1PTRACE(2) Linux Programmer's Manual PTRACE(2)
2
6 ptrace - process trace
7
9 #include <sys/ptrace.h>
10 long ptrace(enum __ptrace_request request, pid_t pid,
12 void *addr, void *data);
13
15 The ptrace() system call provides a means by which one process (the
16 "tracer") may observe and control the execution of another process (the
17 "tracee"), and examine and change the tracee's memory and registers.
18 It is primarily used to implement breakpoint debugging and system call
19 tracing.
20 A tracee first needs to be attached to the tracer. Attachment and sub‐
22 sequent commands are per thread: in a multithreaded process, every
23 thread can be individually attached to a (potentially different)
24 tracer, or left not attached and thus not debugged. Therefore,
25 "tracee" always means "(one) thread", never "a (possibly multithreaded)
26 process". Ptrace commands are always sent to a specific tracee using a
27 call of the form
28 ptrace(PTRACE_foo, pid, ...)
30 where pid is the thread ID of the corresponding Linux thread.
32 (Note that in this page, a "multithreaded process" means a thread group
34 consisting of threads created using the clone(2) CLONE_THREAD flag.)
35 A process can initiate a trace by calling fork(2) and having the
37 resulting child do a PTRACE_TRACEME, followed (typically) by an
38 execve(2). Alternatively, one process may commence tracing another
39 process using PTRACE_ATTACH or PTRACE_SEIZE.
40 While being traced, the tracee will stop each time a signal is deliv‐
42 ered, even if the signal is being ignored. (An exception is SIGKILL,
43 which has its usual effect.) The tracer will be notified at its next
44 call to waitpid(2) (or one of the related "wait" system calls); that
45 call will return a status value containing information that indicates
46 the cause of the stop in the tracee. While the tracee is stopped, the
47 tracer can use various ptrace requests to inspect and modify the
48 tracee. The tracer then causes the tracee to continue, optionally
49 ignoring the delivered signal (or even delivering a different signal
50 instead).
51 If the PTRACE_O_TRACEEXEC option is not in effect, all successful calls
53 to execve(2) by the traced process will cause it to be sent a SIGTRAP
54 signal, giving the parent a chance to gain control before the new pro‐
55 gram begins execution.
56 When the tracer is finished tracing, it can cause the tracee to con‐
58 tinue executing in a normal, untraced mode via PTRACE_DETACH.
59 The value of request determines the action to be performed:
61 PTRACE_TRACEME
63 Indicate that this process is to be traced by its parent. A
64 process probably shouldn't make this request if its parent isn't
65 expecting to trace it. (pid, addr, and data are ignored.)
66 The PTRACE_TRACEME request is used only by the tracee; the
68 remaining requests are used only by the tracer. In the follow‐
69 ing requests, pid specifies the thread ID of the tracee to be
70 acted on. For requests other than PTRACE_ATTACH, PTRACE_SEIZE,
71 PTRACE_INTERRUPT, and PTRACE_KILL, the tracee must be stopped.
72 PTRACE_PEEKTEXT, PTRACE_PEEKDATA
74 Read a word at the address addr in the tracee's memory, return‐
75 ing the word as the result of the ptrace() call. Linux does not
76 have separate text and data address spaces, so these two
77 requests are currently equivalent. (data is ignored; but see
78 NOTES.)
79 PTRACE_PEEKUSER
81 Read a word at offset addr in the tracee's USER area, which
82 holds the registers and other information about the process (see
83 <sys/user.h>). The word is returned as the result of the
84 ptrace() call. Typically, the offset must be word-aligned,
85 though this might vary by architecture. See NOTES. (data is
86 ignored; but see NOTES.)
87 PTRACE_POKETEXT, PTRACE_POKEDATA
89 Copy the word data to the address addr in the tracee's memory.
90 As for PTRACE_PEEKTEXT and PTRACE_PEEKDATA, these two requests
91 are currently equivalent.
92 PTRACE_POKEUSER
94 Copy the word data to offset addr in the tracee's USER area. As
95 for PTRACE_PEEKUSER, the offset must typically be word-aligned.
96 In order to maintain the integrity of the kernel, some modifica‐
97 tions to the USER area are disallowed.
98 PTRACE_GETREGS, PTRACE_GETFPREGS
100 Copy the tracee's general-purpose or floating-point registers,
101 respectively, to the address data in the tracer. See
102 <sys/user.h> for information on the format of this data. (addr
103 is ignored.) Note that SPARC systems have the meaning of data
104 and addr reversed; that is, data is ignored and the registers
105 are copied to the address addr. PTRACE_GETREGS and PTRACE_GETF‐
106 PREGS are not present on all architectures.
107 PTRACE_GETREGSET (since Linux 2.6.34)
109 Read the tracee's registers. addr specifies, in an architec‐
110 ture-dependent way, the type of registers to be read. NT_PRSTA‐
111 TUS (with numerical value 1) usually results in reading of gen‐
112 eral-purpose registers. If the CPU has, for example, floating-
113 point and/or vector registers, they can be retrieved by setting
114 addr to the corresponding NT_foo constant. data points to a
115 struct iovec, which describes the destination buffer's location
116 and length. On return, the kernel modifies iov.len to indicate
117 the actual number of bytes returned.
118 PTRACE_SETREGS, PTRACE_SETFPREGS
120 Modify the tracee's general-purpose or floating-point registers,
121 respectively, from the address data in the tracer. As for
122 PTRACE_POKEUSER, some general-purpose register modifications may
123 be disallowed. (addr is ignored.) Note that SPARC systems have
124 the meaning of data and addr reversed; that is, data is ignored
125 and the registers are copied from the address addr.
126 PTRACE_SETREGS and PTRACE_SETFPREGS are not present on all
127 architectures.
128 PTRACE_SETREGSET (since Linux 2.6.34)
130 Modify the tracee's registers. The meaning of addr and data is
131 analogous to PTRACE_GETREGSET.
132 PTRACE_GETSIGINFO (since Linux 2.3.99-pre6)
134 Retrieve information about the signal that caused the stop.
135 Copy a siginfo_t structure (see sigaction(2)) from the tracee to
136 the address data in the tracer. (addr is ignored.)
137 PTRACE_SETSIGINFO (since Linux 2.3.99-pre6)
139 Set signal information: copy a siginfo_t structure from the
140 address data in the tracer to the tracee. This will affect only
141 signals that would normally be delivered to the tracee and were
142 caught by the tracer. It may be difficult to tell these normal
143 signals from synthetic signals generated by ptrace() itself.
144 (addr is ignored.)
145 PTRACE_PEEKSIGINFO (since Linux 3.10)
147 Retrieve siginfo_t structures without removing signals from a
148 queue. addr points to a ptrace_peeksiginfo_args structure that
149 specifies the ordinal position from which copying of signals
150 should start, and the number of signals to copy. siginfo_t
151 structures are copied into the buffer pointed to by data. The
152 return value contains the number of copied signals (zero indi‐
153 cates that there is no signal corresponding to the specified
154 ordinal position). Within the returned siginfo structures, the
155 si_code field includes information (__SI_CHLD, __SI_FAULT, etc.)
156 that are not otherwise exposed to user space.
157 struct ptrace_peeksiginfo_args {
159 u64 off; /* Ordinal position in queue at which
160 to start copying signals */
161 u32 flags; /* PTRACE_PEEKSIGINFO_SHARED or 0 */
162 s32 nr; /* Number of signals to copy */
163 };
164 Currently, there is only one flag, PTRACE_PEEKSIGINFO_SHARED,
166 for dumping signals from the process-wide signal queue. If this
167 flag is not set, signals are read from the per-thread queue of
168 the specified thread.
169 PTRACE_GETSIGMASK (since Linux 3.11)
171 Place a copy of the mask of blocked signals (see sigprocmask(2))
172 in the buffer pointed to by data, which should be a pointer to a
173 buffer of type sigset_t. The addr argument contains the size of
174 the buffer pointed to by data (i.e., sizeof(sigset_t)).
175 PTRACE_SETSIGMASK (since Linux 3.11)
177 Change the mask of blocked signals (see sigprocmask(2)) to the
178 value specified in the buffer pointed to by data, which should
179 be a pointer to a buffer of type sigset_t. The addr argument
180 contains the size of the buffer pointed to by data (i.e.,
181 sizeof(sigset_t)).
182 PTRACE_SETOPTIONS (since Linux 2.4.6; see BUGS for caveats)
184 Set ptrace options from data. (addr is ignored.) data is
185 interpreted as a bit mask of options, which are specified by the
186 following flags:
187 PTRACE_O_EXITKILL (since Linux 3.8)
189 Send a SIGKILL signal to the tracee if the tracer exits.
190 This option is useful for ptrace jailers that want to
191 ensure that tracees can never escape the tracer's con‐
192 trol.
