink, clear_nlink, inode_dec_link_count). Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-10-28 12:13:28 +00:00
clear_nlink(ip);
jfs: switch to discard_new_inode() we don't want open-by-handle to pick an in-core inode that has failed setup halfway through. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-06-29 15:59:37 +00:00
discard_new_inode(ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
} else {
do d_instantiate/unlock_new_inode combinations safely For anything NFS-exported we do _not_ want to unlock new inode before it has grown an alias; original set of fixes got the ordering right, but missed the nasty complication in case of lockdep being enabled - unlock_new_inode() does lockdep_annotate_inode_mutex_key(inode) which can only be done before anyone gets a chance to touch ->i_mutex. Unfortunately, flipping the order and doing unlock_new_inode() before d_instantiate() opens a window when mkdir can race with open-by-fhandle on a guessed fhandle, leading to multiple aliases for a directory inode and all the breakage that follows from that. Correct solution: a new primitive (d_instantiate_new()) combining these two in the right order - lockdep annotate, then d_instantiate(), then the rest of unlock_new_inode(). All combinations of d_instantiate() with unlock_new_inode() should be converted to that. Cc: stable@kernel.org # 2.6.29 and later Tested-by: Mike Marshall <hubcap@omnibond.com> Reviewed-by: Andreas Dilger <adilger@dilger.ca> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-05-04 12:23:01 +00:00
d_instantiate_new(dentry, ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out2:
free_UCSname(&dname);
out1:
jfs_info("jfs_symlink: rc:%d", rc);
return rc;
}
/*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* NAME: jfs_rename
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* FUNCTION: rename a file or directory
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
fs: port ->rename() to pass mnt_idmap Convert to struct mnt_idmap. Last cycle we merged the necessary infrastructure in 256c8aed2b42 ("fs: introduce dedicated idmap type for mounts"). This is just the conversion to struct mnt_idmap. Currently we still pass around the plain namespace that was attached to a mount. This is in general pretty convenient but it makes it easy to conflate namespaces that are relevant on the filesystem with namespaces that are relevent on the mount level. Especially for non-vfs developers without detailed knowledge in this area this can be a potential source for bugs. Once the conversion to struct mnt_idmap is done all helpers down to the really low-level helpers will take a struct mnt_idmap argument instead of two namespace arguments. This way it becomes impossible to conflate the two eliminating the possibility of any bugs. All of the vfs and all filesystems only operate on struct mnt_idmap. Acked-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2023-01-13 11:49:17 +00:00
static int jfs_rename(struct mnt_idmap *idmap, struct inode *old_dir,
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
struct dentry *old_dentry, struct inode *new_dir,
struct dentry *new_dentry, unsigned int flags)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct btstack btstack;
ino_t ino;
struct component_name new_dname;
struct inode *new_ip;
struct component_name old_dname;
struct inode *old_ip;
int rc;
tid_t tid;
struct tlock *tlck;
struct dt_lock *dtlck;
struct lv *lv;
int ipcount;
struct inode *iplist[4];
struct tblock *tblk;
s64 new_size = 0;
int commit_flag;
fs: support RENAME_NOREPLACE for local filesystems This is trivial to do: - add flags argument to foo_rename() - check if flags doesn't have any other than RENAME_NOREPLACE - assign foo_rename() to .rename2 instead of .rename Filesystems converted: affs, bfs, exofs, ext2, hfs, hfsplus, jffs2, jfs, logfs, minix, msdos, nilfs2, omfs, reiserfs, sysvfs, ubifs, udf, ufs, vfat. Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Acked-by: Boaz Harrosh <ooo@electrozaur.com> Acked-by: Richard Weinberger <richard@nod.at> Acked-by: Bob Copeland <me@bobcopeland.com> Acked-by: Jan Kara <jack@suse.cz> Cc: Theodore Ts'o <tytso@mit.edu> Cc: Jaegeuk Kim <jaegeuk@kernel.org> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Mikulas Patocka <mpatocka@redhat.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Christoph Hellwig <hch@infradead.org>
2016-09-27 09:03:57 +00:00
if (flags & ~RENAME_NOREPLACE)
return -EINVAL;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_rename: %pd %pd", old_dentry, new_dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(old_dir);
if (rc)
goto out1;
rc = dquot_initialize(new_dir);
if (rc)
goto out1;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
old_ip = d_inode(old_dentry);
new_ip = d_inode(new_dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = get_UCSname(&old_dname, old_dentry)))
goto out1;
if ((rc = get_UCSname(&new_dname, new_dentry)))
goto out2;
/*
* Make sure source inode number is what we think it is
*/
rc = dtSearch(old_dir, &old_dname, &ino, &btstack, JFS_LOOKUP);
if (rc || (ino != old_ip->i_ino)) {
rc = -ENOENT;
goto out3;
}
/*
* Make sure dest inode number (if any) is what we think it is
*/
rc = dtSearch(new_dir, &new_dname, &ino, &btstack, JFS_LOOKUP);
JFS: Remove defconfig ptr comparison to 0 Remove sparse warning: Using plain integer as NULL pointer Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-11-14 04:16:08 +00:00
if (!rc) {
JFS: FIx one more plain integer as NULL pointer warning Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-11-14 04:25:41 +00:00
if ((!new_ip) || (ino != new_ip->i_ino)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rc = -ESTALE;
goto out3;
}
} else if (rc != -ENOENT)
goto out3;
else if (new_ip) {
/* no entry exists, but one was expected */
rc = -ESTALE;
goto out3;
}
if (S_ISDIR(old_ip->i_mode)) {
if (new_ip) {
if (!dtEmpty(new_ip)) {
rc = -ENOTEMPTY;
goto out3;
}
}
} else if (new_ip) {
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
IWRITE_LOCK(new_ip, RDWRLOCK_NORMAL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* Init inode for quota operations. */
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(new_ip);
if (rc)
goto out_unlock;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/*
* The real work starts here
*/
tid = txBegin(new_dir->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
/*
* How do we know the locking is safe from deadlocks?
