xref: /freebsd/lib/libsys/procctl.2 (revision e3ebc5f534ee988d98c32b5e256ff08f71931225)
1.\" Copyright (c) 2013 Hudson River Trading LLC
2.\" Written by: John H. Baldwin <jhb@FreeBSD.org>
3.\" All rights reserved.
4.\"
5.\" Copyright (c) 2014 The FreeBSD Foundation
6.\" Portions of this documentation were written by Konstantin Belousov
7.\" under sponsorship from the FreeBSD Foundation.
8.\"
9.\" Redistribution and use in source and binary forms, with or without
10.\" modification, are permitted provided that the following conditions
11.\" are met:
12.\" 1. Redistributions of source code must retain the above copyright
13.\"    notice, this list of conditions and the following disclaimer.
14.\" 2. Redistributions in binary form must reproduce the above copyright
15.\"    notice, this list of conditions and the following disclaimer in the
16.\"    documentation and/or other materials provided with the distribution.
17.\"
18.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
19.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
22.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28.\" SUCH DAMAGE.
29.\"
30.Dd October 26, 2023
31.Dt PROCCTL 2
32.Os
33.Sh NAME
34.Nm procctl
35.Nd control processes
36.Sh LIBRARY
37.Lb libc
38.Sh SYNOPSIS
39.In sys/procctl.h
40.Ft int
41.Fn procctl "idtype_t idtype" "id_t id" "int cmd" "void *data"
42.Sh DESCRIPTION
43The
44.Fn procctl
45system call provides for control over processes.
46The
47.Fa idtype
48and
49.Fa id
50arguments specify the set of processes to control.
51If multiple processes match the identifier,
52.Nm
53will make a
54.Dq best effort
55to control as many of the selected processes as possible.
56An error is only returned if no selected processes successfully complete
57the request.
58The following identifier types are supported:
59.Bl -tag -width P_PGID
60.It Dv P_PID
61Control the process with the process ID
62.Fa id .
63.Fa id
64zero is a shortcut for the calling process ID.
65.It Dv P_PGID
66Control processes belonging to the process group with the ID
67.Fa id .
68.El
69.Pp
70The control request to perform is specified by the
71.Fa cmd
72argument.
73.Pp
74All status changing requests
75.Dv *_CTL
76require the caller to have the right to debug the target.
77All status query requests
78.Dv *_STATUS
79require the caller to have the right to observe the target.
80.Pp
81The following commands are supported:
82.Bl -tag -width PROC_TRAPCAP_STATUS
83.It Dv PROC_ASLR_CTL
84Controls the Address Space Layout Randomization (ASLR) in the program
85images created
86by
87.Xr execve 2
88in the specified process or its descendants that do not either change
89the control or modify it by other means.
90The
91.Fa data
92parameter must point to the integer variable holding one of the following
93values:
94.Bl -tag -width PROC_ASLR_FORCE_DISABLE
95.It Dv PROC_ASLR_FORCE_ENABLE
96Request that ASLR is enabled after execution, even if it is disabled
97system-wide.
98The image flag and set-uid might prevent ASLR enablement still.
99.It Dv PROC_ASLR_FORCE_DISABLE
100Request that ASLR is disabled after execution.
101Same notes as for
102.Dv PROC_ASLR_FORCE_ENABLE
103apply.
104.It Dv PROC_ASLR_NOFORCE
105Use the system-wide configured policy for ASLR.
106.El
107.It Dv PROC_ASLR_STATUS
108Returns the current status of ASLR enablement for the target process.
109The
110.Fa data
111parameter must point to the integer variable, where one of the
112following values is written:
113.Bl -tag -width PROC_ASLR_FORCE_DISABLE
114.It Dv PROC_ASLR_FORCE_ENABLE
115.It Dv PROC_ASLR_FORCE_DISABLE
116.It Dv PROC_ASLR_NOFORCE
117.El
118.Pp
119If the currently executed image in the process itself has ASLR enabled,
120the
121.Dv PROC_ASLR_ACTIVE
122flag is or-ed with the value listed above.
123.It Dv PROC_PROTMAX_CTL
124Controls implicit application of PROT_MAX protection equal to the
125.Fa prot
126argument of the
127.Xr mmap 2
128syscall, in the target process.
