173 lines
7.0 KiB
Plaintext
173 lines
7.0 KiB
Plaintext
|
|
JFFS2 LOCKING DOCUMENTATION
|
|
---------------------------
|
|
|
|
At least theoretically, JFFS2 does not require the Big Kernel Lock
|
|
(BKL), which was always helpfully obtained for it by Linux 2.4 VFS
|
|
code. It has its own locking, as described below.
|
|
|
|
This document attempts to describe the existing locking rules for
|
|
JFFS2. It is not expected to remain perfectly up to date, but ought to
|
|
be fairly close.
|
|
|
|
|
|
alloc_sem
|
|
---------
|
|
|
|
The alloc_sem is a per-filesystem mutex, used primarily to ensure
|
|
contiguous allocation of space on the medium. It is automatically
|
|
obtained during space allocations (jffs2_reserve_space()) and freed
|
|
upon write completion (jffs2_complete_reservation()). Note that
|
|
the garbage collector will obtain this right at the beginning of
|
|
jffs2_garbage_collect_pass() and release it at the end, thereby
|
|
preventing any other write activity on the file system during a
|
|
garbage collect pass.
|
|
|
|
When writing new nodes, the alloc_sem must be held until the new nodes
|
|
have been properly linked into the data structures for the inode to
|
|
which they belong. This is for the benefit of NAND flash - adding new
|
|
nodes to an inode may obsolete old ones, and by holding the alloc_sem
|
|
until this happens we ensure that any data in the write-buffer at the
|
|
time this happens are part of the new node, not just something that
|
|
was written afterwards. Hence, we can ensure the newly-obsoleted nodes
|
|
don't actually get erased until the write-buffer has been flushed to
|
|
the medium.
|
|
|
|
With the introduction of NAND flash support and the write-buffer,
|
|
the alloc_sem is also used to protect the wbuf-related members of the
|
|
jffs2_sb_info structure. Atomically reading the wbuf_len member to see
|
|
if the wbuf is currently holding any data is permitted, though.
|
|
|
|
Ordering constraints: See f->sem.
|
|
|
|
|
|
File Mutex f->sem
|
|
---------------------
|
|
|
|
This is the JFFS2-internal equivalent of the inode mutex i->i_sem.
|
|
It protects the contents of the jffs2_inode_info private inode data,
|
|
including the linked list of node fragments (but see the notes below on
|
|
erase_completion_lock), etc.
|
|
|
|
The reason that the i_sem itself isn't used for this purpose is to
|
|
avoid deadlocks with garbage collection -- the VFS will lock the i_sem
|
|
before calling a function which may need to allocate space. The
|
|
allocation may trigger garbage-collection, which may need to move a
|
|
node belonging to the inode which was locked in the first place by the
|
|
VFS. If the garbage collection code were to attempt to lock the i_sem
|
|
of the inode from which it's garbage-collecting a physical node, this
|
|
lead to deadlock, unless we played games with unlocking the i_sem
|
|
before calling the space allocation functions.
|
|
|
|
Instead of playing such games, we just have an extra internal
|
|
mutex, which is obtained by the garbage collection code and also
|
|
by the normal file system code _after_ allocation of space.
|
|
|
|
Ordering constraints:
|
|
|
|
1. Never attempt to allocate space or lock alloc_sem with
|
|
any f->sem held.
|
|
2. Never attempt to lock two file mutexes in one thread.
|
|
No ordering rules have been made for doing so.
|
|
|
|
|
|
erase_completion_lock spinlock
|
|
------------------------------
|
|
|
|
This is used to serialise access to the eraseblock lists, to the
|
|
per-eraseblock lists of physical jffs2_raw_node_ref structures, and
|
|
(NB) the per-inode list of physical nodes. The latter is a special
|
|
case - see below.
|
|
|
|
As the MTD API no longer permits erase-completion callback functions
|
|
to be called from bottom-half (timer) context (on the basis that nobody
|
|
ever actually implemented such a thing), it's now sufficient to use
|
|
a simple spin_lock() rather than spin_lock_bh().