* The vfs does the hard part for us. Any time we are taking nested
* commit_mutexes, the vfs already has i_mutex held on the parent.
* Here, the vfs has already taken i_mutex on both old_dir and new_dir.
*/
mutex_lock_nested(&JFS_IP(new_dir)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(old_ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (old_dir != new_dir)
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(old_dir)->commit_mutex,
COMMIT_MUTEX_SECOND_PARENT);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (new_ip) {
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(new_ip)->commit_mutex,
COMMIT_MUTEX_VICTIM);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* Change existing directory entry to new inode number
*/
ino = new_ip->i_ino;
rc = dtModify(tid, new_dir, &new_dname, &ino,
old_ip->i_ino, JFS_RENAME);
if (rc)
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
goto out_tx;
[PATCH] r/o bind mounts: unlink: monitor i_nlink When a filesystem decrements i_nlink to zero, it means that a write must be performed in order to drop the inode from the filesystem. We're shortly going to have keep filesystems from being remounted r/o between the time that this i_nlink decrement and that write occurs. So, add a little helper function to do the decrements. We'll tie into it in a bit to note when i_nlink hits zero. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:03 +00:00
drop_nlink(new_ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (S_ISDIR(new_ip->i_mode)) {
[PATCH] r/o bind mounts: unlink: monitor i_nlink When a filesystem decrements i_nlink to zero, it means that a write must be performed in order to drop the inode from the filesystem. We're shortly going to have keep filesystems from being remounted r/o between the time that this i_nlink decrement and that write occurs. So, add a little helper function to do the decrements. We'll tie into it in a bit to note when i_nlink hits zero. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:03 +00:00
drop_nlink(new_ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (new_ip->i_nlink) {
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(new_ip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (old_dir != new_dir)
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(old_dir)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
mutex_unlock(&JFS_IP(old_ip)->commit_mutex);
mutex_unlock(&JFS_IP(new_dir)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!S_ISDIR(old_ip->i_mode) && new_ip)
IWRITE_UNLOCK(new_ip);
jfs_error(new_ip->i_sb,
jfs: Update jfs_error Use a more current logging style. Add __printf format and argument verification. Remove embedded function names from formats. Add %pf, __builtin_return_address(0) to jfs_error. Add newlines to formats for kernel style consistency. (One format already had an erroneous newline) Coalesce formats and align arguments. Object size reduced ~1KiB. $ size fs/jfs/built-in.o* text data bss dec hex filename 201891 35488 63936 301315 49903 fs/jfs/built-in.o.new 202821 35488 64192 302501 49da5 fs/jfs/built-in.o.old Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2013-06-04 23:39:15 +00:00
"new_ip->i_nlink != 0\n");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return -EIO;
}
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_DELETE;
tblk->u.ip = new_ip;
} else if (new_ip->i_nlink == 0) {
assert(!test_cflag(COMMIT_Nolink, new_ip));
/* free block resources */
if ((new_size = commitZeroLink(tid, new_ip)) < 0) {
txAbort(tid, 1); /* Marks FS Dirty */
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
rc = new_size;
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
goto out_tx;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_DELETE;
tblk->u.ip = new_ip;
} else {
jfs: convert to ctime accessor functions In later patches, we're going to change how the inode's ctime field is used. Switch to using accessor functions instead of raw accesses of inode->i_ctime. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Jan Kara <jack@suse.cz> Message-Id: <20230705190309.579783-53-jlayton@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-07-05 19:01:20 +00:00
inode_set_ctime_current(new_ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(new_ip);
}
} else {
/*
* Add new directory entry
*/
rc = dtSearch(new_dir, &new_dname, &ino, &btstack,
JFS_CREATE);
if (rc) {
jfs: Coalesce some formats Formats are better kept as a single line for easier grep. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2016-03-30 12:23:18 +00:00
jfs_err("jfs_rename didn't expect dtSearch to fail w/rc = %d",
rc);
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
goto out_tx;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
ino = old_ip->i_ino;
rc = dtInsert(tid, new_dir, &new_dname, &ino, &btstack);
if (rc) {
if (rc == -EIO)
jfs_err("jfs_rename: dtInsert returned -EIO");
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
goto out_tx;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
if (S_ISDIR(old_ip->i_mode))
[PATCH] r/o bind mount prepwork: inc_nlink() helper This is mostly included for parity with dec_nlink(), where we will have some more hooks. This one should stay pretty darn straightforward for now. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:04 +00:00
inc_nlink(new_dir);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/*
* Remove old directory entry
*/
ino = old_ip->i_ino;
rc = dtDelete(tid, old_dir, &old_dname, &ino, JFS_REMOVE);
if (rc) {
jfs_err("jfs_rename did not expect dtDelete to return rc = %d",
rc);
txAbort(tid, 1); /* Marks Filesystem dirty */
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
goto out_tx;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
if (S_ISDIR(old_ip->i_mode)) {
[PATCH] r/o bind mounts: unlink: monitor i_nlink When a filesystem decrements i_nlink to zero, it means that a write must be performed in order to drop the inode from the filesystem. We're shortly going to have keep filesystems from being remounted r/o between the time that this i_nlink decrement and that write occurs. So, add a little helper function to do the decrements. We'll tie into it in a bit to note when i_nlink hits zero. Signed-off-by: Dave Hansen <haveblue@us.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-01 06:29:03 +00:00
