Commit | Line | Data |
---|---|---|
a1dac767 MCC |
1 | ======================= |
2 | Kernel Probes (Kprobes) | |
3 | ======================= | |
4 | ||
5 | :Author: Jim Keniston <jkenisto@us.ibm.com> | |
6 | :Author: Prasanna S Panchamukhi <prasanna.panchamukhi@gmail.com> | |
7 | :Author: Masami Hiramatsu <mhiramat@redhat.com> | |
8 | ||
9 | .. CONTENTS | |
10 | ||
9b17374e | 11 | 1. Concepts: Kprobes, and Return Probes |
a1dac767 MCC |
12 | 2. Architectures Supported |
13 | 3. Configuring Kprobes | |
14 | 4. API Reference | |
15 | 5. Kprobes Features and Limitations | |
16 | 6. Probe Overhead | |
17 | 7. TODO | |
18 | 8. Kprobes Example | |
9b17374e MH |
19 | 9. Kretprobes Example |
20 | 10. Deprecated Features | |
a1dac767 MCC |
21 | Appendix A: The kprobes debugfs interface |
22 | Appendix B: The kprobes sysctl interface | |
23 | ||
9b17374e | 24 | Concepts: Kprobes and Return Probes |
a1dac767 | 25 | ========================================= |
d27a4ddd JK |
26 | |
27 | Kprobes enables you to dynamically break into any kernel routine and | |
28 | collect debugging and performance information non-disruptively. You | |
a1dac767 | 29 | can trap at almost any kernel code address [1]_, specifying a handler |
d27a4ddd | 30 | routine to be invoked when the breakpoint is hit. |
a1dac767 MCC |
31 | |
32 | .. [1] some parts of the kernel code can not be trapped, see | |
33 | :ref:`kprobes_blacklist`) | |
d27a4ddd | 34 | |
9b17374e MH |
35 | There are currently two types of probes: kprobes, and kretprobes |
36 | (also called return probes). A kprobe can be inserted on virtually | |
37 | any instruction in the kernel. A return probe fires when a specified | |
38 | function returns. | |
d27a4ddd JK |
39 | |
40 | In the typical case, Kprobes-based instrumentation is packaged as | |
41 | a kernel module. The module's init function installs ("registers") | |
42 | one or more probes, and the exit function unregisters them. A | |
43 | registration function such as register_kprobe() specifies where | |
44 | the probe is to be inserted and what handler is to be called when | |
45 | the probe is hit. | |
46 | ||
a1dac767 MCC |
47 | There are also ``register_/unregister_*probes()`` functions for batch |
48 | registration/unregistration of a group of ``*probes``. These functions | |
3b0cb4ca MH |
49 | can speed up unregistration process when you have to unregister |
50 | a lot of probes at once. | |
51 | ||
b26486bf MH |
52 | The next four subsections explain how the different types of |
53 | probes work and how jump optimization works. They explain certain | |
54 | things that you'll need to know in order to make the best use of | |
55 | Kprobes -- e.g., the difference between a pre_handler and | |
56 | a post_handler, and how to use the maxactive and nmissed fields of | |
57 | a kretprobe. But if you're in a hurry to start using Kprobes, you | |
a1dac767 | 58 | can skip ahead to :ref:`kprobes_archs_supported`. |
d27a4ddd | 59 | |
a1dac767 MCC |
60 | How Does a Kprobe Work? |
61 | ----------------------- | |
d27a4ddd JK |
62 | |
63 | When a kprobe is registered, Kprobes makes a copy of the probed | |
64 | instruction and replaces the first byte(s) of the probed instruction | |
65 | with a breakpoint instruction (e.g., int3 on i386 and x86_64). | |
66 | ||
67 | When a CPU hits the breakpoint instruction, a trap occurs, the CPU's | |
68 | registers are saved, and control passes to Kprobes via the | |
69 | notifier_call_chain mechanism. Kprobes executes the "pre_handler" | |
70 | associated with the kprobe, passing the handler the addresses of the | |
71 | kprobe struct and the saved registers. | |
72 | ||
73 | Next, Kprobes single-steps its copy of the probed instruction. | |
74 | (It would be simpler to single-step the actual instruction in place, | |
75 | but then Kprobes would have to temporarily remove the breakpoint | |
76 | instruction. This would open a small time window when another CPU | |
77 | could sail right past the probepoint.) | |
78 | ||
79 | After the instruction is single-stepped, Kprobes executes the | |
80 | "post_handler," if any, that is associated with the kprobe. | |
81 | Execution then continues with the instruction following the probepoint. | |
82 | ||
dcce32d9 MH |
83 | Changing Execution Path |
84 | ----------------------- | |
85 | ||
01bdee64 MH |
86 | Since kprobes can probe into a running kernel code, it can change the |
87 | register set, including instruction pointer. This operation requires | |
88 | maximum care, such as keeping the stack frame, recovering the execution | |
89 | path etc. Since it operates on a running kernel and needs deep knowledge | |
90 | of computer architecture and concurrent computing, you can easily shoot | |
91 | your foot. | |
dcce32d9 MH |
92 | |
93 | If you change the instruction pointer (and set up other related | |
01bdee64 MH |
94 | registers) in pre_handler, you must return !0 so that kprobes stops |
95 | single stepping and just returns to the given address. | |
dcce32d9 MH |
96 | This also means post_handler should not be called anymore. |
97 | ||
01bdee64 MH |
98 | Note that this operation may be harder on some architectures which use |