129The
130.Fa data
131parameter must point to the integer variable holding one of the following
132values:
133.Bl -tag -width PROC_PROTMAX_FORCE_DISABLE
134.It Dv PROC_PROTMAX_FORCE_ENABLE
135Enables implicit PROT_MAX application,
136even if it is disabled system-wide by the sysctl
137.Va vm.imply_prot_max .
138The image flag might still prevent the enablement.
139.It Dv PROC_PROTMAX_FORCE_DISABLE
140Request that implicit application of PROT_MAX be disabled.
141Same notes as for
142.Dv PROC_PROTMAX_FORCE_ENABLE
143apply.
144.It Dv PROC_PROTMAX_NOFORCE
145Use the system-wide configured policy for PROT_MAX.
146.El
147.It Dv PROC_PROTMAX_STATUS
148Returns the current status of implicit PROT_MAX enablement for the
149target process.
150The
151.Fa data
152parameter must point to the integer variable, where one of the
153following values is written:
154.Bl -tag -width PROC_PROTMAX_FORCE_DISABLE
155.It Dv PROC_PROTMAX_FORCE_ENABLE
156.It Dv PROC_PROTMAX_FORCE_DISABLE
157.It Dv PROC_PROTMAX_NOFORCE
158.El
159.Pp
160If the currently executed image in the process itself has implicit PROT_MAX
161application enabled, the
162.Dv PROC_PROTMAX_ACTIVE
163flag is or-ed with the value listed above.
164.It Dv PROC_SPROTECT
165Set process protection state.
166This is used to mark a process as protected from being killed if the system
167exhausts the available memory and swap.
168The
169.Fa data
170parameter must point to an integer containing an operation and zero or more
171optional flags.
172The following operations are supported:
173.Bl -tag -width PPROT_CLEAR
174.It Dv PPROT_SET
175Mark the selected processes as protected.
176.It Dv PPROT_CLEAR
177Clear the protected state of selected processes.
178.El
179.Pp
180The following optional flags are supported:
181.Bl -tag -width PPROT_DESCEND
182.It Dv PPROT_DESCEND
183Apply the requested operation to all child processes of each selected process
184in addition to each selected process.
185.It Dv PPROT_INHERIT
186When used with
187.Dv PPROT_SET ,
188mark all future child processes of each selected process as protected.
189Future child processes will also mark all of their future child processes.
190.El
191.It Dv PROC_REAP_ACQUIRE
192Acquires the reaper status for the current process.
193Reaper status means that children orphaned by the reaper's descendants
194that were forked after the acquisition of reaper status are reparented to the
195reaper process.
196After system initialization,
197.Xr init 8
198is the default reaper.
199.It Dv PROC_REAP_RELEASE
200Release the reaper state for the current process.
201The reaper of the current process becomes the new reaper of the
202current process's descendants.
203.It Dv PROC_REAP_STATUS
204Provides information about the reaper of the specified process,
205or the process itself when it is a reaper.
206The
207.Fa data
208argument must point to a
209.Vt procctl_reaper_status
210structure which is filled in by the syscall on successful return.
211.Bd -literal
212struct procctl_reaper_status {
213	u_int	rs_flags;
214	u_int	rs_children;
215	u_int	rs_descendants;
216	pid_t	rs_reaper;
217	pid_t	rs_pid;
218};
219.Ed
220The
221.Fa rs_flags
222may have the following flags returned:
223.Bl -tag -width REAPER_STATUS_REALINIT
224.It Dv REAPER_STATUS_OWNED
225The specified process has acquired reaper status and has not
226released it.
227When the flag is returned, the specified process
228.Fa id ,
229pid, identifies the reaper, otherwise the
230.Fa rs_reaper
231field of the structure is set to the pid of the reaper
232for the specified process id.
233.It Dv REAPER_STATUS_REALINIT
234The specified process is the root of the reaper tree, i.e.,
235.Xr init 8 .
236.El
237.Pp
238The
239.Fa rs_children
240field returns the number of children of the reaper among the descendants.
241It is possible to have a child whose reaper is not the specified process,
242since the reaper for any existing children is not reset on the
243.Dv PROC_REAP_ACQUIRE
244operation.