193 PTRACE_O_TRACECLONE (since Linux 2.5.46)
195 Stop the tracee at the next clone(2) and automatically
196 start tracing the newly cloned process, which will start
197 with a SIGSTOP, or PTRACE_EVENT_STOP if PTRACE_SEIZE was
198 used. A waitpid(2) by the tracer will return a status
199 value such that
200 status>>8 == (SIGTRAP | (PTRACE_EVENT_CLONE<<8))
202 The PID of the new process can be retrieved with
204 PTRACE_GETEVENTMSG.
205 This option may not catch clone(2) calls in all cases.
207 If the tracee calls clone(2) with the CLONE_VFORK flag,
208 PTRACE_EVENT_VFORK will be delivered instead if
209 PTRACE_O_TRACEVFORK is set; otherwise if the tracee calls
210 clone(2) with the exit signal set to SIGCHLD,
211 PTRACE_EVENT_FORK will be delivered if PTRACE_O_TRACEFORK
212 is set.
213 PTRACE_O_TRACEEXEC (since Linux 2.5.46)
215 Stop the tracee at the next execve(2). A waitpid(2) by
216 the tracer will return a status value such that
217 status>>8 == (SIGTRAP | (PTRACE_EVENT_EXEC<<8))
219 If the execing thread is not a thread group leader, the
221 thread ID is reset to thread group leader's ID before
222 this stop. Since Linux 3.0, the former thread ID can be
223 retrieved with PTRACE_GETEVENTMSG.
224 PTRACE_O_TRACEEXIT (since Linux 2.5.60)
226 Stop the tracee at exit. A waitpid(2) by the tracer will
227 return a status value such that
228 status>>8 == (SIGTRAP | (PTRACE_EVENT_EXIT<<8))
230 The tracee's exit status can be retrieved with
232 PTRACE_GETEVENTMSG.
233 The tracee is stopped early during process exit, when
235 registers are still available, allowing the tracer to see
236 where the exit occurred, whereas the normal exit notifi‐
237 cation is done after the process is finished exiting.
238 Even though context is available, the tracer cannot pre‐
239 vent the exit from happening at this point.
240 PTRACE_O_TRACEFORK (since Linux 2.5.46)
242 Stop the tracee at the next fork(2) and automatically
243 start tracing the newly forked process, which will start
244 with a SIGSTOP, or PTRACE_EVENT_STOP if PTRACE_SEIZE was
245 used. A waitpid(2) by the tracer will return a status
246 value such that
247 status>>8 == (SIGTRAP | (PTRACE_EVENT_FORK<<8))
249 The PID of the new process can be retrieved with
251 PTRACE_GETEVENTMSG.
252 PTRACE_O_TRACESYSGOOD (since Linux 2.4.6)
254 When delivering system call traps, set bit 7 in the sig‐
255 nal number (i.e., deliver SIGTRAP|0x80). This makes it
256 easy for the tracer to distinguish normal traps from
257 those caused by a system call.
258 PTRACE_O_TRACEVFORK (since Linux 2.5.46)
260 Stop the tracee at the next vfork(2) and automatically
261 start tracing the newly vforked process, which will start
262 with a SIGSTOP, or PTRACE_EVENT_STOP if PTRACE_SEIZE was
263 used. A waitpid(2) by the tracer will return a status
264 value such that
265 status>>8 == (SIGTRAP | (PTRACE_EVENT_VFORK<<8))
267 The PID of the new process can be retrieved with
269 PTRACE_GETEVENTMSG.
270 PTRACE_O_TRACEVFORKDONE (since Linux 2.5.60)
272 Stop the tracee at the completion of the next vfork(2).
273 A waitpid(2) by the tracer will return a status value
274 such that
275 status>>8 == (SIGTRAP | (PTRACE_EVENT_VFORK_DONE<<8))
277 The PID of the new process can (since Linux 2.6.18) be
279 retrieved with PTRACE_GETEVENTMSG.
280 PTRACE_O_TRACESECCOMP (since Linux 3.5)
282 Stop the tracee when a seccomp(2) SECCOMP_RET_TRACE rule
283 is triggered. A waitpid(2) by the tracer will return a
284 status value such that
285 status>>8 == (SIGTRAP | (PTRACE_EVENT_SECCOMP<<8))
287 While this triggers a PTRACE_EVENT stop, it is similar to
289 a syscall-enter-stop. For details, see the note on
290 PTRACE_EVENT_SECCOMP below. The seccomp event message
291 data (from the SECCOMP_RET_DATA portion of the seccomp
292 filter rule) can be retrieved with PTRACE_GETEVENTMSG.
293 PTRACE_O_SUSPEND_SECCOMP (since Linux 4.3)
295 Suspend the tracee's seccomp protections. This applies
296 regardless of mode, and can be used when the tracee has
297 not yet installed seccomp filters. That is, a valid use
298 case is to suspend a tracee's seccomp protections before
299 they are installed by the tracee, let the tracee install
300 the filters, and then clear this flag when the filters
301 should be resumed. Setting this option requires that the
302 tracer have the CAP_SYS_ADMIN capability, not have any
303 seccomp protections installed, and not have PTRACE_O_SUS‐
304 PEND_SECCOMP set on itself.
305 PTRACE_GETEVENTMSG (since Linux 2.5.46)
307 Retrieve a message (as an unsigned long) about the ptrace event
308 that just happened, placing it at the address data in the
309 tracer. For PTRACE_EVENT_EXIT, this is the tracee's exit sta‐
310 tus. For PTRACE_EVENT_FORK, PTRACE_EVENT_VFORK,
311 PTRACE_EVENT_VFORK_DONE, and PTRACE_EVENT_CLONE, this is the PID
312 of the new process. For PTRACE_EVENT_SECCOMP, this is the sec‐
313 comp(2) filter's SECCOMP_RET_DATA associated with the triggered
314 rule. (addr is ignored.)
315 PTRACE_CONT
317 Restart the stopped tracee process. If data is nonzero, it is
318 interpreted as the number of a signal to be delivered to the
319 tracee; otherwise, no signal is delivered. Thus, for example,
320 the tracer can control whether a signal sent to the tracee is
321 delivered or not. (addr is ignored.)
322 PTRACE_SYSCALL, PTRACE_SINGLESTEP
324 Restart the stopped tracee as for PTRACE_CONT, but arrange for
325 the tracee to be stopped at the next entry to or exit from a
326 system call, or after execution of a single instruction, respec‐
327 tively. (The tracee will also, as usual, be stopped upon
328 receipt of a signal.) From the tracer's perspective, the tracee
329 will appear to have been stopped by receipt of a SIGTRAP. So,
330 for PTRACE_SYSCALL, for example, the idea is to inspect the
331 arguments to the system call at the first stop, then do another
332 PTRACE_SYSCALL and inspect the return value of the system call
333 at the second stop. The data argument is treated as for
334 PTRACE_CONT. (addr is ignored.)
335 PTRACE_SYSEMU, PTRACE_SYSEMU_SINGLESTEP (since Linux 2.6.14)
337 For PTRACE_SYSEMU, continue and stop on entry to the next system
338 call, which will not be executed. See the documentation on
339 syscall-stops below. For PTRACE_SYSEMU_SINGLESTEP, do the same
340 but also singlestep if not a system call. This call is used by
341 programs like User Mode Linux that want to emulate all the
342 tracee's system calls. The data argument is treated as for
343 PTRACE_CONT. The addr argument is ignored. These requests are
344 currently supported only on x86.
345 PTRACE_LISTEN (since Linux 3.4)
347 Restart the stopped tracee, but prevent it from executing. The
348 resulting state of the tracee is similar to a process which has
349 been stopped by a SIGSTOP (or other stopping signal). See the
350 "group-stop" subsection for additional information. PTRACE_LIS‐
351 TEN works only on tracees attached by PTRACE_SEIZE.
352 PTRACE_KILL
354 Send the tracee a SIGKILL to terminate it. (addr and data are
355 ignored.)
356 This operation is deprecated; do not use it! Instead, send a
358 SIGKILL directly using kill(2) or tgkill(2). The problem with
359 PTRACE_KILL is that it requires the tracee to be in signal-
360 delivery-stop, otherwise it may not work (i.e., may complete
361 successfully but won't kill the tracee). By contrast, sending a
362 SIGKILL directly has no such limitation.
363 PTRACE_INTERRUPT (since Linux 3.4)
365 Stop a tracee. If the tracee is running or sleeping in kernel
366 space and PTRACE_SYSCALL is in effect, the system call is inter‐
367 rupted and syscall-exit-stop is reported. (The interrupted sys‐
368 tem call is restarted when the tracee is restarted.) If the
369 tracee was already stopped by a signal and PTRACE_LISTEN was
370 sent to it, the tracee stops with PTRACE_EVENT_STOP and WSTOP‐
371 SIG(status) returns the stop signal. If any other ptrace-stop
372 is generated at the same time (for example, if a signal is sent
373 to the tracee), this ptrace-stop happens. If none of the above
374 applies (for example, if the tracee is running in user space),
375 it stops with PTRACE_EVENT_STOP with WSTOPSIG(status) == SIG‐
376 TRAP. PTRACE_INTERRUPT only works on tracees attached by
377 PTRACE_SEIZE.
378 PTRACE_ATTACH
380 Attach to the process specified in pid, making it a tracee of
381 the calling process. The tracee is sent a SIGSTOP, but will not
382 necessarily have stopped by the completion of this call; use
383 waitpid(2) to wait for the tracee to stop. See the "Attaching
384 and detaching" subsection for additional information. (addr and
385 data are ignored.)
