xref: /linux/Documentation/dev-tools/kfence.rst (revision 0a94608f0f7de9b1135ffea3546afe68eafef57f)
1.. SPDX-License-Identifier: GPL-2.0
2.. Copyright (C) 2020, Google LLC.
3
4Kernel Electric-Fence (KFENCE)
5==============================
6
7Kernel Electric-Fence (KFENCE) is a low-overhead sampling-based memory safety
8error detector. KFENCE detects heap out-of-bounds access, use-after-free, and
9invalid-free errors.
10
11KFENCE is designed to be enabled in production kernels, and has near zero
12performance overhead. Compared to KASAN, KFENCE trades performance for
13precision. The main motivation behind KFENCE's design, is that with enough
14total uptime KFENCE will detect bugs in code paths not typically exercised by
15non-production test workloads. One way to quickly achieve a large enough total
16uptime is when the tool is deployed across a large fleet of machines.
17
18Usage
19-----
20
21To enable KFENCE, configure the kernel with::
22
23    CONFIG_KFENCE=y
24
25To build a kernel with KFENCE support, but disabled by default (to enable, set
26``kfence.sample_interval`` to non-zero value), configure the kernel with::
27
28    CONFIG_KFENCE=y
29    CONFIG_KFENCE_SAMPLE_INTERVAL=0
30
31KFENCE provides several other configuration options to customize behaviour (see
32the respective help text in ``lib/Kconfig.kfence`` for more info).
33
34Tuning performance
35~~~~~~~~~~~~~~~~~~
36
37The most important parameter is KFENCE's sample interval, which can be set via
38the kernel boot parameter ``kfence.sample_interval`` in milliseconds. The
39sample interval determines the frequency with which heap allocations will be
40guarded by KFENCE. The default is configurable via the Kconfig option
41``CONFIG_KFENCE_SAMPLE_INTERVAL``. Setting ``kfence.sample_interval=0``
42disables KFENCE.
43
44The sample interval controls a timer that sets up KFENCE allocations. By
45default, to keep the real sample interval predictable, the normal timer also
46causes CPU wake-ups when the system is completely idle. This may be undesirable
47on power-constrained systems. The boot parameter ``kfence.deferrable=1``
48instead switches to a "deferrable" timer which does not force CPU wake-ups on
49idle systems, at the risk of unpredictable sample intervals. The default is
50configurable via the Kconfig option ``CONFIG_KFENCE_DEFERRABLE``.
51
52.. warning::
53   The KUnit test suite is very likely to fail when using a deferrable timer
54   since it currently causes very unpredictable sample intervals.
55
56The KFENCE memory pool is of fixed size, and if the pool is exhausted, no
57further KFENCE allocations occur. With ``CONFIG_KFENCE_NUM_OBJECTS`` (default
58255), the number of available guarded objects can be controlled. Each object
59requires 2 pages, one for the object itself and the other one used as a guard
60page; object pages are interleaved with guard pages, and every object page is
61therefore surrounded by two guard pages.
62
63The total memory dedicated to the KFENCE memory pool can be computed as::
64
65    ( #objects + 1 ) * 2 * PAGE_SIZE
66
67Using the default config, and assuming a page size of 4 KiB, results in
68dedicating 2 MiB to the KFENCE memory pool.
69
70Note: On architectures that support huge pages, KFENCE will ensure that the
71pool is using pages of size ``PAGE_SIZE``. This will result in additional page
72tables being allocated.