245The
246.Fa rs_descendants
247field returns the total number of descendants of the reaper(s),
248not counting descendants of the reaper in the subtree.
249The
250.Fa rs_reaper
251field returns the reaper pid.
252The
253.Fa rs_pid
254returns the pid of one reaper child if there are any descendants.
255.It Dv PROC_REAP_GETPIDS
256Queries the list of descendants of the reaper of the specified process.
257The request takes a pointer to a
258.Vt procctl_reaper_pids
259structure in the
260.Fa data
261parameter.
262.Bd -literal
263struct procctl_reaper_pids {
264	u_int	rp_count;
265	struct procctl_reaper_pidinfo *rp_pids;
266};
267.Ed
268When called, the
269.Fa rp_pids
270field must point to an array of
271.Vt procctl_reaper_pidinfo
272structures, to be filled in on return,
273and the
274.Fa rp_count
275field must specify the size of the array,
276into which no more than
277.Fa rp_count
278elements will be filled in by the kernel.
279.Pp
280The
281.Vt "struct procctl_reaper_pidinfo"
282structure provides some information about one of the reaper's descendants.
283Note that for a descendant that is not a child, it may be incorrectly
284identified because of a race in which the original child process exited
285and the exited process's pid was reused for an unrelated process.
286.Bd -literal
287struct procctl_reaper_pidinfo {
288	pid_t	pi_pid;
289	pid_t	pi_subtree;
290	u_int	pi_flags;
291};
292.Ed
293The
294.Fa pi_pid
295field is the process id of the descendant.
296The
297.Fa pi_subtree
298field provides the pid of the child of the reaper, which is the (grand-)parent
299of the process.
300The
301.Fa pi_flags
302field returns the following flags, further describing the descendant:
303.Bl -tag -width REAPER_PIDINFO_EXITING
304.It Dv REAPER_PIDINFO_VALID
305Set to indicate that the
306.Vt procctl_reaper_pidinfo
307structure was filled in by the kernel.
308Zero-filling the
309.Fa rp_pids
310array and testing the
311.Dv REAPER_PIDINFO_VALID
312flag allows the caller to detect the end
313of the returned array.
314.It Dv REAPER_PIDINFO_CHILD
315The
316.Fa pi_pid
317field identifies the direct child of the reaper.
318.It Dv REAPER_PIDINFO_REAPER
319The reported process is itself a reaper.
320The descendants of the subordinate reaper are not reported.
321.It Dv REAPER_PIDINFO_ZOMBIE
322The reported process is in the zombie state, ready to be reaped.
323.It Dv REAPER_PIDINFO_STOPPED
324The reported process is stopped by a SIGSTOP/SIGTSTP signal.
325.It Dv REAPER_PIDINFO_EXITING
326The reported process is in the process of exiting (but not yet a zombie).
327.El
328.It Dv PROC_REAP_KILL
329Request to deliver a signal to some subset of the descendants of the reaper.
330The
331.Fa data
332parameter must point to a
333.Vt procctl_reaper_kill
334structure, which is used both for parameters and status return.
335.Bd -literal
336struct procctl_reaper_kill {
337	int	rk_sig;
338	u_int	rk_flags;
339	pid_t	rk_subtree;
340	u_int	rk_killed;
341	pid_t	rk_fpid;
342};
343.Ed
344The
345.Fa rk_sig
346field specifies the signal to be delivered.
347Zero is not a valid signal number, unlike for
348.Xr kill 2 .
349The
350.Fa rk_flags
351field further directs the operation.
352It is or-ed from the following flags:
353.Bl -tag -width REAPER_KILL_CHILDREN
354.It Dv REAPER_KILL_CHILDREN
355Deliver the specified signal only to direct children of the reaper.
356.It Dv REAPER_KILL_SUBTREE
357Deliver the specified signal only to descendants that were forked by
358the direct child with pid specified in the
359.Fa rk_subtree
360field.
361.El
362If neither the
363.Dv REAPER_KILL_CHILDREN
364nor the
365.Dv REAPER_KILL_SUBTREE
366flags are specified, all current descendants of the reaper are signalled.
367.Pp
368If a signal was delivered to any process, the return value from the request
369is zero.
