- 08 Dec, 2013 2 commits
-
-
Bjorn Helgaas authored
There's no "unlink from sysfs" interface for ksets, so I think callers of kset_unregister() expect the kset to be removed from sysfs immediately, without waiting for the last reference to be released. This patch makes the sysfs removal happen immediately, so the caller may create a new kset with the same name as soon as kset_unregister() returns. Without this, every caller has to call "kobject_del(&kset->kobj)" first unless it knows it will never create a new kset with the same name. This sometimes shows up on module unload and reload, where the reload fails because it tries to create a kobject with the same name as one from the original load that still exists. CONFIG_DEBUG_KOBJECT_RELEASE=y makes this problem easier to hit. Signed-off-by:
Bjorn Helgaas <bhelgaas@google.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Bjorn Helgaas authored
When CONFIG_DEBUG_KOBJECT_RELEASE=y, delay kobject release functions for a random time between 1 and 8 seconds, which effectively changes the order in which they're called. Signed-off-by:
-
- 04 Dec, 2013 2 commits
-
-
Tejun Heo authored
kernfs inherited "security.*" xattr support from sysfs. This patch extends xattr support to "trusted.*" using simple_xattr_*(). As trusted xattrs are restricted to CAP_SYS_ADMIN, simple_xattr_*() which uses kernel memory for storage shouldn't be problematic. Note that the existing "security.*" support doesn't implement get/remove/list and the this patch only implements those ops for "trusted.*". We probably want to extend those ops to include support for "security.*". This patch will allow using kernfs from cgroup which requires "trusted.*" xattr support. Signed-off-by:
Tejun Heo <tj@kernel.org> Cc: David P. Quigley <dpquigl@tycho.nsa.gov> Signed-off-by:
-
Tejun Heo authored
sysfs_init_inode_attrs() is a bit clumsy to use requiring the caller to check whether @sd->s_iattr is already set or not. Rename it to sysfs_inode_attrs(), update it to check whether @sd->s_iattr is already initialized before trying to initialize it and return @sd->s_iattr. This simplifies the callers. While at it, * Rename struct sysfs_inode_attrs pointer variables to "attrs". As kernfs no longer deals with "struct attribute", this isn't confusing and makes it easier to distinguish from struct iattr pointers. * A new field will be added to sysfs_inode_attrs. Reindent in preparation. This patch doesn't introduce any behavior changes. Signed-off-by:
-
- 03 Dec, 2013 1 commit
-
-
Bjorn Helgaas authored
Previously, example_init() leaked a kset if any of the object creations failed. This fixes the leak by calling kset_unregister() in the error path. Signed-off-by:
-
- 30 Nov, 2013 35 commits
-
-
Tejun Heo authored
fs/kernfs/kernfs-internal.h needed to include fs/sysfs/sysfs.h because part of kernfs core implementation was living in sysfs. fs/sysfs/sysfs.h needed to include fs/kernfs/kernfs-internal.h because include/linux/kernfs.h didn't expose enough interface. The separation is complete and neither is true anymore. Remove the cross inclusion and make sysfs a proper user of kernfs. Signed-off-by:
-
Tejun Heo authored
fs/sysfs/symlink.c::sysfs_delete_link() tests @sd->s_flags for SYSFS_FLAG_NS. Let's add kernfs_ns_enabled() so that sysfs doesn't have to test sysfs_dirent flag directly. This makes things tidier for kernfs proper too. This is purely cosmetic. v2: To avoid possible NULL deref, use noop dummy implementation which always returns false when !CONFIG_SYSFS. Signed-off-by: -
Tejun Heo authored
sysfs_dirent includes some information which should be available to kernfs users - the type, flags, name and parent pointer. This patch moves sysfs_dirent definition from kernfs/kernfs-internal.h to include/linux/kernfs.h so that kernfs users can access them. The type part of flags is exported as enum kernfs_node_type, the flags kernfs_node_flag, sysfs_type() and kernfs_enable_ns() are moved to include/linux/kernfs.h and the former is updated to return the enum type. sysfs_dirent->s_parent and ->s_name are marked explicitly as public. This patch doesn't introduce any functional changes. v2: Flags exported too and kernfs_enable_ns() definition moved. v3: While moving kernfs_enable_ns() to include/linux/kernfs.h, v1 and v2 put the definition outside CONFIG_SYSFS replacing the dummy implementation with the actual implementation too. Unfortunately, this can lead to oops when !CONFIG_SYSFS because kernfs_enable_ns() may be called on a NULL @sd and now tries to dereference @sd instead of not doing anything. This issue was reported by Yuanhan Liu. Signed-off-by:Yuanhan Liu <yuanhan.liu@linux.intel.com> Signed-off-by:
-
Tejun Heo authored
Move core mount code to fs/kernfs/mount.c. The respective declarations in fs/sysfs/sysfs.h are moved to fs/kernfs/kernfs-internal.h. This is pure relocation. Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch rearranges mount path so that the kernfs and sysfs parts are separate. * As sysfs_super_info won't be visible outside kernfs proper, kernfs_super_ns() is added to allow kernfs users to access a super_block's namespace tag. * Generic mount operation is separated out into kernfs_mount_ns(). sysfs_mount() now just performs sysfs-specific permission check, acquires namespace tag, and invokes kernfs_mount_ns(). * Generic superblock release is separated out into kernfs_kill_sb() which can be used directly as file_system_type->kill_sb(). As sysfs needs to put the namespace tag, sysfs_kill_sb() wraps kernfs_kill_sb() with ns tag put. * sysfs_dir_cachep init and sysfs_inode_init() are separated out into kernfs_init(). kernfs_init() uses only small amount of memory and trying to handle and propagate kernfs_init() failure doesn't make much sense. Use SLAB_PANIC for sysfs_dir_cachep and make sysfs_inode_init() panic on failure. After this change, kernfs_init() should be called before sysfs_init(), fs/namespace.c::mnt_init() modified accordingly. Signed-off-by:
-
Tejun Heo authored
kernfs is being updated to allow multiple sysfs_dirent hierarchies so that it can also be used by other users. Currently, sysfs super_blocks are always attached to one kernfs_root - sysfs_root - and distinguished only by their namespace tags. This patch adds sysfs_super_info->root and update sysfs_fill/test_super() so that super_blocks are identified by the combination of both the associated kernfs_root and namespace tag. This allows mounting different kernfs hierarchies. Signed-off-by:
-
Tejun Heo authored
kernfs is being updated to allow multiple sysfs_dirent hierarchies so that it can also be used by other users. Currently, inode number is allocated using a global ida, sysfs_ino_ida; however, inos for different hierarchies should be handled separately. This patch makes ino allocation per kernfs_root. sysfs_ino_ida is replaced by kernfs_root->ino_ida and sysfs_new_dirent() is updated to take @root and allocate ino from it. ida_simple_get/remove() are used instead of sysfs_ino_lock and sysfs_alloc/free_ino(). Signed-off-by:
-
Tejun Heo authored
There currently is single kernfs hierarchy in the whole system which is used for sysfs. kernfs needs to support multiple hierarchies to allow other users. This patch introduces struct kernfs_root which serves as the root of each kernfs hierarchy and implements kernfs_create/destroy_root(). * Each kernfs_root is associated with a root sd (sysfs_dentry). The root is freed when the root sd is released and kernfs_destory_root() simply invokes kernfs_remove() on the root sd. sysfs_remove_one() is updated to handle release of the root sd. Note that ps_iattr update in sysfs_remove_one() is trivially updated for readability. * Root sd's are now dynamically allocated using sysfs_new_dirent(). Update sysfs_alloc_ino() so that it gives out ino from 1 so that the root sd still gets ino 1. * While kernfs currently only points to the root sd, it'll soon grow fields which are specific to each hierarchy. As determining a given sd's root will be necessary, sd->s_dir.root is added. This backlink fits better as a separate field in sd; however, sd->s_dir is inside union with space to spare, so use it to save space and provide kernfs_root() accessor to determine the root sd. * As hierarchies may be destroyed now, each mount needs to hold onto the hierarchy it's attached to. Update sysfs_fill_super() and sysfs_kill_sb() so that they get and put the kernfs_root respectively. * sysfs_root is replaced with kernfs_root which is dynamically created by invoking kernfs_create_root() from sysfs_init(). This patch doesn't introduce any visible behavior changes. v2: kernfs_create_root() forgot to set @sd->priv. Fixed. Signed-off-by:
-
Tejun Heo authored
Currently, it's assumed that there's a single kernfs hierarchy in the system anchored at sysfs_root which is defined as a global struct. To allow other users of kernfs, this will be made dynamic. Introduce a new global variable sysfs_root_sd which points to &sysfs_root and convert all &sysfs_root users. This patch doesn't introduce any behavior difference. Signed-off-by:
-
Tejun Heo authored