73
74Error reports
75~~~~~~~~~~~~~
76
77A typical out-of-bounds access looks like this::
78
79    ==================================================================
80    BUG: KFENCE: out-of-bounds read in test_out_of_bounds_read+0xa6/0x234
81
82    Out-of-bounds read at 0xffff8c3f2e291fff (1B left of kfence-#72):
83     test_out_of_bounds_read+0xa6/0x234
84     kunit_try_run_case+0x61/0xa0
85     kunit_generic_run_threadfn_adapter+0x16/0x30
86     kthread+0x176/0x1b0
87     ret_from_fork+0x22/0x30
88
89    kfence-#72: 0xffff8c3f2e292000-0xffff8c3f2e29201f, size=32, cache=kmalloc-32
90
91    allocated by task 484 on cpu 0 at 32.919330s:
92     test_alloc+0xfe/0x738
93     test_out_of_bounds_read+0x9b/0x234
94     kunit_try_run_case+0x61/0xa0
95     kunit_generic_run_threadfn_adapter+0x16/0x30
96     kthread+0x176/0x1b0
97     ret_from_fork+0x22/0x30
98
99    CPU: 0 PID: 484 Comm: kunit_try_catch Not tainted 5.13.0-rc3+ #7
100    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
101    ==================================================================
102
103The header of the report provides a short summary of the function involved in
104the access. It is followed by more detailed information about the access and
105its origin. Note that, real kernel addresses are only shown when using the
106kernel command line option ``no_hash_pointers``.
107
108Use-after-free accesses are reported as::
109
110    ==================================================================
111    BUG: KFENCE: use-after-free read in test_use_after_free_read+0xb3/0x143
112
113    Use-after-free read at 0xffff8c3f2e2a0000 (in kfence-#79):
114     test_use_after_free_read+0xb3/0x143
115     kunit_try_run_case+0x61/0xa0
116     kunit_generic_run_threadfn_adapter+0x16/0x30
117     kthread+0x176/0x1b0
118     ret_from_fork+0x22/0x30
119
120    kfence-#79: 0xffff8c3f2e2a0000-0xffff8c3f2e2a001f, size=32, cache=kmalloc-32
121
122    allocated by task 488 on cpu 2 at 33.871326s:
123     test_alloc+0xfe/0x738
124     test_use_after_free_read+0x76/0x143
125     kunit_try_run_case+0x61/0xa0
126     kunit_generic_run_threadfn_adapter+0x16/0x30
127     kthread+0x176/0x1b0
128     ret_from_fork+0x22/0x30
129
130    freed by task 488 on cpu 2 at 33.871358s:
131     test_use_after_free_read+0xa8/0x143
132     kunit_try_run_case+0x61/0xa0
133     kunit_generic_run_threadfn_adapter+0x16/0x30
134     kthread+0x176/0x1b0
135     ret_from_fork+0x22/0x30
136
137    CPU: 2 PID: 488 Comm: kunit_try_catch Tainted: G    B             5.13.0-rc3+ #7
138    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
139    ==================================================================
140
141KFENCE also reports on invalid frees, such as double-frees::
142
143    ==================================================================
144    BUG: KFENCE: invalid free in test_double_free+0xdc/0x171
145
146    Invalid free of 0xffff8c3f2e2a4000 (in kfence-#81):
147     test_double_free+0xdc/0x171
148     kunit_try_run_case+0x61/0xa0
149     kunit_generic_run_threadfn_adapter+0x16/0x30
150     kthread+0x176/0x1b0
151     ret_from_fork+0x22/0x30
152
153    kfence-#81: 0xffff8c3f2e2a4000-0xffff8c3f2e2a401f, size=32, cache=kmalloc-32
154
155    allocated by task 490 on cpu 1 at 34.175321s:
156     test_alloc+0xfe/0x738
157     test_double_free+0x76/0x171
158     kunit_try_run_case+0x61/0xa0
159     kunit_generic_run_threadfn_adapter+0x16/0x30
160     kthread+0x176/0x1b0
161     ret_from_fork+0x22/0x30
162
163    freed by task 490 on cpu 1 at 34.175348s:
164     test_double_free+0xa8/0x171
165     kunit_try_run_case+0x61/0xa0
166     kunit_generic_run_threadfn_adapter+0x16/0x30
167     kthread+0x176/0x1b0
168     ret_from_fork+0x22/0x30
169
170    CPU: 1 PID: 490 Comm: kunit_try_catch Tainted: G    B             5.13.0-rc3+ #7
171    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
172    ==================================================================
173
174KFENCE also uses pattern-based redzones on the other side of an object's guard
175page, to detect out-of-bounds writes on the unprotected side of the object.