386 Permission to perform a PTRACE_ATTACH is governed by a ptrace
388 access mode PTRACE_MODE_ATTACH_REALCREDS check; see below.
389 PTRACE_SEIZE (since Linux 3.4)
391 Attach to the process specified in pid, making it a tracee of
392 the calling process. Unlike PTRACE_ATTACH, PTRACE_SEIZE does
393 not stop the process. Group-stops are reported as
394 PTRACE_EVENT_STOP and WSTOPSIG(status) returns the stop signal.
395 Automatically attached children stop with PTRACE_EVENT_STOP and
396 WSTOPSIG(status) returns SIGTRAP instead of having SIGSTOP sig‐
397 nal delivered to them. execve(2) does not deliver an extra SIG‐
398 TRAP. Only a PTRACE_SEIZEd process can accept PTRACE_INTERRUPT
399 and PTRACE_LISTEN commands. The "seized" behavior just
400 described is inherited by children that are automatically
401 attached using PTRACE_O_TRACEFORK, PTRACE_O_TRACEVFORK, and
402 PTRACE_O_TRACECLONE. addr must be zero. data contains a bit
403 mask of ptrace options to activate immediately.
404 Permission to perform a PTRACE_SEIZE is governed by a ptrace
406 access mode PTRACE_MODE_ATTACH_REALCREDS check; see below.
407 PTRACE_SECCOMP_GET_FILTER (since Linux 4.4)
409 This operation allows the tracer to dump the tracee's classic
410 BPF filters.
411 addr is an integer specifying the index of the filter to be
413 dumped. The most recently installed filter has the index 0. If
414 addr is greater than the number of installed filters, the opera‐
415 tion fails with the error ENOENT.
416 data is either a pointer to a struct sock_filter array that is
418 large enough to store the BPF program, or NULL if the program is
419 not to be stored.
420 Upon success, the return value is the number of instructions in
422 the BPF program. If data was NULL, then this return value can
423 be used to correctly size the struct sock_filter array passed in
424 a subsequent call.
425 This operation fails with the error EACCES if the caller does
427 not have the CAP_SYS_ADMIN capability or if the caller is in
428 strict or filter seccomp mode. If the filter referred to by
429 addr is not a classic BPF filter, the operation fails with the
430 error EMEDIUMTYPE.
431 This operation is available if the kernel was configured with
433 both the CONFIG_SECCOMP_FILTER and the CONFIG_CHECKPOINT_RESTORE
434 options.
435 PTRACE_DETACH
437 Restart the stopped tracee as for PTRACE_CONT, but first detach
438 from it. Under Linux, a tracee can be detached in this way
439 regardless of which method was used to initiate tracing. (addr
440 is ignored.)
441 PTRACE_GET_THREAD_AREA (since Linux 2.6.0)
443 This operation performs a similar task to get_thread_area(2).
444 It reads the TLS entry in the GDT whose index is given in addr,
445 placing a copy of the entry into the struct user_desc pointed to
446 by data. (By contrast with get_thread_area(2), the entry_number
447 of the struct user_desc is ignored.)
448 PTRACE_SET_THREAD_AREA (since Linux 2.6.0)
450 This operation performs a similar task to set_thread_area(2).
451 It sets the TLS entry in the GDT whose index is given in addr,
452 assigning it the data supplied in the struct user_desc pointed
453 to by data. (By contrast with set_thread_area(2), the
454 entry_number of the struct user_desc is ignored; in other words,
455 this ptrace operation can't be used to allocate a free TLS
456 entry.)
457 Death under ptrace
459 When a (possibly multithreaded) process receives a killing signal (one
460 whose disposition is set to SIG_DFL and whose default action is to kill
461 the process), all threads exit. Tracees report their death to their
462 tracer(s). Notification of this event is delivered via waitpid(2).
463 Note that the killing signal will first cause signal-delivery-stop (on
465 one tracee only), and only after it is injected by the tracer (or after
466 it was dispatched to a thread which isn't traced), will death from the
467 signal happen on all tracees within a multithreaded process. (The term
468 "signal-delivery-stop" is explained below.)
469 SIGKILL does not generate signal-delivery-stop and therefore the tracer
471 can't suppress it. SIGKILL kills even within system calls (syscall-
472 exit-stop is not generated prior to death by SIGKILL). The net effect
473 is that SIGKILL always kills the process (all its threads), even if
474 some threads of the process are ptraced.
475 When the tracee calls _exit(2), it reports its death to its tracer.
477 Other threads are not affected.
478 When any thread executes exit_group(2), every tracee in its thread
480 group reports its death to its tracer.
481 If the PTRACE_O_TRACEEXIT option is on, PTRACE_EVENT_EXIT will happen
483 before actual death. This applies to exits via exit(2), exit_group(2),
484 and signal deaths (except SIGKILL, depending on the kernel version; see
485 BUGS below), and when threads are torn down on execve(2) in a multi‐
486 threaded process.
487 The tracer cannot assume that the ptrace-stopped tracee exists. There
489 are many scenarios when the tracee may die while stopped (such as
490 SIGKILL). Therefore, the tracer must be prepared to handle an ESRCH
491 error on any ptrace operation. Unfortunately, the same error is
492 returned if the tracee exists but is not ptrace-stopped (for commands
493 which require a stopped tracee), or if it is not traced by the process
494 which issued the ptrace call. The tracer needs to keep track of the
495 stopped/running state of the tracee, and interpret ESRCH as "tracee
496 died unexpectedly" only if it knows that the tracee has been observed
497 to enter ptrace-stop. Note that there is no guarantee that wait‐
498 pid(WNOHANG) will reliably report the tracee's death status if a ptrace
499 operation returned ESRCH. waitpid(WNOHANG) may return 0 instead. In
500 other words, the tracee may be "not yet fully dead", but already refus‐
501 ing ptrace requests.
502 The tracer can't assume that the tracee always ends its life by report‐
504 ing WIFEXITED(status) or WIFSIGNALED(status); there are cases where
505 this does not occur. For example, if a thread other than thread group
506 leader does an execve(2), it disappears; its PID will never be seen
507 again, and any subsequent ptrace stops will be reported under the
508 thread group leader's PID.
509 Stopped states
511 A tracee can be in two states: running or stopped. For the purposes of
512 ptrace, a tracee which is blocked in a system call (such as read(2),
513 pause(2), etc.) is nevertheless considered to be running, even if the
514 tracee is blocked for a long time. The state of the tracee after
515 PTRACE_LISTEN is somewhat of a gray area: it is not in any ptrace-stop
516 (ptrace commands won't work on it, and it will deliver waitpid(2) noti‐
517 fications), but it also may be considered "stopped" because it is not
518 executing instructions (is not scheduled), and if it was in group-stop
519 before PTRACE_LISTEN, it will not respond to signals until SIGCONT is
520 received.
521 There are many kinds of states when the tracee is stopped, and in
523 ptrace discussions they are often conflated. Therefore, it is impor‐
524 tant to use precise terms.
525 In this manual page, any stopped state in which the tracee is ready to
527 accept ptrace commands from the tracer is called ptrace-stop. Ptrace-
528 stops can be further subdivided into signal-delivery-stop, group-stop,
529 syscall-stop, PTRACE_EVENT stops, and so on. These stopped states are
530 described in detail below.
531 When the running tracee enters ptrace-stop, it notifies its tracer
533 using waitpid(2) (or one of the other "wait" system calls). Most of
534 this manual page assumes that the tracer waits with:
535 pid = waitpid(pid_or_minus_1, &status, __WALL);
537 Ptrace-stopped tracees are reported as returns with pid greater than 0
539 and WIFSTOPPED(status) true.
540 The __WALL flag does not include the WSTOPPED and WEXITED flags, but
542 implies their functionality.
543 Setting the WCONTINUED flag when calling waitpid(2) is not recommended:
545 the "continued" state is per-process and consuming it can confuse the
546 real parent of the tracee.
547 Use of the WNOHANG flag may cause waitpid(2) to return 0 ("no wait
549 results available yet") even if the tracer knows there should be a
550 notification. Example:
551 errno = 0;
553 ptrace(PTRACE_CONT, pid, 0L, 0L);
554 if (errno == ESRCH) {
555 /* tracee is dead */
556 r = waitpid(tracee, &status, __WALL | WNOHANG);
557 /* r can still be 0 here! */
558 }
559 The following kinds of ptrace-stops exist: signal-delivery-stops,
561 group-stops, PTRACE_EVENT stops, syscall-stops. They all are reported
562 by waitpid(2) with WIFSTOPPED(status) true. They may be differentiated
563 by examining the value status>>8, and if there is ambiguity in that
564 value, by querying PTRACE_GETSIGINFO. (Note: the WSTOPSIG(status)
565 macro can't be used to perform this examination, because it returns the
566 value (status>>8) & 0xff.)