|
|
|
|
Note that the per-inode list of physical nodes (f->nodes) is a special
|
|
case. Any changes to _valid_ nodes (i.e. ->flash_offset & 1 == 0) in
|
|
the list are protected by the file mutex f->sem. But the erase code
|
|
may remove _obsolete_ nodes from the list while holding only the
|
|
erase_completion_lock. So you can walk the list only while holding the
|
|
erase_completion_lock, and can drop the lock temporarily mid-walk as
|
|
long as the pointer you're holding is to a _valid_ node, not an
|
|
obsolete one.
|
|
|
|
The erase_completion_lock is also used to protect the c->gc_task
|
|
pointer when the garbage collection thread exits. The code to kill the
|
|
GC thread locks it, sends the signal, then unlocks it - while the GC
|
|
thread itself locks it, zeroes c->gc_task, then unlocks on the exit path.
|
|
|
|
|
|
inocache_lock spinlock
|
|
----------------------
|
|
|
|
This spinlock protects the hashed list (c->inocache_list) of the
|
|
in-core jffs2_inode_cache objects (each inode in JFFS2 has the
|
|
correspondent jffs2_inode_cache object). So, the inocache_lock
|
|
has to be locked while walking the c->inocache_list hash buckets.
|
|
|
|
This spinlock also covers allocation of new inode numbers, which is
|
|
currently just '++->highest_ino++', but might one day get more complicated
|
|
if we need to deal with wrapping after 4 milliard inode numbers are used.
|
|
|
|
Note, the f->sem guarantees that the correspondent jffs2_inode_cache
|
|
will not be removed. So, it is allowed to access it without locking
|
|
the inocache_lock spinlock.
|
|
|
|
Ordering constraints:
|
|
|
|
If both erase_completion_lock and inocache_lock are needed, the
|
|
c->erase_completion has to be acquired first.
|
|
|
|
|
|
erase_free_sem
|
|
--------------
|
|
|
|
This mutex is only used by the erase code which frees obsolete node
|
|
references and the jffs2_garbage_collect_deletion_dirent() function.
|
|
The latter function on NAND flash must read _obsolete_ nodes to
|
|
determine whether the 'deletion dirent' under consideration can be
|
|
discarded or whether it is still required to show that an inode has
|
|
been unlinked. Because reading from the flash may sleep, the
|
|
erase_completion_lock cannot be held, so an alternative, more
|
|
heavyweight lock was required to prevent the erase code from freeing
|
|
the jffs2_raw_node_ref structures in question while the garbage
|
|
collection code is looking at them.
|
|
|
|
Suggestions for alternative solutions to this problem would be welcomed.
|
|
|
|
|
|
wbuf_sem
|
|
--------
|
|
|
|
This read/write semaphore protects against concurrent access to the
|
|
write-behind buffer ('wbuf') used for flash chips where we must write
|
|
in blocks. It protects both the contents of the wbuf and the metadata
|
|
which indicates which flash region (if any) is currently covered by
|
|
the buffer.
|
|
|
|
Ordering constraints:
|
|
Lock wbuf_sem last, after the alloc_sem or and f->sem.
|
|
|
|
|
|
c->xattr_sem
|
|
------------
|
|
|
|
This read/write semaphore protects against concurrent access to the
|
|
xattr related objects which include stuff in superblock and ic->xref.
|
|
In read-only path, write-semaphore is too much exclusion. It's enough
|
|
by read-semaphore. But you must hold write-semaphore when updating,
|
|
creating or deleting any xattr related object.
|
|
|
|
Once xattr_sem released, there would be no assurance for the existence
|
|
of those objects. Thus, a series of processes is often required to retry,
|
|
when updating such a object is necessary under holding read semaphore.
|
|
For example, do_jffs2_getxattr() holds read-semaphore to scan xref and
|
|
xdatum at first. But it retries this process with holding write-semaphore
|
|
after release read-semaphore, if it's necessary to load name/value pair
|
|
from medium.
|
|
|
|
Ordering constraints:
|
|
Lock xattr_sem last, after the alloc_sem.
|