>> 1 $Id: README.Locking,v 1.4 2002/03/08 16:20:06 dwmw2 Exp $
1 2
2 JFFS2 LOCKING DOCUMENTATION 3 JFFS2 LOCKING DOCUMENTATION
3 --------------------------- 4 ---------------------------
4 5
>> 6 At least theoretically, JFFS2 does not require the Big Kernel Lock
>> 7 (BKL), which was always helpfully obtained for it by Linux 2.4 VFS
>> 8 code. It has its own locking, as described below.
>> 9
5 This document attempts to describe the existin 10 This document attempts to describe the existing locking rules for
6 JFFS2. It is not expected to remain perfectly 11 JFFS2. It is not expected to remain perfectly up to date, but ought to
7 be fairly close. 12 be fairly close.
8 13
9 14
10 alloc_sem 15 alloc_sem
11 --------- 16 ---------
12 17
13 The alloc_sem is a per-filesystem mutex, used !! 18 The alloc_sem is a per-filesystem semaphore, used primarily to ensure
14 contiguous allocation of space on the medium. 19 contiguous allocation of space on the medium. It is automatically
15 obtained during space allocations (jffs2_reser 20 obtained during space allocations (jffs2_reserve_space()) and freed
16 upon write completion (jffs2_complete_reservat 21 upon write completion (jffs2_complete_reservation()). Note that
17 the garbage collector will obtain this right a 22 the garbage collector will obtain this right at the beginning of
18 jffs2_garbage_collect_pass() and release it at 23 jffs2_garbage_collect_pass() and release it at the end, thereby
19 preventing any other write activity on the fil 24 preventing any other write activity on the file system during a
20 garbage collect pass. 25 garbage collect pass.
21 26
22 When writing new nodes, the alloc_sem must be 27 When writing new nodes, the alloc_sem must be held until the new nodes
23 have been properly linked into the data struct 28 have been properly linked into the data structures for the inode to
24 which they belong. This is for the benefit of 29 which they belong. This is for the benefit of NAND flash - adding new
25 nodes to an inode may obsolete old ones, and b 30 nodes to an inode may obsolete old ones, and by holding the alloc_sem
26 until this happens we ensure that any data in 31 until this happens we ensure that any data in the write-buffer at the
27 time this happens are part of the new node, no 32 time this happens are part of the new node, not just something that
28 was written afterwards. Hence, we can ensure t 33 was written afterwards. Hence, we can ensure the newly-obsoleted nodes
29 don't actually get erased until the write-buff 34 don't actually get erased until the write-buffer has been flushed to
30 the medium. 35 the medium.
31 36
32 With the introduction of NAND flash support an 37 With the introduction of NAND flash support and the write-buffer,
33 the alloc_sem is also used to protect the wbuf 38 the alloc_sem is also used to protect the wbuf-related members of the
34 jffs2_sb_info structure. Atomically reading th 39 jffs2_sb_info structure. Atomically reading the wbuf_len member to see
35 if the wbuf is currently holding any data is p 40 if the wbuf is currently holding any data is permitted, though.
36 41
37 Ordering constraints: See f->sem. 42 Ordering constraints: See f->sem.
38 43
39 44
40 File Mutex f->sem !! 45 File Semaphore f->sem
41 --------------------- 46 ---------------------
42 47
43 This is the JFFS2-internal equivalent of the i !! 48 This is the JFFS2-internal equivalent of the inode semaphore i->i_sem.
44 It protects the contents of the jffs2_inode_in 49 It protects the contents of the jffs2_inode_info private inode data,
45 including the linked list of node fragments (b 50 including the linked list of node fragments (but see the notes below on
46 erase_completion_lock), etc. 51 erase_completion_lock), etc.
47 52
48 The reason that the i_sem itself isn't used fo 53 The reason that the i_sem itself isn't used for this purpose is to
49 avoid deadlocks with garbage collection -- the 54 avoid deadlocks with garbage collection -- the VFS will lock the i_sem
50 before calling a function which may need to al 55 before calling a function which may need to allocate space. The
51 allocation may trigger garbage-collection, whi 56 allocation may trigger garbage-collection, which may need to move a
52 node belonging to the inode which was locked i 57 node belonging to the inode which was locked in the first place by the
53 VFS. If the garbage collection code were to at 58 VFS. If the garbage collection code were to attempt to lock the i_sem
54 of the inode from which it's garbage-collectin 59 of the inode from which it's garbage-collecting a physical node, this
55 lead to deadlock, unless we played games with 60 lead to deadlock, unless we played games with unlocking the i_sem
56 before calling the space allocation functions. 61 before calling the space allocation functions.
57 62
58 Instead of playing such games, we just have an 63 Instead of playing such games, we just have an extra internal
59 mutex, which is obtained by the garbage collec !! 64 semaphore, which is obtained by the garbage collection code and also
60 by the normal file system code _after_ allocat 65 by the normal file system code _after_ allocation of space.
61 66
62 Ordering constraints: 67 Ordering constraints:
63 68
64 1. Never attempt to allocate space or 69 1. Never attempt to allocate space or lock alloc_sem with
65 any f->sem held. 70 any f->sem held.
66 2. Never attempt to lock two file mute !! 71 2. Never attempt to lock two file semaphores in one thread.