It has been very long since sysfs depended on vfs to keep track of internal states and whether sysfs is mounted or not doesn't make any difference to sysfs's internal operation. In addition to init and filesystem type registration, sysfs_init() invokes kern_mount() to create in-kernel mount of sysfs. This internal mounting doesn't server any purpose anymore. Remove it. Signed-off-by:
-
Tejun Heo authored
Add const qualifier to sysfs_super_info->ns so that it's consistent with other namespace tag usages in sysfs. Because kobject doesn't use const qualifier for namespace tags, this ends up requiring an explicit cast to drop const qualifier in free_sysfs_super_info(). Signed-off-by:
-
Tejun Heo authored
sysfs_fill_super() takes three params - @sb, @data and @silent - but uses only @sb. Drop the latter two. Signed-off-by:
-
Tejun Heo authored
Move core symlink code to fs/kernfs/symlink.c. fs/sysfs/symlink.c now only contains sysfs wrappers around kernfs interfaces. The respective declarations in fs/sysfs/sysfs.h are moved to fs/kernfs/kernfs-internal.h. This is pure relocation. Signed-off-by:
-
Tejun Heo authored
Move core file code to fs/kernfs/file.c. fs/sysfs/file.c now contains sysfs kernfs_ops callbacks, sysfs wrappers around kernfs interfaces, and sysfs_schedule_callback(). The respective declarations in fs/sysfs/sysfs.h are moved to fs/kernfs/kernfs-internal.h. This is pure relocation. v2: Refreshed on top of the v2 of "sysfs, kernfs: prepare read path for kernfs". v3: Refreshed on top of the v3 of "sysfs, kernfs: prepare read path for kernfs". Signed-off-by: -
Tejun Heo authored
Move core dir code to fs/kernfs/dir.c. fs/sysfs/dir.c now only contains sysfs_warn_dup() and sysfs wrappers around kernfs interfaces. The respective declarations in fs/sysfs/sysfs.h are moved to fs/kernfs/kernfs-internal.h. This is pure relocation. v2: sysfs_symlink_target_lock was mistakenly relocated to kernfs. It should remain with sysfs. Fixed. Signed-off-by: -
Tejun Heo authored
There's nothing sysfs-specific in fs/sysfs/inode.c. Move everything in it to fs/kernfs/inode.c. The respective declarations in fs/sysfs/sysfs.h are moved to fs/kernfs/kernfs-internal.h. This is pure relocation. Signed-off-by:
-
Tejun Heo authored
Move data structure, constant and basic accessor declarations from fs/sysfs/sysfs.h to fs/kernfs/kernfs-internal.h. The two files currently include each other. Once kernfs / sysfs separation is complete, the cross inclusions will be removed. Inclusion protectors are added to fs/sysfs/sysfs.h to allow cross-inclusion. This patch doesn't introduce any functional changes. Signed-off-by:
-
Tejun Heo authored
Introduce kernfs interface for finding, getting and putting sysfs_dirents. * sysfs_find_dirent() is renamed to kernfs_find_ns() and lockdep assertion for sysfs_mutex is added. * sysfs_get_dirent_ns() is renamed to kernfs_find_and_get(). * Macro inline dancing around __sysfs_get/put() are removed and kernfs_get/put() are made proper functions implemented in fs/sysfs/dir.c. While the conversions are mostly equivalent, there's one difference - kernfs_get() doesn't return the input param as its return value. This change is intentional. While passing through the input increases writability in some areas, it is unnecessary and has been shown to cause confusion regarding how the last ref is handled. Signed-off-by:
-
Tejun Heo authored
Currently, sysfs_dirent active_ref lockdep annotation uses attribute->[s]key as the lockdep key, which forces kernfs_create_file_ns() to assume that sysfs_dirent->priv is pointing to a struct attribute which may not be true for non-sysfs users. This patch restructures the lockdep annotation such that * kernfs_ops contains lockdep_key which is used by default for files created kernfs_create_file_ns(). * kernfs_create_file_ns_key() is introduced which takes an extra @key argument. The created file will use the specified key for active_ref lockdep annotation. If NULL is specified, lockdep for the file is disabled. * sysfs_add_file_mode_ns() is updated to use kernfs_create_file_ns_key() with the appropriate key from the attribute or NULL if ignore_lockdep is set. This makes the lockdep annotation properly contained in kernfs while allowing sysfs to cleanly keep its current behavior. This patch doesn't introduce any behavior differences. Signed-off-by:
-
Tejun Heo authored