drop_nlink(old_dir);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (old_dir != new_dir) {
/*
* Change inode number of parent for moved directory
*/
JFS_IP(old_ip)->i_dtroot.header.idotdot =
cpu_to_le32(new_dir->i_ino);
/* Linelock header of dtree */
tlck = txLock(tid, old_ip,
(struct metapage *) &JFS_IP(old_ip)->bxflag,
tlckDTREE | tlckBTROOT | tlckRELINK);
dtlck = (struct dt_lock *) & tlck->lock;
ASSERT(dtlck->index == 0);
lv = & dtlck->lv[0];
lv->offset = 0;
lv->length = 1;
dtlck->index++;
}
}
/*
* Update ctime on changed/moved inodes & mark dirty
*/
jfs: convert to ctime accessor functions In later patches, we're going to change how the inode's ctime field is used. Switch to using accessor functions instead of raw accesses of inode->i_ctime. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Jan Kara <jack@suse.cz> Message-Id: <20230705190309.579783-53-jlayton@kernel.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-07-05 19:01:20 +00:00
inode_set_ctime_current(old_ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(old_ip);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(new_dir, inode_set_ctime_current(new_dir));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(new_dir);
/* Build list of inodes modified by this transaction */
ipcount = 0;
iplist[ipcount++] = old_ip;
if (new_ip)
iplist[ipcount++] = new_ip;
iplist[ipcount++] = old_dir;
if (old_dir != new_dir) {
iplist[ipcount++] = new_dir;
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(old_dir,
inode_set_ctime_current(old_dir));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(old_dir);
}
/*
* Incomplete truncate of file data can
* result in timing problems unless we synchronously commit the
* transaction.
*/
if (new_size)
commit_flag = COMMIT_SYNC;
else
commit_flag = 0;
rc = txCommit(tid, ipcount, iplist, commit_flag);
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
out_tx:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
txEnd(tid);
if (new_ip)
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(new_ip)->commit_mutex);
JFS: commit_mutex cleanups I look at code, and see that 1)locks wasn't release in the opposite order in which they were taken 2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it 3)I see strange expression: "! !" May be this worth to fix? Signed-off-by: Evgeniy Dushistov <dushistov@mail.ru> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-06-05 13:21:03 +00:00
if (old_dir != new_dir)
mutex_unlock(&JFS_IP(old_dir)->commit_mutex);
mutex_unlock(&JFS_IP(old_ip)->commit_mutex);
mutex_unlock(&JFS_IP(new_dir)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
while (new_size && (rc == 0)) {
tid = txBegin(new_ip->i_sb, 0);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_lock(&JFS_IP(new_ip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
new_size = xtTruncate_pmap(tid, new_ip, new_size);
if (new_size < 0) {
txAbort(tid, 1);
JFS: White space cleanup Removed trailing spaces & tabs, and spaces preceding tabs. Also a couple very minor comment cleanups. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)
2006-10-02 14:55:27 +00:00
rc = new_size;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
} else
rc = txCommit(tid, 1, &new_ip, COMMIT_SYNC);
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(new_ip)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
if (new_ip && (new_ip->i_nlink == 0))
set_cflag(COMMIT_Nolink, new_ip);
/*
* Truncating the directory index table is not guaranteed. It
* may need to be done iteratively
*/
if (test_cflag(COMMIT_Stale, old_dir)) {
if (old_dir->i_size > 1)
jfs_truncate_nolock(old_dir, 0);
clear_cflag(COMMIT_Stale, old_dir);
}
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
out_unlock:
jfs: clean up jfs_rename and fix out of order unlock The end of jfs_rename(), which is also used by the error paths, included a call to IWRITE_UNLOCK(new_ip) after labels out1, out2 and out3. If we come in through these labels, IWRITE_LOCK() has not been called yet. In moving that call to the correct spot, I also moved some exceptional truncate code earlier as well, since the early error paths don't need to deal with it, and I renamed out4: to out_tx: so a future patch by Jan Kara doesn't need to deal with renumbering or confusing out-of-order labels. Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 17:52:47 +00:00
if (new_ip && !S_ISDIR(new_ip->i_mode))
IWRITE_UNLOCK(new_ip);
out3:
free_UCSname(&new_dname);
out2:
free_UCSname(&old_dname);
out1:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs_info("jfs_rename: returning %d", rc);
return rc;
}
/*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* NAME: jfs_mknod
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
JFS: Whitespace cleanup and remove some dead code Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2007-06-06 20:28:35 +00:00
* FUNCTION: Create a special file (device)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
fs: port ->mknod() to pass mnt_idmap Convert to struct mnt_idmap. Last cycle we merged the necessary infrastructure in 256c8aed2b42 ("fs: introduce dedicated idmap type for mounts"). This is just the conversion to struct mnt_idmap. Currently we still pass around the plain namespace that was attached to a mount. This is in general pretty convenient but it makes it easy to conflate namespaces that are relevant on the filesystem with namespaces that are relevent on the mount level. Especially for non-vfs developers without detailed knowledge in this area this can be a potential source for bugs. Once the conversion to struct mnt_idmap is done all helpers down to the really low-level helpers will take a struct mnt_idmap argument instead of two namespace arguments. This way it becomes impossible to conflate the two eliminating the possibility of any bugs. All of the vfs and all filesystems only operate on struct mnt_idmap. Acked-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2023-01-13 11:49:16 +00:00
static int jfs_mknod(struct mnt_idmap *idmap, struct inode *dir,
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
struct dentry *dentry, umode_t mode, dev_t rdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct jfs_inode_info *jfs_ip;
struct btstack btstack;