370In this case, the
371.Fa rk_killed
372field identifies the number of processes signalled.
373The
374.Fa rk_fpid
375field is set to the pid of the first process for which signal
376delivery failed, e.g., due to permission problems.
377If no such process exists, the
378.Fa rk_fpid
379field is set to -1.
380.It Dv PROC_TRACE_CTL
381Enable or disable tracing of the specified process(es), according to the
382value of the integer argument.
383Tracing includes attachment to the process using the
384.Xr ptrace 2
385and
386.Xr ktrace 2 ,
387debugging sysctls,
388.Xr hwpmc 4 ,
389.Xr dtrace 1 ,
390and core dumping.
391Possible values for the
392.Fa data
393argument are:
394.Bl -tag -width PROC_TRACE_CTL_DISABLE_EXEC
395.It Dv PROC_TRACE_CTL_ENABLE
396Enable tracing, after it was disabled by
397.Dv PROC_TRACE_CTL_DISABLE .
398Only allowed for self.
399.It Dv PROC_TRACE_CTL_DISABLE
400Disable tracing for the specified process.
401Tracing is re-enabled when the process changes the executing
402program with the
403.Xr execve 2
404syscall.
405A child inherits the trace settings from the parent on
406.Xr fork 2 .
407.It Dv PROC_TRACE_CTL_DISABLE_EXEC
408Same as
409.Dv PROC_TRACE_CTL_DISABLE ,
410but the setting persists for the process even after
411.Xr execve 2 .
412.El
413.It Dv PROC_TRACE_STATUS
414Returns the current tracing status for the specified process in
415the integer variable pointed to by
416.Fa data .
417If tracing is disabled,
418.Fa data
419is set to -1.
420If tracing is enabled, but no debugger is attached by the
421.Xr ptrace 2
422syscall,
423.Fa data
424is set to 0.
425If a debugger is attached,
426.Fa data
427is set to the pid of the debugger process.
428.It Dv PROC_TRAPCAP_CTL
429Controls the capability mode sandbox actions for the specified
430sandboxed processes,
431on a return from any syscall which gives either a
432.Er ENOTCAPABLE
433or
434.Er ECAPMODE
435error.
436If the control is enabled, such errors from the syscalls cause
437delivery of the synchronous
438.Dv SIGTRAP
439signal to the thread immediately before returning from the syscalls.
440.Pp
441Possible values for the
442.Fa data
443argument are:
444.Bl -tag -width PROC_TRAPCAP_CTL_DISABLE
445.It Dv PROC_TRAPCAP_CTL_ENABLE
446Enable the
447.Dv SIGTRAP
448signal delivery on capability mode access violations.
449The enabled mode is inherited by the children of the process,
450and is kept after
451.Xr fexecve 2
452calls.
453.It Dv PROC_TRAPCAP_CTL_DISABLE
454Disable the signal delivery on capability mode access violations.
455Note that the global sysctl
456.Dv kern.trap_enotcap
457might still cause the signal to be delivered.
458See
459.Xr capsicum 4 .
460.El
461.Pp
462On signal delivery, the
463.Va si_errno
464member of the
465.Fa siginfo
466signal handler parameter is set to the syscall error value,
467and the
468.Va si_code
469member is set to
470.Dv TRAP_CAP .
471The system call number is stored in the
472.Va si_syscall
473field of the
474.Fa siginfo
475signal handler parameter.
476The other system call parameters can be read from the
477.Fa ucontext_t
478but the system call number is typically stored in the register
479that also contains the return value and so is unavailable in the
480signal handler.
481.Pp
482See
483.Xr capsicum 4
484for more information about the capability mode.
485.It Dv PROC_TRAPCAP_STATUS
486Return the current status of signalling capability mode access
487violations for the specified process.
488The integer value pointed to by the
489.Fa data
490argument is set to the
491.Dv PROC_TRAPCAP_CTL_ENABLE
492value if the process control enables signal delivery, and to
493.Dv PROC_TRAPCAP_CTL_DISABLE
494otherwise.
495.Pp
496See the note about sysctl
497.Dv kern.trap_enotcap
498above, which gives independent global control of signal delivery.
499.It Dv PROC_PDEATHSIG_CTL
500Request the delivery of a signal when the parent of the calling
501process exits.