99 | TOC (Table of Contents) for function call, since you have to setup a new | |
100 | TOC for your function in your module, and recover the old one after | |
101 | returning from it. | |
dcce32d9 | 102 | |
a1dac767 MCC |
103 | Return Probes |
104 | ------------- | |
f47cd9b5 | 105 | |
a1dac767 MCC |
106 | How Does a Return Probe Work? |
107 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
d27a4ddd JK |
108 | |
109 | When you call register_kretprobe(), Kprobes establishes a kprobe at | |
110 | the entry to the function. When the probed function is called and this | |
111 | probe is hit, Kprobes saves a copy of the return address, and replaces | |
112 | the return address with the address of a "trampoline." The trampoline | |
113 | is an arbitrary piece of code -- typically just a nop instruction. | |
114 | At boot time, Kprobes registers a kprobe at the trampoline. | |
115 | ||
116 | When the probed function executes its return instruction, control | |
117 | passes to the trampoline and that probe is hit. Kprobes' trampoline | |
f47cd9b5 AS |
118 | handler calls the user-specified return handler associated with the |
119 | kretprobe, then sets the saved instruction pointer to the saved return | |
120 | address, and that's where execution resumes upon return from the trap. | |
d27a4ddd JK |
121 | |
122 | While the probed function is executing, its return address is | |
123 | stored in an object of type kretprobe_instance. Before calling | |
124 | register_kretprobe(), the user sets the maxactive field of the | |
125 | kretprobe struct to specify how many instances of the specified | |
126 | function can be probed simultaneously. register_kretprobe() | |
127 | pre-allocates the indicated number of kretprobe_instance objects. | |
128 | ||
129 | For example, if the function is non-recursive and is called with a | |
130 | spinlock held, maxactive = 1 should be enough. If the function is | |
131 | non-recursive and can never relinquish the CPU (e.g., via a semaphore | |
132 | or preemption), NR_CPUS should be enough. If maxactive <= 0, it is | |
133 | set to a default value. If CONFIG_PREEMPT is enabled, the default | |
134 | is max(10, 2*NR_CPUS). Otherwise, the default is NR_CPUS. | |
135 | ||
136 | It's not a disaster if you set maxactive too low; you'll just miss | |
137 | some probes. In the kretprobe struct, the nmissed field is set to | |
138 | zero when the return probe is registered, and is incremented every | |
139 | time the probed function is entered but there is no kretprobe_instance | |
140 | object available for establishing the return probe. | |
141 | ||
a1dac767 MCC |
142 | Kretprobe entry-handler |
143 | ^^^^^^^^^^^^^^^^^^^^^^^ | |
f47cd9b5 AS |
144 | |
145 | Kretprobes also provides an optional user-specified handler which runs | |
146 | on function entry. This handler is specified by setting the entry_handler | |
147 | field of the kretprobe struct. Whenever the kprobe placed by kretprobe at the | |
148 | function entry is hit, the user-defined entry_handler, if any, is invoked. | |
149 | If the entry_handler returns 0 (success) then a corresponding return handler | |
150 | is guaranteed to be called upon function return. If the entry_handler | |
151 | returns a non-zero error then Kprobes leaves the return address as is, and | |
152 | the kretprobe has no further effect for that particular function instance. | |
153 | ||
154 | Multiple entry and return handler invocations are matched using the unique | |
155 | kretprobe_instance object associated with them. Additionally, a user | |
156 | may also specify per return-instance private data to be part of each | |
157 | kretprobe_instance object. This is especially useful when sharing private | |
158 | data between corresponding user entry and return handlers. The size of each | |
159 | private data object can be specified at kretprobe registration time by | |
160 | setting the data_size field of the kretprobe struct. This data can be | |
161 | accessed through the data field of each kretprobe_instance object. | |
162 | ||
163 | In case probed function is entered but there is no kretprobe_instance | |
164 | object available, then in addition to incrementing the nmissed count, | |
165 | the user entry_handler invocation is also skipped. | |
166 | ||
a1dac767 MCC |
167 | .. _kprobes_jump_optimization: |
168 | ||
169 | How Does Jump Optimization Work? | |
170 | -------------------------------- | |
b26486bf | 171 | |
5cc718b9 MH |
172 | If your kernel is built with CONFIG_OPTPROBES=y (currently this flag |
173 | is automatically set 'y' on x86/x86-64, non-preemptive kernel) and | |
b26486bf MH |
174 | the "debug.kprobes_optimization" kernel parameter is set to 1 (see |
175 | sysctl(8)), Kprobes tries to reduce probe-hit overhead by using a jump | |
176 | instruction instead of a breakpoint instruction at each probepoint. | |
177 | ||
a1dac767 MCC |
178 | Init a Kprobe |
179 | ^^^^^^^^^^^^^ | |
b26486bf MH |
180 | |
181 | When a probe is registered, before attempting this optimization, | |
182 | Kprobes inserts an ordinary, breakpoint-based kprobe at the specified | |