567 Signal-delivery-stop
569 When a (possibly multithreaded) process receives any signal except
570 SIGKILL, the kernel selects an arbitrary thread which handles the sig‐
571 nal. (If the signal is generated with tgkill(2), the target thread can
572 be explicitly selected by the caller.) If the selected thread is
573 traced, it enters signal-delivery-stop. At this point, the signal is
574 not yet delivered to the process, and can be suppressed by the tracer.
575 If the tracer doesn't suppress the signal, it passes the signal to the
576 tracee in the next ptrace restart request. This second step of signal
577 delivery is called signal injection in this manual page. Note that if
578 the signal is blocked, signal-delivery-stop doesn't happen until the
579 signal is unblocked, with the usual exception that SIGSTOP can't be
580 blocked.
581 Signal-delivery-stop is observed by the tracer as waitpid(2) returning
583 with WIFSTOPPED(status) true, with the signal returned by WSTOPSIG(sta‐
584 tus). If the signal is SIGTRAP, this may be a different kind of
585 ptrace-stop; see the "Syscall-stops" and "execve" sections below for
586 details. If WSTOPSIG(status) returns a stopping signal, this may be a
587 group-stop; see below.
588 Signal injection and suppression
590 After signal-delivery-stop is observed by the tracer, the tracer should
591 restart the tracee with the call
592 ptrace(PTRACE_restart, pid, 0, sig)
594 where PTRACE_restart is one of the restarting ptrace requests. If sig
596 is 0, then a signal is not delivered. Otherwise, the signal sig is
597 delivered. This operation is called signal injection in this manual
598 page, to distinguish it from signal-delivery-stop.
599 The sig value may be different from the WSTOPSIG(status) value: the
601 tracer can cause a different signal to be injected.
602 Note that a suppressed signal still causes system calls to return pre‐
604 maturely. In this case, system calls will be restarted: the tracer
605 will observe the tracee to reexecute the interrupted system call (or
606 restart_syscall(2) system call for a few system calls which use a dif‐
607 ferent mechanism for restarting) if the tracer uses PTRACE_SYSCALL.
608 Even system calls (such as poll(2)) which are not restartable after
609 signal are restarted after signal is suppressed; however, kernel bugs
610 exist which cause some system calls to fail with EINTR even though no
611 observable signal is injected to the tracee.
612 Restarting ptrace commands issued in ptrace-stops other than signal-
614 delivery-stop are not guaranteed to inject a signal, even if sig is
615 nonzero. No error is reported; a nonzero sig may simply be ignored.
616 Ptrace users should not try to "create a new signal" this way: use
617 tgkill(2) instead.
618 The fact that signal injection requests may be ignored when restarting
620 the tracee after ptrace stops that are not signal-delivery-stops is a
621 cause of confusion among ptrace users. One typical scenario is that
622 the tracer observes group-stop, mistakes it for signal-delivery-stop,
623 restarts the tracee with
624 ptrace(PTRACE_restart, pid, 0, stopsig)
626 with the intention of injecting stopsig, but stopsig gets ignored and
628 the tracee continues to run.
629 The SIGCONT signal has a side effect of waking up (all threads of) a
631 group-stopped process. This side effect happens before signal-deliv‐
632 ery-stop. The tracer can't suppress this side effect (it can only sup‐
633 press signal injection, which only causes the SIGCONT handler to not be
634 executed in the tracee, if such a handler is installed). In fact, wak‐
635 ing up from group-stop may be followed by signal-delivery-stop for sig‐
636 nal(s) other than SIGCONT, if they were pending when SIGCONT was deliv‐
637 ered. In other words, SIGCONT may be not the first signal observed by
638 the tracee after it was sent.
639 Stopping signals cause (all threads of) a process to enter group-stop.
641 This side effect happens after signal injection, and therefore can be
642 suppressed by the tracer.
643 In Linux 2.4 and earlier, the SIGSTOP signal can't be injected.
645 PTRACE_GETSIGINFO can be used to retrieve a siginfo_t structure which
647 corresponds to the delivered signal. PTRACE_SETSIGINFO may be used to
648 modify it. If PTRACE_SETSIGINFO has been used to alter siginfo_t, the
649 si_signo field and the sig parameter in the restarting command must
650 match, otherwise the result is undefined.
651 Group-stop
653 When a (possibly multithreaded) process receives a stopping signal, all
654 threads stop. If some threads are traced, they enter a group-stop.
655 Note that the stopping signal will first cause signal-delivery-stop (on
656 one tracee only), and only after it is injected by the tracer (or after
657 it was dispatched to a thread which isn't traced), will group-stop be
658 initiated on all tracees within the multithreaded process. As usual,
659 every tracee reports its group-stop separately to the corresponding
660 tracer.
661 Group-stop is observed by the tracer as waitpid(2) returning with WIF‐
663 STOPPED(status) true, with the stopping signal available via WSTOP‐
664 SIG(status). The same result is returned by some other classes of
665 ptrace-stops, therefore the recommended practice is to perform the call
666 ptrace(PTRACE_GETSIGINFO, pid, 0, &siginfo)
668 The call can be avoided if the signal is not SIGSTOP, SIGTSTP, SIGTTIN,
670 or SIGTTOU; only these four signals are stopping signals. If the
671 tracer sees something else, it can't be a group-stop. Otherwise, the
672 tracer needs to call PTRACE_GETSIGINFO. If PTRACE_GETSIGINFO fails
673 with EINVAL, then it is definitely a group-stop. (Other failure codes
674 are possible, such as ESRCH ("no such process") if a SIGKILL killed the
675 tracee.)
676 If tracee was attached using PTRACE_SEIZE, group-stop is indicated by
678 PTRACE_EVENT_STOP: status>>16 == PTRACE_EVENT_STOP. This allows detec‐
679 tion of group-stops without requiring an extra PTRACE_GETSIGINFO call.
680 As of Linux 2.6.38, after the tracer sees the tracee ptrace-stop and
682 until it restarts or kills it, the tracee will not run, and will not
683 send notifications (except SIGKILL death) to the tracer, even if the
684 tracer enters into another waitpid(2) call.
685 The kernel behavior described in the previous paragraph causes a prob‐
687 lem with transparent handling of stopping signals. If the tracer
688 restarts the tracee after group-stop, the stopping signal is effec‐
689 tively ignored—the tracee doesn't remain stopped, it runs. If the
690 tracer doesn't restart the tracee before entering into the next wait‐
691 pid(2), future SIGCONT signals will not be reported to the tracer; this
692 would cause the SIGCONT signals to have no effect on the tracee.
693 Since Linux 3.4, there is a method to overcome this problem: instead of
695 PTRACE_CONT, a PTRACE_LISTEN command can be used to restart a tracee in
696 a way where it does not execute, but waits for a new event which it can
697 report via waitpid(2) (such as when it is restarted by a SIGCONT).
698 PTRACE_EVENT stops
700 If the tracer sets PTRACE_O_TRACE_* options, the tracee will enter
701 ptrace-stops called PTRACE_EVENT stops.
702 PTRACE_EVENT stops are observed by the tracer as waitpid(2) returning
704 with WIFSTOPPED(status), and WSTOPSIG(status) returns SIGTRAP. An
705 additional bit is set in the higher byte of the status word: the value
706 status>>8 will be
707 (SIGTRAP | PTRACE_EVENT_foo << 8).
709 The following events exist:
711 PTRACE_EVENT_VFORK
713 Stop before return from vfork(2) or clone(2) with the
714 CLONE_VFORK flag. When the tracee is continued after this stop,
715 it will wait for child to exit/exec before continuing its execu‐
716 tion (in other words, the usual behavior on vfork(2)).
717 PTRACE_EVENT_FORK
719 Stop before return from fork(2) or clone(2) with the exit signal
720 set to SIGCHLD.
721 PTRACE_EVENT_CLONE
723 Stop before return from clone(2).
724 PTRACE_EVENT_VFORK_DONE
726 Stop before return from vfork(2) or clone(2) with the
727 CLONE_VFORK flag, but after the child unblocked this tracee by
728 exiting or execing.
729 For all four stops described above, the stop occurs in the parent
731 (i.e., the tracee), not in the newly created thread.
732 PTRACE_GETEVENTMSG can be used to retrieve the new thread's ID.
733 PTRACE_EVENT_EXEC
735 Stop before return from execve(2). Since Linux 3.0,
736 PTRACE_GETEVENTMSG returns the former thread ID.
737 PTRACE_EVENT_EXIT
739 Stop before exit (including death from exit_group(2)), signal
740 death, or exit caused by execve(2) in a multithreaded process.