struct component_name dname;
ino_t ino;
struct inode *ip;
struct inode *iplist[2];
int rc;
tid_t tid;
struct tblock *tblk;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_mknod: %pd", dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
jfs: Handle error from dquot_initialize() dquot_initialize() can now return error. Handle it where possible Slightly modified by Dave Kleikamp due to needed jfs_rename() error path fix. Signed-off-by: Jan Kara <jack@suse.com> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com>
2015-07-15 18:53:19 +00:00
rc = dquot_initialize(dir);
if (rc)
goto out;
dquot: move dquot initialization responsibility into the filesystem Currently various places in the VFS call vfs_dq_init directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the initialization. For most metadata operations this is a straight forward move into the methods, but for truncate and open it's a bit more complicated. For truncate we currently only call vfs_dq_init for the sys_truncate case because open already takes care of it for ftruncate and open(O_TRUNC) - the new code causes an additional vfs_dq_init for those which is harmless. For open the initialization is moved from do_filp_open into the open method, which means it happens slightly earlier now, and only for regular files. The latter is fine because we don't need to initialize it for operations on special files, and we already do it as part of the namespace operations for directories. Add a dquot_file_open helper that filesystems that support generic quotas can use to fill in ->open. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:06 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if ((rc = get_UCSname(&dname, dentry)))
goto out;
ip = ialloc(dir, mode);
[PATCH] JFS: return correct error when i-node allocation failed I have seen confusing behavior on JFS when I injected many intentional slab allocation errors. The cp command failed with no disk space error with enough disk space. This patch makes: - change the return value in case slab allocation failures happen from -ENOSPC to -ENOMEM - ialloc() return error code so that the caller can know the reason of failures Signed-off-by: Akinobu Mita <mita@miraclelinux.com> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com> (cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)
2006-09-14 14:22:38 +00:00
if (IS_ERR(ip)) {
rc = PTR_ERR(ip);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out1;
}
jfs_ip = JFS_IP(ip);
tid = txBegin(dir->i_sb, 0);
JFS: Add lockdep annotations Yeah, it's about time. Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2007-01-09 20:14:48 +00:00
mutex_lock_nested(&JFS_IP(dir)->commit_mutex, COMMIT_MUTEX_PARENT);
mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
rc = jfs_init_acl(tid, ip, dir);
if (rc)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out3;
fs/vfs/security: pass last path component to LSM on inode creation SELinux would like to implement a new labeling behavior of newly created inodes. We currently label new inodes based on the parent and the creating process. This new behavior would also take into account the name of the new object when deciding the new label. This is not the (supposed) full path, just the last component of the path. This is very useful because creating /etc/shadow is different than creating /etc/passwd but the kernel hooks are unable to differentiate these operations. We currently require that userspace realize it is doing some difficult operation like that and than userspace jumps through SELinux hoops to get things set up correctly. This patch does not implement new behavior, that is obviously contained in a seperate SELinux patch, but it does pass the needed name down to the correct LSM hook. If no such name exists it is fine to pass NULL. Signed-off-by: Eric Paris <eparis@redhat.com>
2011-02-01 16:05:39 +00:00
rc = jfs_init_security(tid, ip, dir, &dentry->d_name);
JFS: Implement jfs_init_security This atomically initializes the security xattr when an object is created Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:05:39 +00:00
if (rc) {
txAbort(tid, 0);
goto out3;
}
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
if ((rc = dtSearch(dir, &dname, &ino, &btstack, JFS_CREATE))) {
txAbort(tid, 0);
goto out3;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
tblk = tid_to_tblock(tid);
tblk->xflag |= COMMIT_CREATE;
tblk->ino = ip->i_ino;
tblk->u.ixpxd = JFS_IP(ip)->ixpxd;
ino = ip->i_ino;
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
if ((rc = dtInsert(tid, dir, &dname, &ino, &btstack))) {
txAbort(tid, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out3;
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
ip->i_op = &jfs_file_inode_operations;
jfs_ip->dev = new_encode_dev(rdev);
init_special_inode(ip, ip->i_mode, rdev);
mark_inode_dirty(ip);
jfs: convert to new timestamp accessors Convert to using the new inode timestamp accessor functions. Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://lore.kernel.org/r/20231004185347.80880-46-jlayton@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-04 18:52:33 +00:00
inode_set_mtime_to_ts(dir, inode_set_ctime_current(dir));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
mark_inode_dirty(dir);
iplist[0] = dir;
iplist[1] = ip;
rc = txCommit(tid, 2, iplist, 0);
out3:
txEnd(tid);
JFS: semaphore to mutex conversion. the conversion was generated via scripts, and the result was validated automatically via a script as well. build and boot tested. Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2006-01-24 21:22:50 +00:00
mutex_unlock(&JFS_IP(ip)->commit_mutex);
mutex_unlock(&JFS_IP(dir)->commit_mutex);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (rc) {
JFS: allow extended attributes to be set within a existing transaction Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
2005-09-01 14:02:43 +00:00
free_ea_wmap(ip);
filesystems: add missing nlink wrappers Replace direct i_nlink updates with the respective updater function (inc_nlink, drop_nlink, clear_nlink, inode_dec_link_count). Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
2011-10-28 12:13:28 +00:00
clear_nlink(ip);
jfs: switch to discard_new_inode() we don't want open-by-handle to pick an in-core inode that has failed setup halfway through. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-06-29 15:59:37 +00:00