183 | address. So, even if it's not possible to optimize this particular | |
184 | probepoint, there'll be a probe there. | |
185 | ||
a1dac767 MCC |
186 | Safety Check |
187 | ^^^^^^^^^^^^ | |
b26486bf MH |
188 | |
189 | Before optimizing a probe, Kprobes performs the following safety checks: | |
190 | ||
191 | - Kprobes verifies that the region that will be replaced by the jump | |
a1dac767 MCC |
192 | instruction (the "optimized region") lies entirely within one function. |
193 | (A jump instruction is multiple bytes, and so may overlay multiple | |
194 | instructions.) | |
b26486bf MH |
195 | |
196 | - Kprobes analyzes the entire function and verifies that there is no | |
a1dac767 MCC |
197 | jump into the optimized region. Specifically: |
198 | ||
b26486bf MH |
199 | - the function contains no indirect jump; |
200 | - the function contains no instruction that causes an exception (since | |
a1dac767 MCC |
201 | the fixup code triggered by the exception could jump back into the |
202 | optimized region -- Kprobes checks the exception tables to verify this); | |
b26486bf | 203 | - there is no near jump to the optimized region (other than to the first |
a1dac767 | 204 | byte). |
b26486bf MH |
205 | |
206 | - For each instruction in the optimized region, Kprobes verifies that | |
a1dac767 | 207 | the instruction can be executed out of line. |
b26486bf | 208 | |
a1dac767 MCC |
209 | Preparing Detour Buffer |
210 | ^^^^^^^^^^^^^^^^^^^^^^^ | |
b26486bf MH |
211 | |
212 | Next, Kprobes prepares a "detour" buffer, which contains the following | |
213 | instruction sequence: | |
a1dac767 | 214 | |
b26486bf MH |
215 | - code to push the CPU's registers (emulating a breakpoint trap) |
216 | - a call to the trampoline code which calls user's probe handlers. | |
217 | - code to restore registers | |
218 | - the instructions from the optimized region | |
219 | - a jump back to the original execution path. | |
220 | ||
a1dac767 MCC |
221 | Pre-optimization |
222 | ^^^^^^^^^^^^^^^^ | |
b26486bf MH |
223 | |
224 | After preparing the detour buffer, Kprobes verifies that none of the | |
225 | following situations exist: | |
a1dac767 | 226 | |
9b17374e | 227 | - The probe has a post_handler. |
b26486bf MH |
228 | - Other instructions in the optimized region are probed. |
229 | - The probe is disabled. | |
a1dac767 | 230 | |
b26486bf MH |
231 | In any of the above cases, Kprobes won't start optimizing the probe. |
232 | Since these are temporary situations, Kprobes tries to start | |
233 | optimizing it again if the situation is changed. | |
234 | ||
235 | If the kprobe can be optimized, Kprobes enqueues the kprobe to an | |
236 | optimizing list, and kicks the kprobe-optimizer workqueue to optimize | |
237 | it. If the to-be-optimized probepoint is hit before being optimized, | |
238 | Kprobes returns control to the original instruction path by setting | |
239 | the CPU's instruction pointer to the copied code in the detour buffer | |
240 | -- thus at least avoiding the single-step. | |
241 | ||
a1dac767 MCC |
242 | Optimization |
243 | ^^^^^^^^^^^^ | |
b26486bf MH |
244 | |
245 | The Kprobe-optimizer doesn't insert the jump instruction immediately; | |
246 | rather, it calls synchronize_sched() for safety first, because it's | |
247 | possible for a CPU to be interrupted in the middle of executing the | |
a1dac767 | 248 | optimized region [3]_. As you know, synchronize_sched() can ensure |
b26486bf MH |
249 | that all interruptions that were active when synchronize_sched() |
250 | was called are done, but only if CONFIG_PREEMPT=n. So, this version | |
a1dac767 | 251 | of kprobe optimization supports only kernels with CONFIG_PREEMPT=n [4]_. |
b26486bf MH |
252 | |
253 | After that, the Kprobe-optimizer calls stop_machine() to replace | |
254 | the optimized region with a jump instruction to the detour buffer, | |
255 | using text_poke_smp(). | |
256 | ||
a1dac767 MCC |
257 | Unoptimization |
258 | ^^^^^^^^^^^^^^ | |
b26486bf MH |
259 | |
260 | When an optimized kprobe is unregistered, disabled, or blocked by | |
261 | another kprobe, it will be unoptimized. If this happens before | |
262 | the optimization is complete, the kprobe is just dequeued from the | |
263 | optimized list. If the optimization has been done, the jump is | |
264 | replaced with the original code (except for an int3 breakpoint in | |
265 | the first byte) by using text_poke_smp(). | |
266 | ||
a1dac767 MCC |
267 | .. [3] Please imagine that the 2nd instruction is interrupted and then |
268 | the optimizer replaces the 2nd instruction with the jump *address* | |
269 | while the interrupt handler is running. When the interrupt | |
270 | returns to original address, there is no valid instruction, | |
271 | and it causes an unexpected result. | |
b26486bf | 272 | |
a1dac767 MCC |
273 | .. [4] This optimization-safety checking may be replaced with the |
274 | stop-machine method that ksplice uses for supporting a CONFIG_PREEMPT=y | |
275 | kernel. | |
b26486bf MH |
276 | |
277 | NOTE for geeks: | |
278 | The jump optimization changes the kprobe's pre_handler behavior. | |
279 | Without optimization, the pre_handler can change the kernel's execution | |
