1Kernel Memory Leak Detector
2===========================
3
4Kmemleak provides a way of detecting possible kernel memory leaks in a
5way similar to a tracing garbage collector
6(https://en.wikipedia.org/wiki/Garbage_collection_%28computer_science%29#Tracing_garbage_collectors),
7with the difference that the orphan objects are not freed but only
8reported via /sys/kernel/debug/kmemleak. A similar method is used by the
9Valgrind tool (``memcheck --leak-check``) to detect the memory leaks in
10user-space applications.
11Kmemleak is supported on x86, arm, powerpc, sparc, sh, microblaze, ppc, mips, s390, metag and tile.
12
13Usage
14-----
15
16CONFIG_DEBUG_KMEMLEAK in "Kernel hacking" has to be enabled. A kernel
17thread scans the memory every 10 minutes (by default) and prints the
18number of new unreferenced objects found. To display the details of all
19the possible memory leaks::
20
21  # mount -t debugfs nodev /sys/kernel/debug/
22  # cat /sys/kernel/debug/kmemleak
23
24To trigger an intermediate memory scan::
25
26  # echo scan > /sys/kernel/debug/kmemleak
27
28To clear the list of all current possible memory leaks::
29
30  # echo clear > /sys/kernel/debug/kmemleak
31
32New leaks will then come up upon reading ``/sys/kernel/debug/kmemleak``
33again.
34
35Note that the orphan objects are listed in the order they were allocated
36and one object at the beginning of the list may cause other subsequent
37objects to be reported as orphan.
38
39Memory scanning parameters can be modified at run-time by writing to the
40``/sys/kernel/debug/kmemleak`` file. The following parameters are supported:
41
42- off
43    disable kmemleak (irreversible)
44- stack=on
45    enable the task stacks scanning (default)
46- stack=off
47    disable the tasks stacks scanning
48- scan=on
49    start the automatic memory scanning thread (default)
50- scan=off
51    stop the automatic memory scanning thread
52- scan=<secs>
53    set the automatic memory scanning period in seconds
54    (default 600, 0 to stop the automatic scanning)
55- scan
56    trigger a memory scan
57- clear
58    clear list of current memory leak suspects, done by
59    marking all current reported unreferenced objects grey,
60    or free all kmemleak objects if kmemleak has been disabled.
61- dump=<addr>
62    dump information about the object found at <addr>
63
64Kmemleak can also be disabled at boot-time by passing ``kmemleak=off`` on
65the kernel command line.
66
67Memory may be allocated or freed before kmemleak is initialised and
68these actions are stored in an early log buffer. The size of this buffer
69is configured via the CONFIG_DEBUG_KMEMLEAK_EARLY_LOG_SIZE option.
70
71If CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF are enabled, the kmemleak is
72disabled by default. Passing ``kmemleak=on`` on the kernel command
73line enables the function.
74
75Basic Algorithm
76---------------
77
78The memory allocations via :c:func:`kmalloc`, :c:func:`vmalloc`,
79:c:func:`kmem_cache_alloc` and
80friends are traced and the pointers, together with additional
81information like size and stack trace, are stored in a rbtree.
82The corresponding freeing function calls are tracked and the pointers
83removed from the kmemleak data structures.
84
85An allocated block of memory is considered orphan if no pointer to its
86start address or to any location inside the block can be found by
87scanning the memory (including saved registers). This means that there
88might be no way for the kernel to pass the address of the allocated
89block to a freeing function and therefore the block is considered a
90memory leak.
91
92The scanning algorithm steps:
93
94  1. mark all objects as white (remaining white objects will later be
95     considered orphan)
96  2. scan the memory starting with the data section and stacks, checking
97     the values against the addresses stored in the rbtree. If
98     a pointer to a white object is found, the object is added to the
99     gray list
100  3. scan the gray objects for matching addresses (some white objects
101     can become gray and added at the end of the gray list) until the
102     gray set is finished
103  4. the remaining white objects are considered orphan and reported via
104     /sys/kernel/debug/kmemleak
105
106Some allocated memory blocks have pointers stored in the kernel's
107internal data structures and they cannot be detected as orphans. To
108avoid this, kmemleak can also store the number of values pointing to an
109address inside the block address range that need to be found so that the
110block is not considered a leak. One example is __vmalloc().
111
112Testing specific sections with kmemleak
113---------------------------------------
114
115Upon initial bootup your /sys/kernel/debug/kmemleak output page may be
116quite extensive. This can also be the case if you have very buggy code
117when doing development. To work around these situations you can use the
118'clear' command to clear all reported unreferenced objects from the
119/sys/kernel/debug/kmemleak output. By issuing a 'scan' after a 'clear'
120you can find new unreferenced objects; this should help with testing
121specific sections of code.