discard_new_inode(ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
} else {
do d_instantiate/unlock_new_inode combinations safely For anything NFS-exported we do _not_ want to unlock new inode before it has grown an alias; original set of fixes got the ordering right, but missed the nasty complication in case of lockdep being enabled - unlock_new_inode() does lockdep_annotate_inode_mutex_key(inode) which can only be done before anyone gets a chance to touch ->i_mutex. Unfortunately, flipping the order and doing unlock_new_inode() before d_instantiate() opens a window when mkdir can race with open-by-fhandle on a guessed fhandle, leading to multiple aliases for a directory inode and all the breakage that follows from that. Correct solution: a new primitive (d_instantiate_new()) combining these two in the right order - lockdep annotate, then d_instantiate(), then the rest of unlock_new_inode(). All combinations of d_instantiate() with unlock_new_inode() should be converted to that. Cc: stable@kernel.org # 2.6.29 and later Tested-by: Mike Marshall <hubcap@omnibond.com> Reviewed-by: Andreas Dilger <adilger@dilger.ca> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2018-05-04 12:23:01 +00:00
d_instantiate_new(dentry, ip);
nfsd race fixes: jfs jfs version of Al Viro's nfsd race patches Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-12-31 04:08:37 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out1:
free_UCSname(&dname);
out:
jfs_info("jfs_mknod: returning %d", rc);
return rc;
}
stop passing nameidata to ->lookup() Just the flags; only NFS cares even about that, but there are legitimate uses for such argument. And getting rid of that completely would require splitting ->lookup() into a couple of methods (at least), so let's leave that alone for now... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 21:13:09 +00:00
static struct dentry *jfs_lookup(struct inode *dip, struct dentry *dentry, unsigned int flags)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct btstack btstack;
ino_t inum;
struct inode *ip;
struct component_name key;
int rc;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
jfs_info("jfs_lookup: name = %pd", dentry);
jfs_lookup(): don't bother with . or .. they'll never be passed to ->lookup() Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-17 14:17:46 +00:00
if ((rc = get_UCSname(&key, dentry)))
return ERR_PTR(rc);
rc = dtSearch(dip, &key, &inum, &btstack, JFS_LOOKUP);
free_UCSname(&key);
if (rc == -ENOENT) {
ip = NULL;
} else if (rc) {
jfs_err("jfs_lookup: dtSearch returned %d", rc);
ip = ERR_PTR(rc);
} else {
ip = jfs_iget(dip->i_sb, inum);
if (IS_ERR(ip))
jfs_err("jfs_lookup: iget failed on inum %d", (uint)inum);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
switch jfs to ->s_d_op, close exportfs races Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2010-12-18 15:59:31 +00:00
return d_splice_alias(ip, dentry);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
jfs: new export ops Trivial switch over to the new generic helpers. Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Neil Brown <neilb@suse.de> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Dave Kleikamp <shaggy@austin.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-21 23:42:09 +00:00
static struct inode *jfs_nfs_get_inode(struct super_block *sb,
u64 ino, u32 generation)
knfsd: exportfs: remove iget abuse When the exportfs interface was added the expectation was that filesystems provide an operation to convert from a file handle to an inode/dentry, but it kept a backwards compat option that still calls into iget. Calling into iget from non-filesystem code is very bad, because it gives too little information to filesystem, and simply crashes if the filesystem doesn't implement the ->read_inode routine. Fortunately there are only two filesystems left using this fallback: efs and jfs. This patch moves a copy of export_iget to each of those to implement the get_dentry method. While this is a temporary increase of lines of code in the kernel it allows for a much cleaner interface and important code restructuring in later patches. [akpm@linux-foundation.org: add jfs_get_inode_flags() declaration] Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 11:04:29 +00:00
{
struct inode *inode;
if (ino == 0)
return ERR_PTR(-ESTALE);
iget: stop JFS from using iget() and read_inode() Stop the JFS filesystem from using iget() and read_inode(). Replace jfs_read_inode() with jfs_iget(), and call that instead of iget(). jfs_iget() then uses iget_locked() directly and returns a proper error code instead of an inode in the event of an error. jfs_fill_super() returns any error incurred when getting the root inode instead of EINVAL. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-07 08:15:43 +00:00
inode = jfs_iget(sb, ino);
if (IS_ERR(inode))
return ERR_CAST(inode);
knfsd: exportfs: remove iget abuse When the exportfs interface was added the expectation was that filesystems provide an operation to convert from a file handle to an inode/dentry, but it kept a backwards compat option that still calls into iget. Calling into iget from non-filesystem code is very bad, because it gives too little information to filesystem, and simply crashes if the filesystem doesn't implement the ->read_inode routine. Fortunately there are only two filesystems left using this fallback: efs and jfs. This patch moves a copy of export_iget to each of those to implement the get_dentry method. While this is a temporary increase of lines of code in the kernel it allows for a much cleaner interface and important code restructuring in later patches. [akpm@linux-foundation.org: add jfs_get_inode_flags() declaration] Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 11:04:29 +00:00
iget: stop JFS from using iget() and read_inode() Stop the JFS filesystem from using iget() and read_inode(). Replace jfs_read_inode() with jfs_iget(), and call that instead of iget(). jfs_iget() then uses iget_locked() directly and returns a proper error code instead of an inode in the event of an error. jfs_fill_super() returns any error incurred when getting the root inode instead of EINVAL. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Acked-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-07 08:15:43 +00:00