280 | path by changing regs->ip and returning 1. However, when the probe | |
281 | is optimized, that modification is ignored. Thus, if you want to | |
282 | tweak the kernel's execution path, you need to suppress optimization, | |
283 | using one of the following techniques: | |
a1dac767 | 284 | |
059053a2 | 285 | - Specify an empty function for the kprobe's post_handler. |
a1dac767 MCC |
286 | |
287 | or | |
288 | ||
b26486bf MH |
289 | - Execute 'sysctl -w debug.kprobes_optimization=n' |
290 | ||
a1dac767 MCC |
291 | .. _kprobes_blacklist: |
292 | ||
293 | Blacklist | |
294 | --------- | |
376e2424 MH |
295 | |
296 | Kprobes can probe most of the kernel except itself. This means | |
297 | that there are some functions where kprobes cannot probe. Probing | |
298 | (trapping) such functions can cause a recursive trap (e.g. double | |
299 | fault) or the nested probe handler may never be called. | |
300 | Kprobes manages such functions as a blacklist. | |
301 | If you want to add a function into the blacklist, you just need | |
302 | to (1) include linux/kprobes.h and (2) use NOKPROBE_SYMBOL() macro | |
303 | to specify a blacklisted function. | |
304 | Kprobes checks the given probe address against the blacklist and | |
305 | rejects registering it, if the given address is in the blacklist. | |
306 | ||
a1dac767 MCC |
307 | .. _kprobes_archs_supported: |
308 | ||
309 | Architectures Supported | |
310 | ======================= | |
d27a4ddd | 311 | |
9b17374e | 312 | Kprobes and return probes are implemented on the following |
d27a4ddd JK |
313 | architectures: |
314 | ||
b26486bf MH |
315 | - i386 (Supports jump optimization) |
316 | - x86_64 (AMD-64, EM64T) (Supports jump optimization) | |
d27a4ddd | 317 | - ppc64 |
8861da31 | 318 | - ia64 (Does not support probes on instruction slot1.) |
d27a4ddd | 319 | - sparc64 (Return probes not yet implemented.) |
5de865b4 | 320 | - arm |
f8279621 | 321 | - ppc |
9bb4d9df | 322 | - mips |
369e8c35 | 323 | - s390 |
d27a4ddd | 324 | |
a1dac767 MCC |
325 | Configuring Kprobes |
326 | =================== | |
d27a4ddd JK |
327 | |
328 | When configuring the kernel using make menuconfig/xconfig/oldconfig, | |
080684c8 LB |
329 | ensure that CONFIG_KPROBES is set to "y". Under "General setup", look |
330 | for "Kprobes". | |
8861da31 JK |
331 | |
332 | So that you can load and unload Kprobes-based instrumentation modules, | |
333 | make sure "Loadable module support" (CONFIG_MODULES) and "Module | |
334 | unloading" (CONFIG_MODULE_UNLOAD) are set to "y". | |
d27a4ddd | 335 | |
09b18203 AM |
336 | Also make sure that CONFIG_KALLSYMS and perhaps even CONFIG_KALLSYMS_ALL |
337 | are set to "y", since kallsyms_lookup_name() is used by the in-kernel | |
338 | kprobe address resolution code. | |
d27a4ddd JK |
339 | |
340 | If you need to insert a probe in the middle of a function, you may find | |
341 | it useful to "Compile the kernel with debug info" (CONFIG_DEBUG_INFO), | |
342 | so you can use "objdump -d -l vmlinux" to see the source-to-object | |
343 | code mapping. | |
344 | ||
a1dac767 MCC |
345 | API Reference |
346 | ============= | |
d27a4ddd JK |
347 | |
348 | The Kprobes API includes a "register" function and an "unregister" | |
3b0cb4ca MH |
349 | function for each type of probe. The API also includes "register_*probes" |
350 | and "unregister_*probes" functions for (un)registering arrays of probes. | |
351 | Here are terse, mini-man-page specifications for these functions and | |
352 | the associated probe handlers that you'll write. See the files in the | |
353 | samples/kprobes/ sub-directory for examples. | |
d27a4ddd | 354 | |
a1dac767 MCC |
355 | register_kprobe |
356 | --------------- | |
d27a4ddd | 357 | |
a1dac767 MCC |
358 | :: |
359 | ||
360 | #include <linux/kprobes.h> | |
361 | int register_kprobe(struct kprobe *kp); | |
d27a4ddd JK |
362 | |
363 | Sets a breakpoint at the address kp->addr. When the breakpoint is | |
364 | hit, Kprobes calls kp->pre_handler. After the probed instruction | |
365 | is single-stepped, Kprobe calls kp->post_handler. If a fault | |
366 | occurs during execution of kp->pre_handler or kp->post_handler, | |
367 | or during single-stepping of the probed instruction, Kprobes calls | |
de5bd88d MH |
368 | kp->fault_handler. Any or all handlers can be NULL. If kp->flags |
369 | is set KPROBE_FLAG_DISABLED, that kp will be registered but disabled, | |
a33f3224 | 370 | so, its handlers aren't hit until calling enable_kprobe(kp). |
d27a4ddd | 371 | |
a1dac767 MCC |
372 | .. note:: |
373 | ||
374 | 1. With the introduction of the "symbol_name" field to struct kprobe, | |
375 | the probepoint address resolution will now be taken care of by the kernel. | |
376 | The following will now work:: | |
09b18203 AM |
377 | |
378 | kp.symbol_name = "symbol_name"; | |
379 | ||
a1dac767 MCC |
380 | (64-bit powerpc intricacies such as function descriptors are handled |
381 | transparently) | |
09b18203 | 382 | |
a1dac767 MCC |
383 | 2. Use the "offset" field of struct kprobe if the offset into the symbol |
384 | to install a probepoint is known. This field is used to calculate the | |
385 | probepoint. | |