502.Fa idtype
503must be
504.Dv P_PID
505and
506.Fa id
507must be the either caller's pid or zero, with no difference in effect.
508The value is cleared for child processes
509and when executing set-user-ID or set-group-ID binaries.
510.Fa data
511must point to a value of type
512.Vt int
513indicating the signal
514that should be delivered to the caller.
515Use zero to cancel a previously requested signal delivery.
516.It Dv PROC_PDEATHSIG_STATUS
517Query the current signal number that will be delivered when the parent
518of the calling process exits.
519.Fa idtype
520must be
521.Dv P_PID
522and
523.Fa id
524must be the either caller's pid or zero, with no difference in effect.
525.Fa data
526must point to a memory location that can hold a value of type
527.Vt int .
528If signal delivery has not been requested, it will contain zero
529on return.
530.It Dv PROC_STACKGAP_CTL
531Controls the stack gaps in the specified process.
532A stack gap is the part of the growth area for a
533.Dv MAP_STACK
534mapped region that is reserved and never filled by memory.
535Instead, the process is guaranteed to receive a
536.Dv SIGSEGV
537signal on accessing pages in the gap.
538Gaps protect against stack overflow corrupting memory adjacent
539to the stack.
540.Pp
541The
542.Fa data
543argument must point to an integer variable containing flags.
544The following flags are allowed:
545.Bl -tag -width PROC_STACKGAP_DISABLE_EXEC
546.It Dv PROC_STACKGAP_ENABLE
547This flag is only accepted for consistency with
548.Dv PROC_STACKGAP_STATUS .
549If stack gaps are enabled, the flag is ignored.
550If disabled, the flag causes an
551.Ev EINVAL
552error to be returned.
553After gaps are disabled in a process, they can only be re-enabled when an
554.Xr execve 2
555is performed.
556.It Dv PROC_STACKGAP_DISABLE
557Disable stack gaps for the process.
558For existing stacks, the gap is no longer a reserved part of the growth
559area and can be filled by memory on access.
560.It Dv PROC_STACKGAP_ENABLE_EXEC
561Enable stack gaps for programs started after an
562.Xr execve 2
563by the specified process.
564.It Dv PROC_STACKGAP_DISABLE_EXEC
565Inherit disabled stack gaps state after
566.Xr execve 2 .
567In other words, if the currently executing program has stack gaps disabled,
568they are kept disabled on exec.
569If gaps were enabled, they are kept enabled after exec.
570.El
571.Pp
572The stack gap state is inherited from the parent on
573.Xr fork 2 .
574.It Dv PROC_STACKGAP_STATUS
575Returns the current stack gap state for the specified process.
576.Fa data
577must point to an integer variable, which is used to return a bitmask
578consisting of the following flags:
579.Bl -tag -width PROC_STACKGAP_DISABLE_EXEC
580.It Dv PROC_STACKGAP_ENABLE
581Stack gaps are enabled.
582.It Dv PROC_STACKGAP_DISABLE
583Stack gaps are disabled.
584.It Dv PROC_STACKGAP_ENABLE_EXEC
585Stack gaps are enabled in the process after
586.Xr execve 2 .
587.It Dv PROC_STACKGAP_DISABLE_EXEC
588Stack gaps are disabled in the process after
589.Xr execve 2 .
590.El
591.It Dv PROC_NO_NEW_PRIVS_CTL
592Allows one to ignore the SUID and SGID bits on the program
593images activated by
594.Xr execve 2
595in the specified process and its future descendants.
596The
597.Fa data
598parameter must point to the integer variable holding the following
599value:
600.Bl -tag -width PROC_NO_NEW_PRIVS_ENABLE
601.It Dv PROC_NO_NEW_PRIVS_ENABLE
602Request SUID and SGID bits to be ignored.
603.El
604.Pp
605It is not possible to disable it once it has been enabled.
606.It Dv PROC_NO_NEW_PRIVS_STATUS
607Returns the current status of SUID/SGID enablement for the target process.