if (generation && inode->i_generation != generation) {
jfs: new export ops Trivial switch over to the new generic helpers. Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Neil Brown <neilb@suse.de> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Dave Kleikamp <shaggy@austin.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-21 23:42:09 +00:00
iput(inode);
return ERR_PTR(-ESTALE);
knfsd: exportfs: remove iget abuse When the exportfs interface was added the expectation was that filesystems provide an operation to convert from a file handle to an inode/dentry, but it kept a backwards compat option that still calls into iget. Calling into iget from non-filesystem code is very bad, because it gives too little information to filesystem, and simply crashes if the filesystem doesn't implement the ->read_inode routine. Fortunately there are only two filesystems left using this fallback: efs and jfs. This patch moves a copy of export_iget to each of those to implement the get_dentry method. While this is a temporary increase of lines of code in the kernel it allows for a much cleaner interface and important code restructuring in later patches. [akpm@linux-foundation.org: add jfs_get_inode_flags() declaration] Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 11:04:29 +00:00
}
jfs: new export ops Trivial switch over to the new generic helpers. Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Neil Brown <neilb@suse.de> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Dave Kleikamp <shaggy@austin.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-21 23:42:09 +00:00
return inode;
}
knfsd: exportfs: remove iget abuse When the exportfs interface was added the expectation was that filesystems provide an operation to convert from a file handle to an inode/dentry, but it kept a backwards compat option that still calls into iget. Calling into iget from non-filesystem code is very bad, because it gives too little information to filesystem, and simply crashes if the filesystem doesn't implement the ->read_inode routine. Fortunately there are only two filesystems left using this fallback: efs and jfs. This patch moves a copy of export_iget to each of those to implement the get_dentry method. While this is a temporary increase of lines of code in the kernel it allows for a much cleaner interface and important code restructuring in later patches. [akpm@linux-foundation.org: add jfs_get_inode_flags() declaration] Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 11:04:29 +00:00
jfs: new export ops Trivial switch over to the new generic helpers. Signed-off-by: Christoph Hellwig <hch@lst.de> Cc: Neil Brown <neilb@suse.de> Cc: "J. Bruce Fields" <bfields@fieldses.org> Cc: Dave Kleikamp <shaggy@austin.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-21 23:42:09 +00:00
struct dentry *jfs_fh_to_dentry(struct super_block *sb, struct fid *fid,
int fh_len, int fh_type)
{
return generic_fh_to_dentry(sb, fid, fh_len, fh_type,
jfs_nfs_get_inode);
}
struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid,
int fh_len, int fh_type)
{
return generic_fh_to_parent(sb, fid, fh_len, fh_type,
jfs_nfs_get_inode);
knfsd: exportfs: remove iget abuse When the exportfs interface was added the expectation was that filesystems provide an operation to convert from a file handle to an inode/dentry, but it kept a backwards compat option that still calls into iget. Calling into iget from non-filesystem code is very bad, because it gives too little information to filesystem, and simply crashes if the filesystem doesn't implement the ->read_inode routine. Fortunately there are only two filesystems left using this fallback: efs and jfs. This patch moves a copy of export_iget to each of those to implement the get_dentry method. While this is a temporary increase of lines of code in the kernel it allows for a much cleaner interface and important code restructuring in later patches. [akpm@linux-foundation.org: add jfs_get_inode_flags() declaration] Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 11:04:29 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct dentry *jfs_get_parent(struct dentry *dentry)
{
unsigned long parent_ino;
parent_ino =
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
le32_to_cpu(JFS_IP(d_inode(dentry))->i_dtroot.header.idotdot);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
don't bother with ->d_inode->i_sb - it's always equal to ->d_sb ... and neither can ever be NULL Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-04-10 05:33:30 +00:00
return d_obtain_alias(jfs_iget(dentry->d_sb, parent_ino));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[PATCH] mark struct inode_operations const 2 Many struct inode_operations in the kernel can be "const". Marking them const moves these to the .rodata section, which avoids false sharing with potential dirty data. In addition it'll catch accidental writes at compile time to these shared resources. Signed-off-by: Arjan van de Ven <arjan@linux.intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-12 08:55:39 +00:00
const struct inode_operations jfs_dir_inode_operations = {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
.create = jfs_create,
.lookup = jfs_lookup,
.link = jfs_link,
.unlink = jfs_unlink,
.symlink = jfs_symlink,
.mkdir = jfs_mkdir,
.rmdir = jfs_rmdir,
.mknod = jfs_mknod,
.rename = jfs_rename,
.listxattr = jfs_listxattr,
.setattr = jfs_setattr,
jfs: convert to fileattr Use the fileattr API to let the VFS handle locking, permission checking and conversion. Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org>
2021-04-07 12:36:44 +00:00
.fileattr_get = jfs_fileattr_get,
.fileattr_set = jfs_fileattr_set,
dquot: move dquot transfer responsibility into the filesystem Currently notify_change calls vfs_dq_transfer directly. This means we tie the quota code into the VFS. Get rid of that and make the filesystem responsible for the transfer. Most filesystems already do this, only ufs and udf need the code added, and for jfs it needs to be enabled unconditionally instead of only when ACLs are enabled. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jan Kara <jack@suse.cz>