09b18203 | 386 | |
a1dac767 MCC |
387 | 3. Specify either the kprobe "symbol_name" OR the "addr". If both are |
388 | specified, kprobe registration will fail with -EINVAL. | |
09b18203 | 389 | |
a1dac767 MCC |
390 | 4. With CISC architectures (such as i386 and x86_64), the kprobes code |
391 | does not validate if the kprobe.addr is at an instruction boundary. | |
392 | Use "offset" with caution. | |
09b18203 | 393 | |
d27a4ddd JK |
394 | register_kprobe() returns 0 on success, or a negative errno otherwise. |
395 | ||
a1dac767 MCC |
396 | User's pre-handler (kp->pre_handler):: |
397 | ||
398 | #include <linux/kprobes.h> | |
399 | #include <linux/ptrace.h> | |
400 | int pre_handler(struct kprobe *p, struct pt_regs *regs); | |
d27a4ddd JK |
401 | |
402 | Called with p pointing to the kprobe associated with the breakpoint, | |
403 | and regs pointing to the struct containing the registers saved when | |
404 | the breakpoint was hit. Return 0 here unless you're a Kprobes geek. | |
405 | ||
a1dac767 MCC |
406 | User's post-handler (kp->post_handler):: |
407 | ||
408 | #include <linux/kprobes.h> | |
409 | #include <linux/ptrace.h> | |
410 | void post_handler(struct kprobe *p, struct pt_regs *regs, | |
411 | unsigned long flags); | |
d27a4ddd JK |
412 | |
413 | p and regs are as described for the pre_handler. flags always seems | |
414 | to be zero. | |
415 | ||
a1dac767 MCC |
416 | User's fault-handler (kp->fault_handler):: |
417 | ||
418 | #include <linux/kprobes.h> | |
419 | #include <linux/ptrace.h> | |
420 | int fault_handler(struct kprobe *p, struct pt_regs *regs, int trapnr); | |
d27a4ddd JK |
421 | |
422 | p and regs are as described for the pre_handler. trapnr is the | |
423 | architecture-specific trap number associated with the fault (e.g., | |
424 | on i386, 13 for a general protection fault or 14 for a page fault). | |
425 | Returns 1 if it successfully handled the exception. | |
426 | ||
a1dac767 MCC |
427 | register_kretprobe |
428 | ------------------ | |
d27a4ddd | 429 | |
a1dac767 MCC |
430 | :: |
431 | ||
432 | #include <linux/kprobes.h> | |
433 | int register_kretprobe(struct kretprobe *rp); | |
d27a4ddd JK |
434 | |
435 | Establishes a return probe for the function whose address is | |
436 | rp->kp.addr. When that function returns, Kprobes calls rp->handler. | |
437 | You must set rp->maxactive appropriately before you call | |
438 | register_kretprobe(); see "How Does a Return Probe Work?" for details. | |
439 | ||
440 | register_kretprobe() returns 0 on success, or a negative errno | |
441 | otherwise. | |
442 | ||
a1dac767 MCC |
443 | User's return-probe handler (rp->handler):: |
444 | ||
445 | #include <linux/kprobes.h> | |
446 | #include <linux/ptrace.h> | |
447 | int kretprobe_handler(struct kretprobe_instance *ri, | |
448 | struct pt_regs *regs); | |
d27a4ddd JK |
449 | |
450 | regs is as described for kprobe.pre_handler. ri points to the | |
451 | kretprobe_instance object, of which the following fields may be | |
452 | of interest: | |
a1dac767 | 453 | |
d27a4ddd JK |
454 | - ret_addr: the return address |
455 | - rp: points to the corresponding kretprobe object | |
456 | - task: points to the corresponding task struct | |
f47cd9b5 AS |
457 | - data: points to per return-instance private data; see "Kretprobe |
458 | entry-handler" for details. | |
09b18203 AM |
459 | |
460 | The regs_return_value(regs) macro provides a simple abstraction to | |
461 | extract the return value from the appropriate register as defined by | |
462 | the architecture's ABI. | |
463 | ||
d27a4ddd JK |
464 | The handler's return value is currently ignored. |
465 | ||
a1dac767 MCC |
466 | unregister_*probe |
467 | ------------------ | |
d27a4ddd | 468 | |
a1dac767 MCC |
469 | :: |
470 | ||
471 | #include <linux/kprobes.h> | |
472 | void unregister_kprobe(struct kprobe *kp); | |
a1dac767 | 473 | void unregister_kretprobe(struct kretprobe *rp); |
d27a4ddd JK |
474 | |
475 | Removes the specified probe. The unregister function can be called | |
476 | at any time after the probe has been registered. | |
477 | ||
a1dac767 MCC |
478 | .. note:: |
479 | ||
480 | If the functions find an incorrect probe (ex. an unregistered probe), | |
481 | they clear the addr field of the probe. | |
3b0cb4ca | 482 | |
a1dac767 MCC |
483 | register_*probes |
484 | ---------------- | |
3b0cb4ca | 485 | |
a1dac767 MCC |
486 | :: |
487 | ||
488 | #include <linux/kprobes.h> | |
489 | int register_kprobes(struct kprobe **kps, int num); | |
490 | int register_kretprobes(struct kretprobe **rps, int num); | |
3b0cb4ca MH |
491 | |
492 | Registers each of the num probes in the specified array. If any | |
493 | error occurs during registration, all probes in the array, up to | |
494 | the bad probe, are safely unregistered before the register_*probes | |
495 | function returns. | |
a1dac767 | 496 | |
bc8c9da5 | 497 | - kps/rps: an array of pointers to ``*probe`` data structures |
3b0cb4ca MH |
498 | - num: the number of the array entries. |
499 | ||
a1dac767 MCC |
500 | .. note:: |
501 | ||
502 | You have to allocate(or define) an array of pointers and set all | |
503 | of the array entries before using these functions. | |