608The
609.Fa data
610parameter must point to the integer variable, where one of the
611following values is written:
612.Bl -tag -width PROC_NO_NEW_PRIVS_DISABLE
613.It Dv PROC_NO_NEW_PRIVS_ENABLE
614.It Dv PROC_NO_NEW_PRIVS_DISABLE
615.El
616.It Dv PROC_WXMAP_CTL
617Controls the 'write exclusive against execution' permissions for the
618mappings in the process address space.
619It overrides the global settings established by the
620.Dv kern.elf{32/64}.allow_wx
621sysctl,
622and the corresponding bit in the ELF control note, see
623.Xr elfctl 1 .
624.Pp
625The
626.Fa data
627parameter must point to the integer variable holding one of the
628following values:
629.Bl -tag -width PROC_WX_MAPPINGS_DISALLOW_EXEC
630.It Dv PROC_WX_MAPPINGS_PERMIT
631Enable creation of mappings that have both write and execute
632protection attributes, in the specified process' address space.
633.It Dv PROC_WX_MAPPINGS_DISALLOW_EXEC
634In the new address space created by
635.Xr execve 2 ,
636disallow creation of mappings that have both write and execute
637permissions.
638.El
639.Pp
640Once creation of writeable and executable mappings is allowed,
641it is impossible (and pointless) to disallow it.
642The only way to ensure the absence of such mappings after they
643were enabled in a given process, is to set the
644.Dv PROC_WX_MAPPINGS_DISALLOW_EXEC
645flag and
646.Xr execve 2
647an image.
648.It Dv PROC_WXMAP_STATUS
649Returns the current status of the 'write exclusive against execution'
650enforcement for the specified process.
651The
652.Dv data
653parameter must point to the integer variable, where one of the
654following values is written:
655.Bl -tag -width PROC_WX_MAPPINGS_DISALLOW_EXEC
656.It Dv PROC_WX_MAPPINGS_PERMIT
657Creation of simultaneously writable and executable mapping is permitted,
658otherwise the process cannot create such mappings.
659.It Dv PROC_WX_MAPPINGS_DISALLOW_EXEC
660After
661.Xr execve 2 ,
662the new address space should disallow creation of simultaneously
663writable and executable mappings.
664.El
665.Pp
666Additionally, if the address space of the process disallows
667creation of simultaneously writable and executable mappings and
668it is guaranteed that no such mapping was created since address space
669creation, the
670.Dv PROC_WXORX_ENFORCE
671flag is set in the returned value.
672.El
673.Sh x86 MACHINE-SPECIFIC REQUESTS
674.Bl -tag -width PROC_KPTI_STATUS
675.It Dv PROC_KPTI_CTL
676AMD64 only.
677Controls the Kernel Page Table Isolation (KPTI) option for the children
678of the specified process.
679For the command to work, the
680.Va vm.pmap.kpti
681tunable must be enabled on boot.
682It is not possible to change the KPTI setting for a running process,
683except at the
684.Xr execve 2 ,
685where the address space is reinitialized.
686.Pp
687The
688.Fa data
689parameter must point to an integer variable containing one of the
690following commands:
691.Bl -tag -width PROC_KPTI_CTL_DISABLE_ON_EXEC
692.It Dv PROC_KPTI_CTL_ENABLE_ON_EXEC
693Enable KPTI after
694.Xr execve 2 .
695.It Dv PROC_KPTI_CTL_DISABLE_ON_EXEC
696Disable KPTI after
697.Xr execve 2 .
698Only root or a process having the
699.Va PRIV_IO
700privilege might use this option.
701.El
702.It Dv PROC_KPTI_STATUS
703Returns the current KPTI status for the specified process.
704.Fa data
705must point to the integer variable, which returns the
706following statuses:
707.Bl -tag -width PROC_KPTI_CTL_DISABLE_ON_EXEC
708.It Dv PROC_KPTI_CTL_ENABLE_ON_EXEC
709.It Dv PROC_KPTI_CTL_DISABLE_ON_EXEC
710.El
711.Pp
712The status is or-ed with the
713.Va PROC_KPTI_STATUS_ACTIVE
714in case KPTI is active for the current address space of the process.
715.El
716.Sh NOTES
717Disabling tracing on a process should not be considered a security
718feature, as it is bypassable both by the kernel and privileged processes,
719and via other system mechanisms.
720As such, it should not be utilized to reliably protect cryptographic
721keying material or other confidential data.