2010-03-03 14:05:02 +00:00
#ifdef CONFIG_JFS_POSIX_ACL
fs: rename current get acl method The current way of setting and getting posix acls through the generic xattr interface is error prone and type unsafe. The vfs needs to interpret and fixup posix acls before storing or reporting it to userspace. Various hacks exist to make this work. The code is hard to understand and difficult to maintain in it's current form. Instead of making this work by hacking posix acls through xattr handlers we are building a dedicated posix acl api around the get and set inode operations. This removes a lot of hackiness and makes the codepaths easier to maintain. A lot of background can be found in [1]. The current inode operation for getting posix acls takes an inode argument but various filesystems (e.g., 9p, cifs, overlayfs) need access to the dentry. In contrast to the ->set_acl() inode operation we cannot simply extend ->get_acl() to take a dentry argument. The ->get_acl() inode operation is called from: acl_permission_check() -> check_acl() -> get_acl() which is part of generic_permission() which in turn is part of inode_permission(). Both generic_permission() and inode_permission() are called in the ->permission() handler of various filesystems (e.g., overlayfs). So simply passing a dentry argument to ->get_acl() would amount to also having to pass a dentry argument to ->permission(). We should avoid this unnecessary change. So instead of extending the existing inode operation rename it from ->get_acl() to ->get_inode_acl() and add a ->get_acl() method later that passes a dentry argument and which filesystems that need access to the dentry can implement instead of ->get_inode_acl(). Filesystems like cifs which allow setting and getting posix acls but not using them for permission checking during lookup can simply not implement ->get_inode_acl(). This is intended to be a non-functional change. Link: https://lore.kernel.org/all/20220801145520.1532837-1-brauner@kernel.org [1] Suggested-by/Inspired-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2022-09-22 15:17:00 +00:00
.get_inode_acl = jfs_get_acl,
jfs: use generic posix ACL infrastructure Copy the scheme I introduced to btrfs many years ago to only use the xattr handler for ACLs, but pass plain attrs straight through. Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-12-20 13:16:51 +00:00
.set_acl = jfs_set_acl,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#endif
};
vfs: get rid of old '->iterate' directory operation All users now just use '->iterate_shared()', which only takes the directory inode lock for reading. Filesystems that never got convered to shared mode now instead use a wrapper that drops the lock, re-takes it in write mode, calls the old function, and then downgrades the lock back to read mode. This way the VFS layer and other callers no longer need to care about filesystems that never got converted to the modern era. The filesystems that use the new wrapper are ceph, coda, exfat, jfs, ntfs, ocfs2, overlayfs, and vboxsf. Honestly, several of them look like they really could just iterate their directories in shared mode and skip the wrapper entirely, but the point of this change is to not change semantics or fix filesystems that haven't been fixed in the last 7+ years, but to finally get rid of the dual iterators. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-08-05 19:25:01 +00:00
WRAP_DIR_ITER(jfs_readdir) // FIXME!
[PATCH] Make most file operations structs in fs/ const This is a conversion to make the various file_operations structs in fs/ const. Basically a regexp job, with a few manual fixups The goal is both to increase correctness (harder to accidentally write to shared datastructures) and reducing the false sharing of cachelines with things that get dirty in .data (while .rodata is nicely read only and thus cache clean) Signed-off-by: Arjan van de Ven <arjan@infradead.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-28 09:56:42 +00:00
const struct file_operations jfs_dir_operations = {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
.read = generic_read_dir,
vfs: get rid of old '->iterate' directory operation All users now just use '->iterate_shared()', which only takes the directory inode lock for reading. Filesystems that never got convered to shared mode now instead use a wrapper that drops the lock, re-takes it in write mode, calls the old function, and then downgrades the lock back to read mode. This way the VFS layer and other callers no longer need to care about filesystems that never got converted to the modern era. The filesystems that use the new wrapper are ceph, coda, exfat, jfs, ntfs, ocfs2, overlayfs, and vboxsf. Honestly, several of them look like they really could just iterate their directories in shared mode and skip the wrapper entirely, but the point of this change is to not change semantics or fix filesystems that haven't been fixed in the last 7+ years, but to finally get rid of the dual iterators. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-08-05 19:25:01 +00:00
.iterate_shared = shared_jfs_readdir,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
.fsync = jfs_fsync,
BKL-removal: Use unlocked_ioctl for jfs Convert jfs_ioctl over to not use the BKL. The only potential race I could see was with two ioctls in parallel changing the flags and losing the updates. Use the i_mutex to protect against this. Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
2008-01-27 22:58:51 +00:00
.unlocked_ioctl = jfs_ioctl,
jfs: convert to fileattr Use the fileattr API to let the VFS handle locking, permission checking and conversion. Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org>
2021-04-07 12:36:44 +00:00
.compat_ioctl = compat_ptr_ioctl,
[PATCH] fix ->llseek for more directories With this patch all directory fops instances that have a readdir that doesn't take the BKL are switched to generic_file_llseek. Signed-off-by: Christoph Hellwig <hch@lst.de>
2008-09-03 19:53:01 +00:00
.llseek = generic_file_llseek,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
};