504 | ||
505 | unregister_*probes | |
506 | ------------------ | |
3b0cb4ca | 507 | |
a1dac767 | 508 | :: |
3b0cb4ca | 509 | |
a1dac767 MCC |
510 | #include <linux/kprobes.h> |
511 | void unregister_kprobes(struct kprobe **kps, int num); | |
512 | void unregister_kretprobes(struct kretprobe **rps, int num); | |
3b0cb4ca MH |
513 | |
514 | Removes each of the num probes in the specified array at once. | |
515 | ||
a1dac767 | 516 | .. note:: |
3b0cb4ca | 517 | |
a1dac767 MCC |
518 | If the functions find some incorrect probes (ex. unregistered |
519 | probes) in the specified array, they clear the addr field of those | |
520 | incorrect probes. However, other probes in the array are | |
521 | unregistered correctly. | |
de5bd88d | 522 | |
a1dac767 MCC |
523 | disable_*probe |
524 | -------------- | |
de5bd88d | 525 | |
a1dac767 MCC |
526 | :: |
527 | ||
528 | #include <linux/kprobes.h> | |
529 | int disable_kprobe(struct kprobe *kp); | |
530 | int disable_kretprobe(struct kretprobe *rp); | |
a1dac767 MCC |
531 | |
532 | Temporarily disables the specified ``*probe``. You can enable it again by using | |
8f9b1528 | 533 | enable_*probe(). You must specify the probe which has been registered. |
de5bd88d | 534 | |
a1dac767 MCC |
535 | enable_*probe |
536 | ------------- | |
de5bd88d | 537 | |
a1dac767 | 538 | :: |
de5bd88d | 539 | |
a1dac767 MCC |
540 | #include <linux/kprobes.h> |
541 | int enable_kprobe(struct kprobe *kp); | |
542 | int enable_kretprobe(struct kretprobe *rp); | |
a1dac767 MCC |
543 | |
544 | Enables ``*probe`` which has been disabled by disable_*probe(). You must specify | |
8f9b1528 | 545 | the probe which has been registered. |
de5bd88d | 546 | |
a1dac767 MCC |
547 | Kprobes Features and Limitations |
548 | ================================ | |
d27a4ddd | 549 | |
9b17374e MH |
550 | Kprobes allows multiple probes at the same address. Also, |
551 | a probepoint for which there is a post_handler cannot be optimized. | |
552 | So if you install a kprobe with a post_handler, at an optimized | |
553 | probepoint, the probepoint will be unoptimized automatically. | |
d27a4ddd JK |
554 | |
555 | In general, you can install a probe anywhere in the kernel. | |
556 | In particular, you can probe interrupt handlers. Known exceptions | |
557 | are discussed in this section. | |
558 | ||
8861da31 JK |
559 | The register_*probe functions will return -EINVAL if you attempt |
560 | to install a probe in the code that implements Kprobes (mostly | |
a1dac767 | 561 | kernel/kprobes.c and ``arch/*/kernel/kprobes.c``, but also functions such |
8861da31 | 562 | as do_page_fault and notifier_call_chain). |
d27a4ddd JK |
563 | |
564 | If you install a probe in an inline-able function, Kprobes makes | |
565 | no attempt to chase down all inline instances of the function and | |
566 | install probes there. gcc may inline a function without being asked, | |
567 | so keep this in mind if you're not seeing the probe hits you expect. | |
568 | ||
569 | A probe handler can modify the environment of the probed function | |
570 | -- e.g., by modifying kernel data structures, or by modifying the | |
571 | contents of the pt_regs struct (which are restored to the registers | |
572 | upon return from the breakpoint). So Kprobes can be used, for example, | |
573 | to install a bug fix or to inject faults for testing. Kprobes, of | |
574 | course, has no way to distinguish the deliberately injected faults | |
575 | from the accidental ones. Don't drink and probe. | |
576 | ||
577 | Kprobes makes no attempt to prevent probe handlers from stepping on | |
578 | each other -- e.g., probing printk() and then calling printk() from a | |
8861da31 JK |
579 | probe handler. If a probe handler hits a probe, that second probe's |
580 | handlers won't be run in that instance, and the kprobe.nmissed member | |
581 | of the second probe will be incremented. | |
582 | ||
583 | As of Linux v2.6.15-rc1, multiple handlers (or multiple instances of | |
584 | the same handler) may run concurrently on different CPUs. | |
585 | ||
586 | Kprobes does not use mutexes or allocate memory except during | |
d27a4ddd JK |
587 | registration and unregistration. |
588 | ||
2bbda764 MH |
589 | Probe handlers are run with preemption disabled or interrupt disabled, |
590 | which depends on the architecture and optimization state. (e.g., | |
591 | kretprobe handlers and optimized kprobe handlers run without interrupt | |
592 | disabled on x86/x86-64). In any case, your handler should not yield | |
593 | the CPU (e.g., by attempting to acquire a semaphore, or waiting I/O). | |
d27a4ddd JK |
594 | |
595 | Since a return probe is implemented by replacing the return | |
596 | address with the trampoline's address, stack backtraces and calls | |
597 | to __builtin_return_address() will typically yield the trampoline's | |
598 | address instead of the real return address for kretprobed functions. | |
599 | (As far as we can tell, __builtin_return_address() is used only | |
600 | for instrumentation and error reporting.) | |
601 | ||
8861da31 JK |
602 | If the number of times a function is called does not match the number |