176These are reported on frees::
177
178    ==================================================================
179    BUG: KFENCE: memory corruption in test_kmalloc_aligned_oob_write+0xef/0x184
180
181    Corrupted memory at 0xffff8c3f2e33aff9 [ 0xac . . . . . . ] (in kfence-#156):
182     test_kmalloc_aligned_oob_write+0xef/0x184
183     kunit_try_run_case+0x61/0xa0
184     kunit_generic_run_threadfn_adapter+0x16/0x30
185     kthread+0x176/0x1b0
186     ret_from_fork+0x22/0x30
187
188    kfence-#156: 0xffff8c3f2e33afb0-0xffff8c3f2e33aff8, size=73, cache=kmalloc-96
189
190    allocated by task 502 on cpu 7 at 42.159302s:
191     test_alloc+0xfe/0x738
192     test_kmalloc_aligned_oob_write+0x57/0x184
193     kunit_try_run_case+0x61/0xa0
194     kunit_generic_run_threadfn_adapter+0x16/0x30
195     kthread+0x176/0x1b0
196     ret_from_fork+0x22/0x30
197
198    CPU: 7 PID: 502 Comm: kunit_try_catch Tainted: G    B             5.13.0-rc3+ #7
199    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
200    ==================================================================
201
202For such errors, the address where the corruption occurred as well as the
203invalidly written bytes (offset from the address) are shown; in this
204representation, '.' denote untouched bytes. In the example above ``0xac`` is
205the value written to the invalid address at offset 0, and the remaining '.'
206denote that no following bytes have been touched. Note that, real values are
207only shown if the kernel was booted with ``no_hash_pointers``; to avoid
208information disclosure otherwise, '!' is used instead to denote invalidly
209written bytes.
210
211And finally, KFENCE may also report on invalid accesses to any protected page
212where it was not possible to determine an associated object, e.g. if adjacent
213object pages had not yet been allocated::
214
215    ==================================================================
216    BUG: KFENCE: invalid read in test_invalid_access+0x26/0xe0
217
218    Invalid read at 0xffffffffb670b00a:
219     test_invalid_access+0x26/0xe0
220     kunit_try_run_case+0x51/0x85
221     kunit_generic_run_threadfn_adapter+0x16/0x30
222     kthread+0x137/0x160
223     ret_from_fork+0x22/0x30
224
225    CPU: 4 PID: 124 Comm: kunit_try_catch Tainted: G        W         5.8.0-rc6+ #7
226    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1 04/01/2014
227    ==================================================================
228
229DebugFS interface
230~~~~~~~~~~~~~~~~~
231
232Some debugging information is exposed via debugfs:
233
234* The file ``/sys/kernel/debug/kfence/stats`` provides runtime statistics.
235
236* The file ``/sys/kernel/debug/kfence/objects`` provides a list of objects
237  allocated via KFENCE, including those already freed but protected.
238
239Implementation Details
240----------------------
241
242Guarded allocations are set up based on the sample interval. After expiration
243of the sample interval, the next allocation through the main allocator (SLAB or
244SLUB) returns a guarded allocation from the KFENCE object pool (allocation
245sizes up to PAGE_SIZE are supported). At this point, the timer is reset, and
246the next allocation is set up after the expiration of the interval.
247
248When using ``CONFIG_KFENCE_STATIC_KEYS=y``, KFENCE allocations are "gated"
249through the main allocator's fast-path by relying on static branches via the
250static keys infrastructure. The static branch is toggled to redirect the
251allocation to KFENCE. Depending on sample interval, target workloads, and
252system architecture, this may perform better than the simple dynamic branch.
253Careful benchmarking is recommended.
254
255KFENCE objects each reside on a dedicated page, at either the left or right
256page boundaries selected at random. The pages to the left and right of the
257object page are "guard pages", whose attributes are changed to a protected
258state, and cause page faults on any attempted access. Such page faults are then
259intercepted by KFENCE, which handles the fault gracefully by reporting an
260out-of-bounds access, and marking the page as accessible so that the faulting
261code can (wrongly) continue executing (set ``panic_on_warn`` to panic instead).