We want to add one more SYSFS_FLAG_* but we can't use the next higher bit, 0x10000, as the flag field is 16bits wide. The flags are currently arranged weirdly - 8 bits are set aside for the type flags when there are only three three used, the first flag starts at 0x1000 instead of 0x0100 and flag literals have 5 digits (20 bits) when only 4 digits can be used. Rearrange them so that type bits are only the lowest four, flags start at 0x0010 and similar flags are grouped. This patch doesn't cause any behavior difference. Signed-off-by:
-
Tejun Heo authored
Introduce kernfs interface to wake up poll(2) which takes and returns sysfs_dirents. sysfs_notify_dirent() is renamed to kernfs_notify() and sysfs_notify() is updated so that it doesn't directly grab sysfs_mutex but acquires the target sysfs_dirents using sysfs_get_dirent(). sysfs_notify_dirent() is reimplemented as a dumb inline wrapper around kernfs_notify(). This patch doesn't introduce any behavior changes. Signed-off-by:
-
Tejun Heo authored
kernfs_ops currently only supports single_open() behavior which is pretty restrictive. Add optional callbacks ->seq_{start|next|stop}() which, when implemented, are invoked for seq_file traversal. This allows full seq_file functionality for kernfs users. This currently doesn't have any user and doesn't change any behavior. v2: Refreshed on top of the updated "sysfs, kernfs: prepare read path for kernfs". Signed-off-by: -
Tejun Heo authored
sysfs_add_one() is a wrapper around __sysfs_add_one() which prints out duplicate name warning if __sysfs_add_one() fails with -EEXIST. The previous kernfs conversions moved all dup warnings to sysfs interface functions and sysfs_add_one() doesn't have any user left. Remove sysfs_add_one() and update __sysfs_add_one() to take its name. This patch doesn't make any functional changes. Signed-off-by:
-
Tejun Heo authored
Introduce kernfs interface to create a file which takes and returns sysfs_dirents. The actual file creation part is separated out from sysfs_add_file_mode_ns() into kernfs_create_file_ns(). The former now only decides the kernfs_ops to use and the file's size and invokes the latter. This patch doesn't introduce behavior changes. v2: Dummy implementation for !CONFIG_SYSFS updated to return -ENOSYS. Signed-off-by:
-
Tejun Heo authored
After kernfs_ops and sysfs_dirent->s_attr.size addition, the distinction between SYSFS_KOBJ_BIN_ATTR and SYSFS_KOBJ_ATTR is only necessary while creating files to decide which kernfs_ops to use. Afterwards, they behave exactly the same. This patch removes SYSFS_KOBJ_BIN_ATTR along with sysfs_is_bin(). sysfs_add_file[_mode_ns]() are updated to take bool @is_bin instead of @type. This patch doesn't introduce any behavior changes. This completely isolates the distinction between the two sysfs file types in the sysfs layer proper. Signed-off-by:
-
Tejun Heo authored
sysfs sets the size of regular files unconditionally at PAGE_SIZE and takes the size of bin files from bin_attribute. The latter is a pretty bad interface which forces bin_attribute users to create a separate copy of bin_attribute for each instance of the file - e.g. pci resource files. Add sysfs_dirent->s_attr.size so that the size can be specified separately. This unifies inode init paths of ATTR and BIN_ATTR identical and allows for generic size handling for kernfs. Unfortunately, this grows the size of sysfs_dirent by sizeof(loff_t). Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch introduces kernfs_ops which hosts methods kernfs users implement and updates fs/sysfs/file.c such that sysfs_kf_*() functions populate kernfs_ops and kernfs_file_*() functions call the matching entries from kernfs_ops. kernfs_ops contains the following groups of methods. * seq_show() - for kernfs files which use seq_file for reads. * read() - for direct read implementations. Used iff seq_show() is not implemented. * write() - for writes. * mmap() - for mmaps. Notes: * sysfs_elem_attr->ops is added so that kernfs_ops can be accessed from sysfs_dirent. kernfs_ops() helper is added to verify locking and access the field. * SYSFS_FLAG_HAS_(SEQ_SHOW|MMAP) added. sd->s_attr->ops is accessible only while holding active_ref and there are cases where we want to take different actions depending on which ops are implemented. These two flags cache whether the two ops are implemented for those. * kernfs_file_*() no longer test sysfs type but chooses different behaviors depending on which methods in kernfs_ops are implemented. The conversions are trivial except for the open path. As kernfs_file_open() now decides whether to allow read/write accesses depending on the kernfs_ops implemented, the presence of methods in kobjs and attribute_bin should be propagated to kernfs_ops. sysfs_add_file_mode_ns() is updated so that it propagates presence / absence of the callbacks through _empty, _ro, _wo, _rw kernfs_ops. Signed-off-by:
-
Tejun Heo authored
sysfs_open_file will be used as the primary handle for kernfs methods. Move its definition from fs/sysfs/file.c to include/linux/kernfs.h and mark the public and private fields. This is pure relocation. Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch prepares the rest - open, release and poll. There isn't much to do. Just renaming is enough. As sysfs_file_operations and sysfs_bin_operations are identical now, use the same file_operations for both - kernfs_file_operations. This patch doesn't introduce any behavior changes. Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch rearranges mmap path so that the kernfs and sysfs parts are separate. sysfs_kf_bin_mmap() which handles the interaction with bin_attribute mmap method is factored out of sysfs_bin_mmap(), which is renamed to kernfs_file_mmap(). All vma ops are renamed accordingly. sysfs_bin_mmap() is updated such that it can be used for both file types. This will eventually allow using the same file_operations for both file types, which is necessary to separate out kernfs. This patch doesn't introduce any behavior changes. Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch rearranges write path so that the kernfs and sysfs parts are separate. kernfs_file_write() handles all boilerplate work including buffer management and locking and invokes sysfs_kf_write() or sysfs_kf_bin_write() depending on the file type which deals with the interaction with kobj store or bin_attribute write method. While this patch changes the order of some operations, it shouldn't change any visible behavior. Signed-off-by:
-
Tejun Heo authored
We're in the process of separating out core sysfs functionality into kernfs which will deal with sysfs_dirents directly. This patch rearranges read path so that the kernfs and sysfs parts are separate. * Regular file read path is refactored such that kernfs_seq_start/next/stop/show() handle all the boilerplate work including locking and updating event count for poll, while sysfs_kf_seq_show() deals with interaction with kobj show method. * Bin file read path is refactored such that kernfs_file_direct_read() handles all the boilerplate work including buffer management and locking, while sysfs_kf_bin_read() deals with interaction with bin_attribute read method. kernfs_file_read() is added. It invokes either the seq_file or direct read path depending on the file type. This will eventually allow using the same file_operations for both file types, which is necessary to separate out kernfs. While this patch changes the order of some operations, it shouldn't change any visible behavior. v2: Dropped unnecessary zeroing of @count from sysfs_kf_seq_show(). Add comments explaining single_open() behavior. Both suggested by Pavel. v3: seq_stop() is called even after seq_start() failed. kernfs_seq_start() updated so that it doesn't unlock sysfs_open_file->mutex on failure so that kernfs_seq_stop() doesn't try to unlock an already unlocked mutex. Reported by Fengguang. Signed-off-by: -
Tejun Heo authored
Introduce kernfs interface to manipulate a directory which takes and returns sysfs_dirents. create_dir() is renamed to kernfs_create_dir_ns() and its argumantes and return value are updated. create_dir() usages are replaced with kernfs_create_dir_ns() and sysfs_create_subdir() usages are replaced with kernfs_create_dir(). Dup warnings are handled explicitly by sysfs users of the kernfs interface. sysfs_enable_ns() is renamed to kernfs_enable_ns(). This patch doesn't introduce any behavior changes. v2: Dummy implementation for !CONFIG_SYSFS updated to return -ENOSYS. v3: kernfs_enable_ns() added. v4: Refreshed on top of "sysfs: drop kobj_ns_type handling, take #2" so that this patch removes sysfs_enable_ns(). Signed-off-by: -
Tejun Heo authored
A directory sysfs_dirent points to the associated kobj. A regular or bin file points to the associated [bin_]attribute. This patch replaces sysfs_dirent->s_dir.kobj and ->s_attr.[bin_]attr with void * ->priv. This is to prepare for kernfs interface so that sysfs can specify the private data in the same way for directories and files. This lower debuggability but not by much - the whole thing was overlaid in a union anyway. If debuggability becomes an issue, we can later add ->priv accessors which explicitly check for the sysfs_dirent type and performs casting. This patch doesn't introduce any behavior difference. Signed-off-by:
-
Greg Kroah-Hartman authored
We want those fixes in here as well. Signed-off-by:
-