Don't pass inode to ->d_hash() and ->d_compare() Instances either don't look at it at all (the majority of cases) or only want it to find the superblock (which can be had as dentry->d_sb). A few cases that want more are actually safe with dentry->d_inode - the only precaution needed is the check that it hadn't been replaced with NULL by rmdir() or by overwriting rename(), which case should be simply treated as cache miss. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-21 22:22:44 +00:00
static int jfs_ci_hash(const struct dentry *dir, struct qstr *this)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
unsigned long hash;
int i;
vfs: make the string hashes salt the hash We always mixed in the parent pointer into the dentry name hash, but we did it late at lookup time. It turns out that we can simplify that lookup-time action by salting the hash with the parent pointer early instead of late. A few other users of our string hashes also wanted to mix in their own pointers into the hash, and those are updated to use the same mechanism. Hash users that don't have any particular initial salt can just use the NULL pointer as a no-salt. Cc: Vegard Nossum <vegard.nossum@oracle.com> Cc: George Spelvin <linux@sciencehorizons.net> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-06-10 14:51:30 +00:00
hash = init_name_hash(dir);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
for (i=0; i < this->len; i++)
hash = partial_name_hash(tolower(this->name[i]), hash);
this->hash = end_name_hash(hash);
return 0;
}
get rid of 'parent' argument of ->d_compare() Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-07-31 20:37:25 +00:00
static int jfs_ci_compare(const struct dentry *dentry,
fs: change d_compare for rcu-walk Change d_compare so it may be called from lock-free RCU lookups. This does put significant restrictions on what may be done from the callback, however there don't seem to have been any problems with in-tree fses. If some strange use case pops up that _really_ cannot cope with the rcu-walk rules, we can just add new rcu-unaware callbacks, which would cause name lookup to drop out of rcu-walk mode. For in-tree filesystems, this is just a mechanical change. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:27 +00:00
unsigned int len, const char *str, const struct qstr *name)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
int i, result = 1;
fs: change d_compare for rcu-walk Change d_compare so it may be called from lock-free RCU lookups. This does put significant restrictions on what may be done from the callback, however there don't seem to have been any problems with in-tree fses. If some strange use case pops up that _really_ cannot cope with the rcu-walk rules, we can just add new rcu-unaware callbacks, which would cause name lookup to drop out of rcu-walk mode. For in-tree filesystems, this is just a mechanical change. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:27 +00:00
if (len != name->len)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out;
fs: change d_compare for rcu-walk Change d_compare so it may be called from lock-free RCU lookups. This does put significant restrictions on what may be done from the callback, however there don't seem to have been any problems with in-tree fses. If some strange use case pops up that _really_ cannot cope with the rcu-walk rules, we can just add new rcu-unaware callbacks, which would cause name lookup to drop out of rcu-walk mode. For in-tree filesystems, this is just a mechanical change. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:27 +00:00
for (i=0; i < len; i++) {
if (tolower(str[i]) != tolower(name->name[i]))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out;
}
result = 0;
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
out:
return result;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
stop passing nameidata * to ->d_revalidate() Just the lookup flags. Die, bastard, die... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 20:03:43 +00:00
static int jfs_ci_revalidate(struct dentry *dentry, unsigned int flags)
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
{
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
* This is not negative dentry. Always valid.
*
* Note, rename() to existing directory entry will have ->d_inode,
* and will use existing name which isn't specified name by user.
*
* We may be able to drop this positive dentry here. But dropping
* positive dentry isn't good idea. So it's unsupported like
* rename("filename", "FILENAME") for now.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
if (d_really_is_positive(dentry))
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
return 1;
/*
* This may be nfsd (or something), anyway, we can't see the
* intent of this. So, since this can be for creation, drop it.
*/
stop passing nameidata * to ->d_revalidate() Just the lookup flags. Die, bastard, die... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 20:03:43 +00:00
if (!flags)
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
return 0;
/*
* Drop the negative dentry, in order to make sure to use the
* case sensitive name which is specified by user if this is
* for creation.
*/
stop passing nameidata * to ->d_revalidate() Just the lookup flags. Die, bastard, die... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 20:03:43 +00:00
if (flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
LOOKUP_CREATE and LOOKUP_RENAME_TARGET can be set only on the last step Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-06-26 01:37:18 +00:00
return 0;
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
return 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
constify dentry_operations: JFS Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2009-02-20 06:00:49 +00:00
const struct dentry_operations jfs_ci_dentry_operations =
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
.d_hash = jfs_ci_hash,
.d_compare = jfs_ci_compare,
jfs: dont overwrite dentry name in d_revalidate Use vfat's method for dealing with negative dentries to preserve case, rather than overwrite dentry name in d_revalidate, which is a bit ugly and also gets in the way of doing lock-free path walking. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:25 +00:00
.d_revalidate = jfs_ci_revalidate,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
};
Reference in New Issue Copy Permalink