603 | of times it returns, registering a return probe on that function may | |
bf8f6e5b AM |
604 | produce undesirable results. In such a case, a line: |
605 | kretprobe BUG!: Processing kretprobe d000000000041aa8 @ c00000000004f48c | |
606 | gets printed. With this information, one will be able to correlate the | |
607 | exact instance of the kretprobe that caused the problem. We have the | |
608 | do_exit() case covered. do_execve() and do_fork() are not an issue. | |
609 | We're unaware of other specific cases where this could be a problem. | |
8861da31 JK |
610 | |
611 | If, upon entry to or exit from a function, the CPU is running on | |
612 | a stack other than that of the current task, registering a return | |
613 | probe on that function may produce undesirable results. For this | |
9b17374e | 614 | reason, Kprobes doesn't support return probes (or kprobes) |
8861da31 JK |
615 | on the x86_64 version of __switch_to(); the registration functions |
616 | return -EINVAL. | |
d27a4ddd | 617 | |
b26486bf MH |
618 | On x86/x86-64, since the Jump Optimization of Kprobes modifies |
619 | instructions widely, there are some limitations to optimization. To | |
620 | explain it, we introduce some terminology. Imagine a 3-instruction | |
621 | sequence consisting of a two 2-byte instructions and one 3-byte | |
622 | instruction. | |
623 | ||
a1dac767 MCC |
624 | :: |
625 | ||
626 | IA | |
627 | | | |
628 | [-2][-1][0][1][2][3][4][5][6][7] | |
629 | [ins1][ins2][ ins3 ] | |
630 | [<- DCR ->] | |
631 | [<- JTPR ->] | |
b26486bf | 632 | |
a1dac767 MCC |
633 | ins1: 1st Instruction |
634 | ins2: 2nd Instruction | |
635 | ins3: 3rd Instruction | |
636 | IA: Insertion Address | |
637 | JTPR: Jump Target Prohibition Region | |
638 | DCR: Detoured Code Region | |
b26486bf MH |
639 | |
640 | The instructions in DCR are copied to the out-of-line buffer | |
641 | of the kprobe, because the bytes in DCR are replaced by | |
642 | a 5-byte jump instruction. So there are several limitations. | |
643 | ||
644 | a) The instructions in DCR must be relocatable. | |
645 | b) The instructions in DCR must not include a call instruction. | |
646 | c) JTPR must not be targeted by any jump or call instruction. | |
b595076a | 647 | d) DCR must not straddle the border between functions. |
b26486bf MH |
648 | |
649 | Anyway, these limitations are checked by the in-kernel instruction | |
650 | decoder, so you don't need to worry about that. | |
651 | ||
a1dac767 MCC |
652 | Probe Overhead |
653 | ============== | |
d27a4ddd JK |
654 | |
655 | On a typical CPU in use in 2005, a kprobe hit takes 0.5 to 1.0 | |
656 | microseconds to process. Specifically, a benchmark that hits the same | |
657 | probepoint repeatedly, firing a simple handler each time, reports 1-2 | |
9b17374e MH |
658 | million hits per second, depending on the architecture. A return-probe |
659 | hit typically takes 50-75% longer than a kprobe hit. | |
d27a4ddd JK |
660 | When you have a return probe set on a function, adding a kprobe at |
661 | the entry to that function adds essentially no overhead. | |
662 | ||
a1dac767 | 663 | Here are sample overhead figures (in usec) for different architectures:: |
d27a4ddd | 664 | |
9b17374e MH |
665 | k = kprobe; r = return probe; kr = kprobe + return probe |
666 | on same function | |
d27a4ddd | 667 | |
a1dac767 | 668 | i386: Intel Pentium M, 1495 MHz, 2957.31 bogomips |
9b17374e | 669 | k = 0.57 usec; r = 0.92; kr = 0.99 |
d27a4ddd | 670 | |
a1dac767 | 671 | x86_64: AMD Opteron 246, 1994 MHz, 3971.48 bogomips |
9b17374e | 672 | k = 0.49 usec; r = 0.80; kr = 0.82 |
d27a4ddd | 673 | |
a1dac767 | 674 | ppc64: POWER5 (gr), 1656 MHz (SMT disabled, 1 virtual CPU per physical CPU) |
9b17374e | 675 | k = 0.77 usec; r = 1.26; kr = 1.45 |
a1dac767 MCC |
676 | |
677 | Optimized Probe Overhead | |
678 | ------------------------ | |
b26486bf MH |
679 | |
680 | Typically, an optimized kprobe hit takes 0.07 to 0.1 microseconds to | |
a1dac767 MCC |
681 | process. Here are sample overhead figures (in usec) for x86 architectures:: |
682 | ||
683 | k = unoptimized kprobe, b = boosted (single-step skipped), o = optimized kprobe, | |
684 | r = unoptimized kretprobe, rb = boosted kretprobe, ro = optimized kretprobe. | |
b26486bf | 685 | |
a1dac767 MCC |
686 | i386: Intel(R) Xeon(R) E5410, 2.33GHz, 4656.90 bogomips |
687 | k = 0.80 usec; b = 0.33; o = 0.05; r = 1.10; rb = 0.61; ro = 0.33 | |
b26486bf | 688 | |
a1dac767 MCC |
689 | x86-64: Intel(R) Xeon(R) E5410, 2.33GHz, 4656.90 bogomips |
690 | k = 0.99 usec; b = 0.43; o = 0.06; r = 1.24; rb = 0.68; ro = 0.30 | |
b26486bf | 691 | |
a1dac767 MCC |
692 | TODO |
693 | ==== | |
d27a4ddd | 694 | |
8861da31 | 695 | a. SystemTap (http://sourceware.org/systemtap): Provides a simplified |
a1dac767 | 696 | programming interface for probe-based instrumentation. Try it out. |
8861da31 JK |
697 | b. Kernel return probes for sparc64. |
698 | c. Support for other architectures. | |
699 | d. User-space probes. | |
700 | e. Watchpoint probes (which fire on data references). | |
d27a4ddd | 701 | |
a1dac767 MCC |
702 | Kprobes Example |
703 | =============== | |
d27a4ddd | 704 | |