741 PTRACE_GETEVENTMSG returns the exit status. Registers can be
742 examined (unlike when "real" exit happens). The tracee is still
743 alive; it needs to be PTRACE_CONTed or PTRACE_DETACHed to finish
744 exiting.
745 PTRACE_EVENT_STOP
747 Stop induced by PTRACE_INTERRUPT command, or group-stop, or ini‐
748 tial ptrace-stop when a new child is attached (only if attached
749 using PTRACE_SEIZE).
750 PTRACE_EVENT_SECCOMP
752 Stop triggered by a seccomp(2) rule on tracee syscall entry when
753 PTRACE_O_TRACESECCOMP has been set by the tracer. The seccomp
754 event message data (from the SECCOMP_RET_DATA portion of the
755 seccomp filter rule) can be retrieved with PTRACE_GETEVENTMSG.
756 The semantics of this stop are described in detail in a separate
757 section below.
758 PTRACE_GETSIGINFO on PTRACE_EVENT stops returns SIGTRAP in si_signo,
760 with si_code set to (event<<8) | SIGTRAP.
761 Syscall-stops
763 If the tracee was restarted by PTRACE_SYSCALL or PTRACE_SYSEMU, the
764 tracee enters syscall-enter-stop just prior to entering any system call
765 (which will not be executed if the restart was using PTRACE_SYSEMU,
766 regardless of any change made to registers at this point or how the
767 tracee is restarted after this stop). No matter which method caused
768 the syscall-entry-stop, if the tracer restarts the tracee with
769 PTRACE_SYSCALL, the tracee enters syscall-exit-stop when the system
770 call is finished, or if it is interrupted by a signal. (That is, sig‐
771 nal-delivery-stop never happens between syscall-enter-stop and syscall-
772 exit-stop; it happens after syscall-exit-stop.). If the tracee is con‐
773 tinued using any other method (including PTRACE_SYSEMU), no syscall-
774 exit-stop occurs. Note that all mentions PTRACE_SYSEMU apply equally
775 to PTRACE_SYSEMU_SINGLESTEP.
776 However, even if the tracee was continued using PTRACE_SYSCALL, it is
778 not guaranteed that the next stop will be a syscall-exit-stop. Other
779 possibilities are that the tracee may stop in a PTRACE_EVENT stop
780 (including seccomp stops), exit (if it entered _exit(2) or
781 exit_group(2)), be killed by SIGKILL, or die silently (if it is a
782 thread group leader, the execve(2) happened in another thread, and that
783 thread is not traced by the same tracer; this situation is discussed
784 later).
785 Syscall-enter-stop and syscall-exit-stop are observed by the tracer as
787 waitpid(2) returning with WIFSTOPPED(status) true, and WSTOPSIG(status)
788 giving SIGTRAP. If the PTRACE_O_TRACESYSGOOD option was set by the
789 tracer, then WSTOPSIG(status) will give the value (SIGTRAP | 0x80).
790 Syscall-stops can be distinguished from signal-delivery-stop with SIG‐
792 TRAP by querying PTRACE_GETSIGINFO for the following cases:
793 si_code <= 0
795 SIGTRAP was delivered as a result of a user-space action, for
796 example, a system call (tgkill(2), kill(2), sigqueue(3), etc.),
797 expiration of a POSIX timer, change of state on a POSIX message
798 queue, or completion of an asynchronous I/O request.
799 si_code == SI_KERNEL (0x80)
801 SIGTRAP was sent by the kernel.
802 si_code == SIGTRAP or si_code == (SIGTRAP|0x80)
804 This is a syscall-stop.
805 However, syscall-stops happen very often (twice per system call), and
807 performing PTRACE_GETSIGINFO for every syscall-stop may be somewhat
808 expensive.
809 Some architectures allow the cases to be distinguished by examining
811 registers. For example, on x86, rax == -ENOSYS in syscall-enter-stop.
812 Since SIGTRAP (like any other signal) always happens after syscall-
813 exit-stop, and at this point rax almost never contains -ENOSYS, the
814 SIGTRAP looks like "syscall-stop which is not syscall-enter-stop"; in
815 other words, it looks like a "stray syscall-exit-stop" and can be
816 detected this way. But such detection is fragile and is best avoided.
817 Using the PTRACE_O_TRACESYSGOOD option is the recommended method to
819 distinguish syscall-stops from other kinds of ptrace-stops, since it is
820 reliable and does not incur a performance penalty.
821 Syscall-enter-stop and syscall-exit-stop are indistinguishable from
823 each other by the tracer. The tracer needs to keep track of the
824 sequence of ptrace-stops in order to not misinterpret syscall-enter-
825 stop as syscall-exit-stop or vice versa. In general, a syscall-enter-
826 stop is always followed by syscall-exit-stop, PTRACE_EVENT stop, or the
827 tracee's death; no other kinds of ptrace-stop can occur in between.
828 However, note that seccomp stops (see below) can cause syscall-exit-
829 stops, without preceding syscall-entry-stops. If seccomp is in use,
830 care needs to be taken not to misinterpret such stops as syscall-entry-
831 stops.
832 If after syscall-enter-stop, the tracer uses a restarting command other
834 than PTRACE_SYSCALL, syscall-exit-stop is not generated.
835 PTRACE_GETSIGINFO on syscall-stops returns SIGTRAP in si_signo, with
837 si_code set to SIGTRAP or (SIGTRAP|0x80).
838 PTRACE_EVENT_SECCOMP stops (Linux 3.5 to 4.7)
840 The behavior of PTRACE_EVENT_SECCOMP stops and their interaction with
841 other kinds of ptrace stops has changed between kernel versions. This
842 documents the behavior from their introduction until Linux 4.7 (inclu‐
843 sive). The behavior in later kernel versions is documented in the next
844 section.
845 A PTRACE_EVENT_SECCOMP stop occurs whenever a SECCOMP_RET_TRACE rule is
847 triggered. This is independent of which methods was used to restart
848 the system call. Notably, seccomp still runs even if the tracee was
849 restarted using PTRACE_SYSEMU and this system call is unconditionally
850 skipped.
851 Restarts from this stop will behave as if the stop had occurred right
853 before the system call in question. In particular, both PTRACE_SYSCALL
854 and PTRACE_SYSEMU will normally cause a subsequent syscall-entry-stop.
855 However, if after the PTRACE_EVENT_SECCOMP the system call number is
856 negative, both the syscall-entry-stop and the system call itself will
857 be skipped. This means that if the system call number is negative
858 after a PTRACE_EVENT_SECCOMP and the tracee is restarted using
859 PTRACE_SYSCALL, the next observed stop will be a syscall-exit-stop,
860 rather than the syscall-entry-stop that might have been expected.
861 PTRACE_EVENT_SECCOMP stops (since Linux 4.8)
863 Starting with Linux 4.8, the PTRACE_EVENT_SECCOMP stop was reordered to
864 occur between syscall-entry-stop and syscall-exit-stop. Note that sec‐
865 comp no longer runs (and no PTRACE_EVENT_SECCOMP will be reported) if
866 the system call is skipped due to PTRACE_SYSEMU.
867 Functionally, a PTRACE_EVENT_SECCOMP stop functions comparably to a
869 syscall-entry-stop (i.e., continuations using PTRACE_SYSCALL will cause
870 syscall-exit-stops, the system call number may be changed and any other
871 modified registers are visible to the to-be-executed system call as
872 well). Note that there may be, but need not have been a preceding
873 syscall-entry-stop.
874 After a PTRACE_EVENT_SECCOMP stop, seccomp will be rerun, with a SEC‐
876 COMP_RET_TRACE rule now functioning the same as a SECCOMP_RET_ALLOW.
877 Specifically, this means that if registers are not modified during the
878 PTRACE_EVENT_SECCOMP stop, the system call will then be allowed.
879 PTRACE_SINGLESTEP stops
881 [Details of these kinds of stops are yet to be documented.]
882 Informational and restarting ptrace commands
884 Most ptrace commands (all except PTRACE_ATTACH, PTRACE_SEIZE,
885 PTRACE_TRACEME, PTRACE_INTERRUPT, and PTRACE_KILL) require the tracee
886 to be in a ptrace-stop, otherwise they fail with ESRCH.