722.Pp
723Note that processes can trivially bypass the 'no simultaneously
724writable and executable mappings' policy by first marking some mapping
725as writeable and write code to it, then removing write and adding
726execute permission.
727This may be legitimately required by some programs, such as JIT compilers.
728.Sh RETURN VALUES
729If an error occurs, a value of -1 is returned and
730.Va errno
731is set to indicate the error.
732.Sh ERRORS
733The
734.Fn procctl
735system call
736will fail if:
737.Bl -tag -width Er
738.It Bq Er EFAULT
739The
740.Fa data
741parameter points outside the process's allocated address space.
742.It Bq Er EINVAL
743The
744.Fa cmd
745argument specifies an unsupported command.
746.Pp
747The
748.Fa idtype
749argument specifies an unsupported identifier type.
750.It Bq Er EPERM
751The calling process does not have permission to perform the requested
752operation on any of the selected processes.
753.It Bq Er ESRCH
754No processes matched the requested
755.Fa idtype
756and
757.Fa id .
758.It Bq Er ESRCH
759No descendant processes can be found matching criteria specified in the
760.Dv PROC_REAP_KILL
761request.
762.It Bq Er EINVAL
763An invalid operation or flag was passed in
764.Fa data
765for a
766.Dv PROC_SPROTECT
767command.
768.It Bq Er EPERM
769The
770.Fa idtype
771argument is not equal to
772.Dv P_PID ,
773or
774.Fa id
775is not equal to the pid of the calling process, for
776.Dv PROC_REAP_ACQUIRE
777or
778.Dv PROC_REAP_RELEASE
779requests.
780.It Bq Er EINVAL
781Invalid or undefined flags were passed to a
782.Dv PROC_REAP_KILL
783request.
784.It Bq Er EINVAL
785An invalid or zero signal number was requested for a
786.Dv PROC_REAP_KILL
787request.
788.It Bq Er EINVAL
789The
790.Dv PROC_REAP_RELEASE
791request was issued by the
792.Xr init 8
793process.
794.It Bq Er EBUSY
795The
796.Dv PROC_REAP_ACQUIRE
797request was issued by a process that had already acquired reaper status
798and has not yet released it.
799.It Bq Er EBUSY
800The
801.Dv PROC_TRACE_CTL
802request was issued for a process already being traced.
803.It Bq Er EPERM
804The
805.Dv PROC_TRACE_CTL
806request to re-enable tracing of the process
807.Po Dv PROC_TRACE_CTL_ENABLE Pc ,
808or to disable persistence of
809.Dv PROC_TRACE_CTL_DISABLE
810on
811.Xr execve 2
812was issued for a non-current process.
813.It Bq Er EINVAL
814The value of the integer
815.Fa data
816parameter for the
817.Dv PROC_TRACE_CTL
818or
819.Dv PROC_TRAPCAP_CTL
820request is invalid.
821.It Bq Er EINVAL
822The
823.Dv PROC_PDEATHSIG_CTL
824or
825.Dv PROC_PDEATHSIG_STATUS
826request referenced an unsupported
827.Fa id ,
828.Fa idtype
829or invalid signal number.
830.El
831.Sh SEE ALSO
832.Xr dtrace 1 ,
833.Xr proccontrol 1 ,
834.Xr protect 1 ,
835.Xr cap_enter 2 ,
836.Xr kill 2 ,
837.Xr ktrace 2 ,
838.Xr mmap 2 ,
839.Xr mprotect 2 ,
840.Xr ptrace 2 ,
841.Xr wait 2 ,
842.Xr capsicum 4 ,
843.Xr hwpmc 4 ,
844.Xr init 8
845.Sh HISTORY
846The
847.Fn procctl
848function appeared in
849.Fx 9.3 .
850.Pp
851The reaper facility is based on a similar feature of Linux and
852DragonflyBSD, and first appeared in
853.Fx 10.2 .
854.Pp
855The
856.Dv PROC_PDEATHSIG_CTL
857facility is based on the prctl(PR_SET_PDEATHSIG, ...) feature of Linux,
858and first appeared in
859.Fx 11.2 .
860.Pp
861The ASLR support was added to system for the checklists compliance in
862.Fx 13.0 .
863