804defea | 705 | See samples/kprobes/kprobe_example.c |
d27a4ddd | 706 | |
a1dac767 MCC |
707 | Kretprobes Example |
708 | ================== | |
d27a4ddd | 709 | |
804defea | 710 | See samples/kprobes/kretprobe_example.c |
d27a4ddd JK |
711 | |
712 | For additional information on Kprobes, refer to the following URLs: | |
bf8f6e5b | 713 | |
a1dac767 MCC |
714 | - http://www-106.ibm.com/developerworks/library/l-kprobes.html?ca=dgr-lnxw42Kprobe |
715 | - http://www.redhat.com/magazine/005mar05/features/kprobes/ | |
716 | - http://www-users.cs.umn.edu/~boutcher/kprobes/ | |
717 | - http://www.linuxsymposium.org/2006/linuxsymposium_procv2.pdf (pages 101-115) | |
718 | ||
9b17374e MH |
719 | Deprecated Features |
720 | =================== | |
721 | ||
722 | Jprobes is now a deprecated feature. People who are depending on it should | |
723 | migrate to other tracing features or use older kernels. Please consider to | |
724 | migrate your tool to one of the following options: | |
725 | ||
726 | - Use trace-event to trace target function with arguments. | |
727 | ||
728 | trace-event is a low-overhead (and almost no visible overhead if it | |
729 | is off) statically defined event interface. You can define new events | |
730 | and trace it via ftrace or any other tracing tools. | |
731 | ||
732 | See the following urls: | |
733 | ||
734 | - https://lwn.net/Articles/379903/ | |
735 | - https://lwn.net/Articles/381064/ | |
736 | - https://lwn.net/Articles/383362/ | |
737 | ||
738 | - Use ftrace dynamic events (kprobe event) with perf-probe. | |
739 | ||
740 | If you build your kernel with debug info (CONFIG_DEBUG_INFO=y), you can | |
741 | find which register/stack is assigned to which local variable or arguments | |
742 | by using perf-probe and set up new event to trace it. | |
743 | ||
744 | See following documents: | |
745 | ||
5fb94e9c MCC |
746 | - Documentation/trace/kprobetrace.rst |
747 | - Documentation/trace/events.rst | |
9b17374e MH |
748 | - tools/perf/Documentation/perf-probe.txt |
749 | ||
a1dac767 MCC |
750 | |
751 | The kprobes debugfs interface | |
752 | ============================= | |
bf8f6e5b | 753 | |
bf8f6e5b AM |
754 | |
755 | With recent kernels (> 2.6.20) the list of registered kprobes is visible | |
156f5a78 | 756 | under the /sys/kernel/debug/kprobes/ directory (assuming debugfs is mounted at //sys/kernel/debug). |
bf8f6e5b | 757 | |
a1dac767 | 758 | /sys/kernel/debug/kprobes/list: Lists all registered probes on the system:: |
bf8f6e5b | 759 | |
a1dac767 | 760 | c015d71a k vfs_read+0x0 |
a1dac767 | 761 | c03dedc5 r tcp_v4_rcv+0x0 |
bf8f6e5b AM |
762 | |
763 | The first column provides the kernel address where the probe is inserted. | |
9b17374e MH |
764 | The second column identifies the type of probe (k - kprobe and r - kretprobe) |
765 | while the third column specifies the symbol+offset of the probe. | |
766 | If the probed function belongs to a module, the module name is also | |
767 | specified. Following columns show probe status. If the probe is on | |
e8386a0c MH |
768 | a virtual address that is no longer valid (module init sections, module |
769 | virtual addresses that correspond to modules that've been unloaded), | |
de5bd88d | 770 | such probes are marked with [GONE]. If the probe is temporarily disabled, |
b26486bf | 771 | such probes are marked with [DISABLED]. If the probe is optimized, it is |
9ed330d3 WL |
772 | marked with [OPTIMIZED]. If the probe is ftrace-based, it is marked with |
773 | [FTRACE]. | |
bf8f6e5b | 774 | |
156f5a78 | 775 | /sys/kernel/debug/kprobes/enabled: Turn kprobes ON/OFF forcibly. |
bf8f6e5b | 776 | |
de5bd88d MH |
777 | Provides a knob to globally and forcibly turn registered kprobes ON or OFF. |
778 | By default, all kprobes are enabled. By echoing "0" to this file, all | |
779 | registered probes will be disarmed, till such time a "1" is echoed to this | |
780 | file. Note that this knob just disarms and arms all kprobes and doesn't | |
781 | change each probe's disabling state. This means that disabled kprobes (marked | |
782 | [DISABLED]) will be not enabled if you turn ON all kprobes by this knob. | |
b26486bf MH |
783 | |
784 | ||
a1dac767 MCC |
785 | The kprobes sysctl interface |
786 | ============================ | |
b26486bf MH |
787 | |
788 | /proc/sys/debug/kprobes-optimization: Turn kprobes optimization ON/OFF. | |
789 | ||
790 | When CONFIG_OPTPROBES=y, this sysctl interface appears and it provides | |
791 | a knob to globally and forcibly turn jump optimization (see section | |
a1dac767 MCC |
792 | :ref:`kprobes_jump_optimization`) ON or OFF. By default, jump optimization |
793 | is allowed (ON). If you echo "0" to this file or set | |
794 | "debug.kprobes_optimization" to 0 via sysctl, all optimized probes will be | |
795 | unoptimized, and any new probes registered after that will not be optimized. | |
43e5f7e1 MCC |
796 | |
797 | Note that this knob *changes* the optimized state. This means that optimized | |
a1dac767 | 798 | probes (marked [OPTIMIZED]) will be unoptimized ([OPTIMIZED] tag will be |
b26486bf MH |
799 | removed). If the knob is turned on, they will be optimized again. |
800 |