887 When the tracee is in ptrace-stop, the tracer can read and write data
889 to the tracee using informational commands. These commands leave the
890 tracee in ptrace-stopped state:
891 ptrace(PTRACE_PEEKTEXT/PEEKDATA/PEEKUSER, pid, addr, 0);
893 ptrace(PTRACE_POKETEXT/POKEDATA/POKEUSER, pid, addr, long_val);
894 ptrace(PTRACE_GETREGS/GETFPREGS, pid, 0, &struct);
895 ptrace(PTRACE_SETREGS/SETFPREGS, pid, 0, &struct);
896 ptrace(PTRACE_GETREGSET, pid, NT_foo, &iov);
897 ptrace(PTRACE_SETREGSET, pid, NT_foo, &iov);
898 ptrace(PTRACE_GETSIGINFO, pid, 0, &siginfo);
899 ptrace(PTRACE_SETSIGINFO, pid, 0, &siginfo);
900 ptrace(PTRACE_GETEVENTMSG, pid, 0, &long_var);
901 ptrace(PTRACE_SETOPTIONS, pid, 0, PTRACE_O_flags);
902 Note that some errors are not reported. For example, setting signal
904 information (siginfo) may have no effect in some ptrace-stops, yet the
905 call may succeed (return 0 and not set errno); querying
906 PTRACE_GETEVENTMSG may succeed and return some random value if current
907 ptrace-stop is not documented as returning a meaningful event message.
908 The call
910 ptrace(PTRACE_SETOPTIONS, pid, 0, PTRACE_O_flags);
912 affects one tracee. The tracee's current flags are replaced. Flags
914 are inherited by new tracees created and "auto-attached" via active
915 PTRACE_O_TRACEFORK, PTRACE_O_TRACEVFORK, or PTRACE_O_TRACECLONE
916 options.
917 Another group of commands makes the ptrace-stopped tracee run. They
919 have the form:
920 ptrace(cmd, pid, 0, sig);
922 where cmd is PTRACE_CONT, PTRACE_LISTEN, PTRACE_DETACH, PTRACE_SYSCALL,
924 PTRACE_SINGLESTEP, PTRACE_SYSEMU, or PTRACE_SYSEMU_SINGLESTEP. If the
925 tracee is in signal-delivery-stop, sig is the signal to be injected (if
926 it is nonzero). Otherwise, sig may be ignored. (When restarting a
927 tracee from a ptrace-stop other than signal-delivery-stop, recommended
928 practice is to always pass 0 in sig.)
929 Attaching and detaching
931 A thread can be attached to the tracer using the call
932 ptrace(PTRACE_ATTACH, pid, 0, 0);
934 or
936 ptrace(PTRACE_SEIZE, pid, 0, PTRACE_O_flags);
938 PTRACE_ATTACH sends SIGSTOP to this thread. If the tracer wants this
940 SIGSTOP to have no effect, it needs to suppress it. Note that if other
941 signals are concurrently sent to this thread during attach, the tracer
942 may see the tracee enter signal-delivery-stop with other signal(s)
943 first! The usual practice is to reinject these signals until SIGSTOP
944 is seen, then suppress SIGSTOP injection. The design bug here is that
945 a ptrace attach and a concurrently delivered SIGSTOP may race and the
946 concurrent SIGSTOP may be lost.
947 Since attaching sends SIGSTOP and the tracer usually suppresses it,
949 this may cause a stray EINTR return from the currently executing system
950 call in the tracee, as described in the "Signal injection and suppres‐
951 sion" section.
952 Since Linux 3.4, PTRACE_SEIZE can be used instead of PTRACE_ATTACH.
954 PTRACE_SEIZE does not stop the attached process. If you need to stop
955 it after attach (or at any other time) without sending it any signals,
956 use PTRACE_INTERRUPT command.
957 The request
959 ptrace(PTRACE_TRACEME, 0, 0, 0);
961 turns the calling thread into a tracee. The thread continues to run
963 (doesn't enter ptrace-stop). A common practice is to follow the
964 PTRACE_TRACEME with
965 raise(SIGSTOP);
967 and allow the parent (which is our tracer now) to observe our signal-
969 delivery-stop.
970 If the PTRACE_O_TRACEFORK, PTRACE_O_TRACEVFORK, or PTRACE_O_TRACECLONE
972 options are in effect, then children created by, respectively, vfork(2)
973 or clone(2) with the CLONE_VFORK flag, fork(2) or clone(2) with the
974 exit signal set to SIGCHLD, and other kinds of clone(2), are automati‐
975 cally attached to the same tracer which traced their parent. SIGSTOP
976 is delivered to the children, causing them to enter signal-delivery-
977 stop after they exit the system call which created them.
978 Detaching of the tracee is performed by:
980 ptrace(PTRACE_DETACH, pid, 0, sig);
982 PTRACE_DETACH is a restarting operation; therefore it requires the
984 tracee to be in ptrace-stop. If the tracee is in signal-delivery-stop,
985 a signal can be injected. Otherwise, the sig parameter may be silently
986 ignored.
987 If the tracee is running when the tracer wants to detach it, the usual
989 solution is to send SIGSTOP (using tgkill(2), to make sure it goes to
990 the correct thread), wait for the tracee to stop in signal-delivery-
991 stop for SIGSTOP and then detach it (suppressing SIGSTOP injection). A
992 design bug is that this can race with concurrent SIGSTOPs. Another
993 complication is that the tracee may enter other ptrace-stops and needs
994 to be restarted and waited for again, until SIGSTOP is seen. Yet
995 another complication is to be sure that the tracee is not already
996 ptrace-stopped, because no signal delivery happens while it is—not even
997 SIGSTOP.
998 If the tracer dies, all tracees are automatically detached and
1000 restarted, unless they were in group-stop. Handling of restart from
1001 group-stop is currently buggy, but the "as planned" behavior is to
1002 leave tracee stopped and waiting for SIGCONT. If the tracee is
1003 restarted from signal-delivery-stop, the pending signal is injected.
1004 execve(2) under ptrace
1006 When one thread in a multithreaded process calls execve(2), the kernel
1007 destroys all other threads in the process, and resets the thread ID of
1008 the execing thread to the thread group ID (process ID). (Or, to put
1009 things another way, when a multithreaded process does an execve(2), at
1010 completion of the call, it appears as though the execve(2) occurred in
1011 the thread group leader, regardless of which thread did the execve(2).)
1012 This resetting of the thread ID looks very confusing to tracers:
1013 * All other threads stop in PTRACE_EVENT_EXIT stop, if the
1015 PTRACE_O_TRACEEXIT option was turned on. Then all other threads
1016 except the thread group leader report death as if they exited via
1017 _exit(2) with exit code 0.
1018 * The execing tracee changes its thread ID while it is in the
1020 execve(2). (Remember, under ptrace, the "pid" returned from wait‐
1021 pid(2), or fed into ptrace calls, is the tracee's thread ID.) That
1022 is, the tracee's thread ID is reset to be the same as its process
1023 ID, which is the same as the thread group leader's thread ID.
1024 * Then a PTRACE_EVENT_EXEC stop happens, if the PTRACE_O_TRACEEXEC
1026 option was turned on.
1027 * If the thread group leader has reported its PTRACE_EVENT_EXIT stop
1029 by this time, it appears to the tracer that the dead thread leader
1030 "reappears from nowhere". (Note: the thread group leader does not
1031 report death via WIFEXITED(status) until there is at least one other
1032 live thread. This eliminates the possibility that the tracer will
1033 see it dying and then reappearing.) If the thread group leader was
1034 still alive, for the tracer this may look as if thread group leader
1035 returns from a different system call than it entered, or even
1036 "returned from a system call even though it was not in any system
1037 call". If the thread group leader was not traced (or was traced by
1038 a different tracer), then during execve(2) it will appear as if it
1039 has become a tracee of the tracer of the execing tracee.
1040 All of the above effects are the artifacts of the thread ID change in
1042 the tracee.
1043 The PTRACE_O_TRACEEXEC option is the recommended tool for dealing with
1045 this situation. First, it enables PTRACE_EVENT_EXEC stop, which occurs
1046 before execve(2) returns. In this stop, the tracer can use
1047 PTRACE_GETEVENTMSG to retrieve the tracee's former thread ID. (This
1048 feature was introduced in Linux 3.0.) Second, the PTRACE_O_TRACEEXEC
1049 option disables legacy SIGTRAP generation on execve(2).
1050 When the tracer receives PTRACE_EVENT_EXEC stop notification, it is
1052 guaranteed that except this tracee and the thread group leader, no
1053 other threads from the process are alive.
1054 On receiving the PTRACE_EVENT_EXEC stop notification, the tracer should
1056 clean up all its internal data structures describing the threads of
1057 this process, and retain only one data structure—one which describes
1058 the single still running tracee, with
1059 thread ID == thread group ID == process ID.
1061 Example: two threads call execve(2) at the same time:
1063 *** we get syscall-enter-stop in thread 1: **
1065 PID1 execve("/bin/foo", "foo" <unfinished ...>
1066 *** we issue PTRACE_SYSCALL for thread 1 **
1067 *** we get syscall-enter-stop in thread 2: **
1068 PID2 execve("/bin/bar", "bar" <unfinished ...>
1069 *** we issue PTRACE_SYSCALL for thread 2 **
1070 *** we get PTRACE_EVENT_EXEC for PID0, we issue PTRACE_SYSCALL **
1071 *** we get syscall-exit-stop for PID0: **
1072 PID0 <... execve resumed> ) = 0
1073 If the PTRACE_O_TRACEEXEC option is not in effect for the execing
1075 tracee, and if the tracee was PTRACE_ATTACHed rather that
1076 PTRACE_SEIZEd, the kernel delivers an extra SIGTRAP to the tracee after
1077 execve(2) returns. This is an ordinary signal (similar to one which
1078 can be generated by kill -TRAP), not a special kind of ptrace-stop.