67 No ordering rules have been made fo 72 No ordering rules have been made for doing so.
68 3. Never lock a page cache page with f <<
69 73
70 74
71 erase_completion_lock spinlock 75 erase_completion_lock spinlock
72 ------------------------------ 76 ------------------------------
73 77
74 This is used to serialise access to the eraseb 78 This is used to serialise access to the eraseblock lists, to the
75 per-eraseblock lists of physical jffs2_raw_nod 79 per-eraseblock lists of physical jffs2_raw_node_ref structures, and
76 (NB) the per-inode list of physical nodes. The 80 (NB) the per-inode list of physical nodes. The latter is a special
77 case - see below. 81 case - see below.
78 82
79 As the MTD API no longer permits erase-complet !! 83 As the MTD API permits erase-completion callback functions to be
80 to be called from bottom-half (timer) context !! 84 called from bottom-half (timer) context, and these functions access
81 ever actually implemented such a thing), it's !! 85 the data structures protected by this lock, it must be locked with
82 a simple spin_lock() rather than spin_lock_bh( !! 86 spin_lock_bh().
83 87
84 Note that the per-inode list of physical nodes 88 Note that the per-inode list of physical nodes (f->nodes) is a special
85 case. Any changes to _valid_ nodes (i.e. ->fla 89 case. Any changes to _valid_ nodes (i.e. ->flash_offset & 1 == 0) in
86 the list are protected by the file mutex f->se !! 90 the list are protected by the file semaphore f->sem. But the erase
87 may remove _obsolete_ nodes from the list whil !! 91 code may remove _obsolete_ nodes from the list while holding only the
88 erase_completion_lock. So you can walk the lis 92 erase_completion_lock. So you can walk the list only while holding the
89 erase_completion_lock, and can drop the lock t 93 erase_completion_lock, and can drop the lock temporarily mid-walk as
90 long as the pointer you're holding is to a _va 94 long as the pointer you're holding is to a _valid_ node, not an
91 obsolete one. 95 obsolete one.
92 96
93 The erase_completion_lock is also used to prot 97 The erase_completion_lock is also used to protect the c->gc_task
94 pointer when the garbage collection thread exi 98 pointer when the garbage collection thread exits. The code to kill the
95 GC thread locks it, sends the signal, then unl 99 GC thread locks it, sends the signal, then unlocks it - while the GC
96 thread itself locks it, zeroes c->gc_task, the 100 thread itself locks it, zeroes c->gc_task, then unlocks on the exit path.
97 101
>> 102 node_free_sem
>> 103 -------------
98 104
99 inocache_lock spinlock !! 105 This semaphore is only used by the erase code which frees obsolete
100 ---------------------- !! 106 node references and the jffs2_garbage_collect_deletion_dirent()
101 !! 107 function. The latter function on NAND flash must read _obsolete_ nodes
102 This spinlock protects the hashed list (c->ino !! 108 to determine whether the 'deletion dirent' under consideration can be
103 in-core jffs2_inode_cache objects (each inode <<
104 correspondent jffs2_inode_cache object). So, t <<
105 has to be locked while walking the c->inocache <<
106 <<
107 This spinlock also covers allocation of new in <<
108 currently just '++->highest_ino++', but might <<
109 if we need to deal with wrapping after 4 milli <<
110 <<
111 Note, the f->sem guarantees that the correspon <<
112 will not be removed. So, it is allowed to acce <<
113 the inocache_lock spinlock. <<
114 <<
115 Ordering constraints: <<
116 <<
117 If both erase_completion_lock and inoc <<
118 c->erase_completion has to be acquired <<
119 <<
120 <<
121 erase_free_sem <<
122 -------------- <<
123 <<
124 This mutex is only used by the erase code whic <<
125 references and the jffs2_garbage_collect_delet <<
126 The latter function on NAND flash must read _o <<
127 determine whether the 'deletion dirent' under <<
128 discarded or whether it is still required to s 109 discarded or whether it is still required to show that an inode has
129 been unlinked. Because reading from the flash 110 been unlinked. Because reading from the flash may sleep, the
130 erase_completion_lock cannot be held, so an al 111 erase_completion_lock cannot be held, so an alternative, more
131 heavyweight lock was required to prevent the e 112 heavyweight lock was required to prevent the erase code from freeing
132 the jffs2_raw_node_ref structures in question 113 the jffs2_raw_node_ref structures in question while the garbage
133 collection code is looking at them. 114 collection code is looking at them.
134 115
135 Suggestions for alternative solutions to this 116 Suggestions for alternative solutions to this problem would be welcomed.
136 <<
137 <<
138 wbuf_sem <<
139 -------- <<
140 <<
141 This read/write semaphore protects against con <<
142 write-behind buffer ('wbuf') used for flash ch <<
143 in blocks. It protects both the contents of th <<
144 which indicates which flash region (if any) is <<
145 the buffer. <<
146 <<
147 Ordering constraints: <<
148 Lock wbuf_sem last, after the alloc_se <<
149 <<
150 <<
151 c->xattr_sem <<
152 ------------ <<
153 <<
154 This read/write semaphore protects against con <<
155 xattr related objects which include stuff in s <<
156 In read-only path, write-semaphore is too much <<
157 by read-semaphore. But you must hold write-sem <<
158 creating or deleting any xattr related object. <<
159 <<
160 Once xattr_sem released, there would be no ass <<
161 of those objects. Thus, a series of processes <<
162 when updating such a object is necessary under <<
163 For example, do_jffs2_getxattr() holds read-se <<
164 xdatum at first. But it retries this process w <<
165 after release read-semaphore, if it's necessar <<
166 from medium. <<
167 <<
168 Ordering constraints: <<
169 Lock xattr_sem last, after the alloc_s <<
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.