122
123To test a critical section on demand with a clean kmemleak do::
124
125  # echo clear > /sys/kernel/debug/kmemleak
126  ... test your kernel or modules ...
127  # echo scan > /sys/kernel/debug/kmemleak
128
129Then as usual to get your report with::
130
131  # cat /sys/kernel/debug/kmemleak
132
133Freeing kmemleak internal objects
134---------------------------------
135
136To allow access to previously found memory leaks after kmemleak has been
137disabled by the user or due to an fatal error, internal kmemleak objects
138won't be freed when kmemleak is disabled, and those objects may occupy
139a large part of physical memory.
140
141In this situation, you may reclaim memory with::
142
143  # echo clear > /sys/kernel/debug/kmemleak
144
145Kmemleak API
146------------
147
148See the include/linux/kmemleak.h header for the functions prototype.
149
150- ``kmemleak_init``		 - initialize kmemleak
151- ``kmemleak_alloc``		 - notify of a memory block allocation
152- ``kmemleak_alloc_percpu``	 - notify of a percpu memory block allocation
153- ``kmemleak_vmalloc``		 - notify of a vmalloc() memory allocation
154- ``kmemleak_free``		 - notify of a memory block freeing
155- ``kmemleak_free_part``	 - notify of a partial memory block freeing
156- ``kmemleak_free_percpu``	 - notify of a percpu memory block freeing
157- ``kmemleak_update_trace``	 - update object allocation stack trace
158- ``kmemleak_not_leak``	 - mark an object as not a leak
159- ``kmemleak_ignore``		 - do not scan or report an object as leak
160- ``kmemleak_scan_area``	 - add scan areas inside a memory block
161- ``kmemleak_no_scan``	 - do not scan a memory block
162- ``kmemleak_erase``		 - erase an old value in a pointer variable
163- ``kmemleak_alloc_recursive`` - as kmemleak_alloc but checks the recursiveness
164- ``kmemleak_free_recursive``	 - as kmemleak_free but checks the recursiveness
165
166The following functions take a physical address as the object pointer
167and only perform the corresponding action if the address has a lowmem
168mapping:
169
170- ``kmemleak_alloc_phys``
171- ``kmemleak_free_part_phys``
172- ``kmemleak_not_leak_phys``
173- ``kmemleak_ignore_phys``
174
175Dealing with false positives/negatives
176--------------------------------------
177
178The false negatives are real memory leaks (orphan objects) but not
179reported by kmemleak because values found during the memory scanning
180point to such objects. To reduce the number of false negatives, kmemleak
181provides the kmemleak_ignore, kmemleak_scan_area, kmemleak_no_scan and
182kmemleak_erase functions (see above). The task stacks also increase the
183amount of false negatives and their scanning is not enabled by default.
184
185The false positives are objects wrongly reported as being memory leaks
186(orphan). For objects known not to be leaks, kmemleak provides the
187kmemleak_not_leak function. The kmemleak_ignore could also be used if
188the memory block is known not to contain other pointers and it will no
189longer be scanned.
190
191Some of the reported leaks are only transient, especially on SMP
192systems, because of pointers temporarily stored in CPU registers or
193stacks. Kmemleak defines MSECS_MIN_AGE (defaulting to 1000) representing
194the minimum age of an object to be reported as a memory leak.
195
196Limitations and Drawbacks
197-------------------------
198
199The main drawback is the reduced performance of memory allocation and
200freeing. To avoid other penalties, the memory scanning is only performed
201when the /sys/kernel/debug/kmemleak file is read. Anyway, this tool is
202intended for debugging purposes where the performance might not be the
203most important requirement.
204
205To keep the algorithm simple, kmemleak scans for values pointing to any
206address inside a block's address range. This may lead to an increased
207number of false negatives. However, it is likely that a real memory leak
208will eventually become visible.
209
210Another source of false negatives is the data stored in non-pointer
211values. In a future version, kmemleak could only scan the pointer
212members in the allocated structures. This feature would solve many of
213the false negative cases described above.
214
215The tool can report false positives. These are cases where an allocated
216block doesn't need to be freed (some cases in the init_call functions),
217the pointer is calculated by other methods than the usual container_of
218macro or the pointer is stored in a location not scanned by kmemleak.
219
220Page allocations and ioremap are not tracked.
221