1079 Employing PTRACE_GETSIGINFO for this signal returns si_code set to 0
1080 (SI_USER). This signal may be blocked by signal mask, and thus may be
1081 delivered (much) later.
1082 Usually, the tracer (for example, strace(1)) would not want to show
1084 this extra post-execve SIGTRAP signal to the user, and would suppress
1085 its delivery to the tracee (if SIGTRAP is set to SIG_DFL, it is a
1086 killing signal). However, determining which SIGTRAP to suppress is not
1087 easy. Setting the PTRACE_O_TRACEEXEC option or using PTRACE_SEIZE and
1088 thus suppressing this extra SIGTRAP is the recommended approach.
1089 Real parent
1091 The ptrace API (ab)uses the standard UNIX parent/child signaling over
1092 waitpid(2). This used to cause the real parent of the process to stop
1093 receiving several kinds of waitpid(2) notifications when the child
1094 process is traced by some other process.
1095 Many of these bugs have been fixed, but as of Linux 2.6.38 several
1097 still exist; see BUGS below.
1098 As of Linux 2.6.38, the following is believed to work correctly:
1100 * exit/death by signal is reported first to the tracer, then, when the
1102 tracer consumes the waitpid(2) result, to the real parent (to the
1103 real parent only when the whole multithreaded process exits). If
1104 the tracer and the real parent are the same process, the report is
1105 sent only once.
1106
1108 On success, the PTRACE_PEEK* requests return the requested data (but
1109 see NOTES), the PTRACE_SECCOMP_GET_FILTER request returns the number of
1110 instructions in the BPF program, and other requests return zero.
1111 On error, all requests return -1, and errno is set appropriately.
1113 Since the value returned by a successful PTRACE_PEEK* request may be
1114 -1, the caller must clear errno before the call, and then check it
1115 afterward to determine whether or not an error occurred.
1116
1118 EBUSY (i386 only) There was an error with allocating or freeing a
1119 debug register.
1120 EFAULT There was an attempt to read from or write to an invalid area in
1122 the tracer's or the tracee's memory, probably because the area
1123 wasn't mapped or accessible. Unfortunately, under Linux, dif‐
1124 ferent variations of this fault will return EIO or EFAULT more
1125 or less arbitrarily.
1126 EINVAL An attempt was made to set an invalid option.
1128 EIO request is invalid, or an attempt was made to read from or write
1130 to an invalid area in the tracer's or the tracee's memory, or
1131 there was a word-alignment violation, or an invalid signal was
1132 specified during a restart request.
1133 EPERM The specified process cannot be traced. This could be because
1135 the tracer has insufficient privileges (the required capability
1136 is CAP_SYS_PTRACE); unprivileged processes cannot trace pro‐
1137 cesses that they cannot send signals to or those running set-
1138 user-ID/set-group-ID programs, for obvious reasons. Alterna‐
1139 tively, the process may already be being traced, or (on kernels
1140 before 2.6.26) be init(1) (PID 1).
1141 ESRCH The specified process does not exist, or is not currently being
1143 traced by the caller, or is not stopped (for requests that
1144 require a stopped tracee).
1145
1147 SVr4, 4.3BSD.
1148
1150 Although arguments to ptrace() are interpreted according to the proto‐
1151 type given, glibc currently declares ptrace() as a variadic function
1152 with only the request argument fixed. It is recommended to always sup‐
1153 ply four arguments, even if the requested operation does not use them,
1154 setting unused/ignored arguments to 0L or (void *) 0.
1155 In Linux kernels before 2.6.26, init(1), the process with PID 1, may
1157 not be traced.
1158 A tracees parent continues to be the tracer even if that tracer calls
1160 execve(2).
1161 The layout of the contents of memory and the USER area are quite oper‐
1163 ating-system- and architecture-specific. The offset supplied, and the
1164 data returned, might not entirely match with the definition of struct
1165 user.
1166 The size of a "word" is determined by the operating-system variant
1168 (e.g., for 32-bit Linux it is 32 bits).
1169 This page documents the way the ptrace() call works currently in Linux.
1171 Its behavior differs significantly on other flavors of UNIX. In any
1172 case, use of ptrace() is highly specific to the operating system and
1173 architecture.
1174 Ptrace access mode checking
1176 Various parts of the kernel-user-space API (not just ptrace() opera‐
1177 tions), require so-called "ptrace access mode" checks, whose outcome
1178 determines whether an operation is permitted (or, in a few cases,
1179 causes a "read" operation to return sanitized data). These checks are
1180 performed in cases where one process can inspect sensitive information
1181 about, or in some cases modify the state of, another process. The
1182 checks are based on factors such as the credentials and capabilities of
1183 the two processes, whether or not the "target" process is dumpable, and
1184 the results of checks performed by any enabled Linux Security Module
1185 (LSM)—for example, SELinux, Yama, or Smack—and by the commoncap LSM
1186 (which is always invoked).
1187 Prior to Linux 2.6.27, all access checks were of a single type. Since
1189 Linux 2.6.27, two access mode levels are distinguished:
1190 PTRACE_MODE_READ
1192 For "read" operations or other operations that are less danger‐
1193 ous, such as: get_robust_list(2); kcmp(2); reading
1194 /proc/[pid]/auxv, /proc/[pid]/environ, or /proc/[pid]/stat; or
1195 readlink(2) of a /proc/[pid]/ns/* file.
1196 PTRACE_MODE_ATTACH
1198 For "write" operations, or other operations that are more dan‐
1199 gerous, such as: ptrace attaching (PTRACE_ATTACH) to another
1200 process or calling process_vm_writev(2). (PTRACE_MODE_ATTACH
1201 was effectively the default before Linux 2.6.27.)
1202 Since Linux 4.5, the above access mode checks are combined (ORed) with
1204 one of the following modifiers:
1205 PTRACE_MODE_FSCREDS
1207 Use the caller's filesystem UID and GID (see credentials(7)) or
1208 effective capabilities for LSM checks.
1209 PTRACE_MODE_REALCREDS
1211 Use the caller's real UID and GID or permitted capabilities for
1212 LSM checks. This was effectively the default before Linux 4.5.
1213 Because combining one of the credential modifiers with one of the
1215 aforementioned access modes is typical, some macros are defined in the
1216 kernel sources for the combinations:
1217 PTRACE_MODE_READ_FSCREDS
1219 Defined as PTRACE_MODE_READ | PTRACE_MODE_FSCREDS.
1220 PTRACE_MODE_READ_REALCREDS
1222 Defined as PTRACE_MODE_READ | PTRACE_MODE_REALCREDS.
1223 PTRACE_MODE_ATTACH_FSCREDS
1225 Defined as PTRACE_MODE_ATTACH | PTRACE_MODE_FSCREDS.
1226 PTRACE_MODE_ATTACH_REALCREDS
1228 Defined as PTRACE_MODE_ATTACH | PTRACE_MODE_REALCREDS.
1229 One further modifier can be ORed with the access mode:
1231 PTRACE_MODE_NOAUDIT (since Linux 3.3)
1233 Don't audit this access mode check. This modifier is employed
1234 for ptrace access mode checks (such as checks when reading
1235 /proc/[pid]/stat) that merely cause the output to be filtered or
1236 sanitized, rather than causing an error to be returned to the
1237 caller. In these cases, accessing the file is not a security
1238 violation and there is no reason to generate a security audit
1239 record. This modifier suppresses the generation of such an
1240 audit record for the particular access check.
1241 Note that all of the PTRACE_MODE_* constants described in this subsec‐
1243 tion are kernel-internal, and not visible to user space. The constant
1244 names are mentioned here in order to label the various kinds of ptrace
1245 access mode checks that are performed for various system calls and
1246 accesses to various pseudofiles (e.g., under /proc). These names are
1247 used in other manual pages to provide a simple shorthand for labeling
1248 the different kernel checks.
1249 The algorithm employed for ptrace access mode checking determines
1251 whether the calling process is allowed to perform the corresponding
1252 action on the target process. (In the case of opening /proc/[pid]
1253 files, the "calling process" is the one opening the file, and the
1254 process with the corresponding PID is the "target process".) The algo‐
1255 rithm is as follows:
1256 1. If the calling thread and the target thread are in the same thread
1258 group, access is always allowed.
1259 2. If the access mode specifies PTRACE_MODE_FSCREDS, then, for the
1261 check in the next step, employ the caller's filesystem UID and GID.
1262 (As noted in credentials(7), the filesystem UID and GID almost
1263 always have the same values as the corresponding effective IDs.)