262
263To detect out-of-bounds writes to memory within the object's page itself,
264KFENCE also uses pattern-based redzones. For each object page, a redzone is set
265up for all non-object memory. For typical alignments, the redzone is only
266required on the unguarded side of an object. Because KFENCE must honor the
267cache's requested alignment, special alignments may result in unprotected gaps
268on either side of an object, all of which are redzoned.
269
270The following figure illustrates the page layout::
271
272    ---+-----------+-----------+-----------+-----------+-----------+---
273       | xxxxxxxxx | O :       | xxxxxxxxx |       : O | xxxxxxxxx |
274       | xxxxxxxxx | B :       | xxxxxxxxx |       : B | xxxxxxxxx |
275       | x GUARD x | J : RED-  | x GUARD x | RED-  : J | x GUARD x |
276       | xxxxxxxxx | E :  ZONE | xxxxxxxxx |  ZONE : E | xxxxxxxxx |
277       | xxxxxxxxx | C :       | xxxxxxxxx |       : C | xxxxxxxxx |
278       | xxxxxxxxx | T :       | xxxxxxxxx |       : T | xxxxxxxxx |
279    ---+-----------+-----------+-----------+-----------+-----------+---
280
281Upon deallocation of a KFENCE object, the object's page is again protected and
282the object is marked as freed. Any further access to the object causes a fault
283and KFENCE reports a use-after-free access. Freed objects are inserted at the
284tail of KFENCE's freelist, so that the least recently freed objects are reused
285first, and the chances of detecting use-after-frees of recently freed objects
286is increased.
287
288If pool utilization reaches 75% (default) or above, to reduce the risk of the
289pool eventually being fully occupied by allocated objects yet ensure diverse
290coverage of allocations, KFENCE limits currently covered allocations of the
291same source from further filling up the pool. The "source" of an allocation is
292based on its partial allocation stack trace. A side-effect is that this also
293limits frequent long-lived allocations (e.g. pagecache) of the same source
294filling up the pool permanently, which is the most common risk for the pool
295becoming full and the sampled allocation rate dropping to zero. The threshold
296at which to start limiting currently covered allocations can be configured via
297the boot parameter ``kfence.skip_covered_thresh`` (pool usage%).
298
299Interface
300---------
301
302The following describes the functions which are used by allocators as well as
303page handling code to set up and deal with KFENCE allocations.
304
305.. kernel-doc:: include/linux/kfence.h
306   :functions: is_kfence_address
307               kfence_shutdown_cache
308               kfence_alloc kfence_free __kfence_free
309               kfence_ksize kfence_object_start
310               kfence_handle_page_fault
311
312Related Tools
313-------------
314
315In userspace, a similar approach is taken by `GWP-ASan
316<http://llvm.org/docs/GwpAsan.html>`_. GWP-ASan also relies on guard pages and
317a sampling strategy to detect memory unsafety bugs at scale. KFENCE's design is
318directly influenced by GWP-ASan, and can be seen as its kernel sibling. Another
319similar but non-sampling approach, that also inspired the name "KFENCE", can be
320found in the userspace `Electric Fence Malloc Debugger
321<https://linux.die.net/man/3/efence>`_.
322
323In the kernel, several tools exist to debug memory access errors, and in
324particular KASAN can detect all bug classes that KFENCE can detect. While KASAN
325is more precise, relying on compiler instrumentation, this comes at a
326performance cost.
327
328It is worth highlighting that KASAN and KFENCE are complementary, with
329different target environments. For instance, KASAN is the better debugging-aid,
330where test cases or reproducers exists: due to the lower chance to detect the
331error, it would require more effort using KFENCE to debug. Deployments at scale
332that cannot afford to enable KASAN, however, would benefit from using KFENCE to
333discover bugs due to code paths not exercised by test cases or fuzzers.
334