1264 Otherwise, the access mode specifies PTRACE_MODE_REALCREDS, so use
1266 the caller's real UID and GID for the checks in the next step.
1267 (Most APIs that check the caller's UID and GID use the effective
1268 IDs. For historical reasons, the PTRACE_MODE_REALCREDS check uses
1269 the real IDs instead.)
1270 3. Deny access if neither of the following is true:
1272 · The real, effective, and saved-set user IDs of the target match
1274 the caller's user ID, and the real, effective, and saved-set group
1275 IDs of the target match the caller's group ID.
1276 · The caller has the CAP_SYS_PTRACE capability in the user namespace
1278 of the target.
1279 4. Deny access if the target process "dumpable" attribute has a value
1281 other than 1 (SUID_DUMP_USER; see the discussion of PR_SET_DUMPABLE
1282 in prctl(2)), and the caller does not have the CAP_SYS_PTRACE capa‐
1283 bility in the user namespace of the target process.
1284 5. The kernel LSM security_ptrace_access_check() interface is invoked
1286 to see if ptrace access is permitted. The results depend on the
1287 LSM(s). The implementation of this interface in the commoncap LSM
1288 performs the following steps:
1289 a) If the access mode includes PTRACE_MODE_FSCREDS, then use the
1291 caller's effective capability set in the following check; other‐
1292 wise (the access mode specifies PTRACE_MODE_REALCREDS, so) use
1293 the caller's permitted capability set.
1294 b) Deny access if neither of the following is true:
1296 · The caller and the target process are in the same user names‐
1298 pace, and the caller's capabilities are a superset of the tar‐
1299 get process's permitted capabilities.
1300 · The caller has the CAP_SYS_PTRACE capability in the target
1302 process's user namespace.
1303 Note that the commoncap LSM does not distinguish between
1305 PTRACE_MODE_READ and PTRACE_MODE_ATTACH.
1306 6. If access has not been denied by any of the preceding steps, then
1308 access is allowed.
1309 /proc/sys/kernel/yama/ptrace_scope
1311 On systems with the Yama Linux Security Module (LSM) installed (i.e.,
1312 the kernel was configured with CONFIG_SECURITY_YAMA), the
1313 /proc/sys/kernel/yama/ptrace_scope file (available since Linux 3.4) can
1314 be used to restrict the ability to trace a process with ptrace() (and
1315 thus also the ability to use tools such as strace(1) and gdb(1)). The
1316 goal of such restrictions is to prevent attack escalation whereby a
1317 compromised process can ptrace-attach to other sensitive processes
1318 (e.g., a GPG agent or an SSH session) owned by the user in order to
1319 gain additional credentials that may exist in memory and thus expand
1320 the scope of the attack.
1321 More precisely, the Yama LSM limits two types of operations:
1323 * Any operation that performs a ptrace access mode PTRACE_MODE_ATTACH
1325 check—for example, ptrace() PTRACE_ATTACH. (See the "Ptrace access
1326 mode checking" discussion above.)
1327 * ptrace() PTRACE_TRACEME.
1329 A process that has the CAP_SYS_PTRACE capability can update the
1331 /proc/sys/kernel/yama/ptrace_scope file with one of the following val‐
1332 ues:
1333 0 ("classic ptrace permissions")
1335 No additional restrictions on operations that perform
1336 PTRACE_MODE_ATTACH checks (beyond those imposed by the commoncap
1337 and other LSMs).
1338 The use of PTRACE_TRACEME is unchanged.
1340 1 ("restricted ptrace") [default value]
1342 When performing an operation that requires a PTRACE_MODE_ATTACH
1343 check, the calling process must either have the CAP_SYS_PTRACE
1344 capability in the user namespace of the target process or it
1345 must have a predefined relationship with the target process. By
1346 default, the predefined relationship is that the target process
1347 must be a descendant of the caller.
1348 A target process can employ the prctl(2) PR_SET_PTRACER opera‐
1350 tion to declare an additional PID that is allowed to perform
1351 PTRACE_MODE_ATTACH operations on the target. See the kernel
1352 source file Documentation/admin-guide/LSM/Yama.rst (or Documen‐
1353 tation/security/Yama.txt before Linux 4.13) for further details.
1354 The use of PTRACE_TRACEME is unchanged.
1356 2 ("admin-only attach")
1358 Only processes with the CAP_SYS_PTRACE capability in the user
1359 namespace of the target process may perform PTRACE_MODE_ATTACH
1360 operations or trace children that employ PTRACE_TRACEME.
1361 3 ("no attach")
1363 No process may perform PTRACE_MODE_ATTACH operations or trace
1364 children that employ PTRACE_TRACEME.
1365 Once this value has been written to the file, it cannot be
1367 changed.
1368 With respect to values 1 and 2, note that creating a new user namespace
1370 effectively removes the protection offered by Yama. This is because a
1371 process in the parent user namespace whose effective UID matches the
1372 UID of the creator of a child namespace has all capabilities (including
1373 CAP_SYS_PTRACE) when performing operations within the child user names‐
1374 pace (and further-removed descendants of that namespace). Conse‐
1375 quently, when a process tries to use user namespaces to sandbox itself,
1376 it inadvertently weakens the protections offered by the Yama LSM.
1377 C library/kernel differences
1379 At the system call level, the PTRACE_PEEKTEXT, PTRACE_PEEKDATA, and
1380 PTRACE_PEEKUSER requests have a different API: they store the result at
1381 the address specified by the data parameter, and the return value is
1382 the error flag. The glibc wrapper function provides the API given in
1383 DESCRIPTION above, with the result being returned via the function
1384 return value.
1385
1387 On hosts with 2.6 kernel headers, PTRACE_SETOPTIONS is declared with a
1388 different value than the one for 2.4. This leads to applications com‐
1389 piled with 2.6 kernel headers failing when run on 2.4 kernels. This
1390 can be worked around by redefining PTRACE_SETOPTIONS to PTRACE_OLDSE‐
1391 TOPTIONS, if that is defined.
1392 Group-stop notifications are sent to the tracer, but not to real par‐
1394 ent. Last confirmed on 2.6.38.6.
1395 If a thread group leader is traced and exits by calling _exit(2), a
1397 PTRACE_EVENT_EXIT stop will happen for it (if requested), but the sub‐
1398 sequent WIFEXITED notification will not be delivered until all other
1399 threads exit. As explained above, if one of other threads calls
1400 execve(2), the death of the thread group leader will never be reported.
1401 If the execed thread is not traced by this tracer, the tracer will
1402 never know that execve(2) happened. One possible workaround is to
1403 PTRACE_DETACH the thread group leader instead of restarting it in this
1404 case. Last confirmed on 2.6.38.6.
1405 A SIGKILL signal may still cause a PTRACE_EVENT_EXIT stop before actual
1407 signal death. This may be changed in the future; SIGKILL is meant to
1408 always immediately kill tasks even under ptrace. Last confirmed on
1409 Linux 3.13.
1410 Some system calls return with EINTR if a signal was sent to a tracee,
1412 but delivery was suppressed by the tracer. (This is very typical oper‐
1413 ation: it is usually done by debuggers on every attach, in order to not
1414 introduce a bogus SIGSTOP). As of Linux 3.2.9, the following system
1415 calls are affected (this list is likely incomplete): epoll_wait(2), and
1416 read(2) from an inotify(7) file descriptor. The usual symptom of this
1417 bug is that when you attach to a quiescent process with the command
1418 strace -p <process-ID>
1420 then, instead of the usual and expected one-line output such as
1422 restart_syscall(<... resuming interrupted call ...>_
1424 or
1426 select(6, [5], NULL, [5], NULL_
1428 ('_' denotes the cursor position), you observe more than one line. For
1430 example:
1431 clock_gettime(CLOCK_MONOTONIC, {15370, 690928118}) = 0
1433 epoll_wait(4,_
1434 What is not visible here is that the process was blocked in
1436 epoll_wait(2) before strace(1) has attached to it. Attaching caused
1437 epoll_wait(2) to return to user space with the error EINTR. In this
1438 particular case, the program reacted to EINTR by checking the current
1439 time, and then executing epoll_wait(2) again. (Programs which do not
1440 expect such "stray" EINTR errors may behave in an unintended way upon
1441 an strace(1) attach.)
1442 Contrary to the normal rules, the glibc wrapper for ptrace() can set
1444 errno to zero.
1445
1447 gdb(1), ltrace(1), strace(1), clone(2), execve(2), fork(2), gettid(2),
1448 prctl(2), seccomp(2), sigaction(2), tgkill(2), vfork(2), waitpid(2),
1449 exec(3), capabilities(7), signal(7)
1450
1452 This page is part of release 5.02 of the Linux man-pages project. A
1453 description of the project, information about reporting bugs, and the
1454 latest version of this page, can be found at
1455 https://www.kernel.org/doc/man-pages/.
1456Linux 2018-04-30 PTRACE(2)