man2/prctl.2

   1 .\" Copyright (C) 1998 Andries Brouwer (aeb@cwi.nl)
   2 .\" and Copyright (C) 2002, 2006, 2008, 2012, 2013 Michael Kerrisk <mtk.manpages@gmail.com>
   3 .\" and Copyright Guillem Jover <guillem@hadrons.org>
   4 .\" and Copyright (C) 2014 Dave Hansen / Intel
   5 .\"
   6 .\" %%%LICENSE_START(VERBATIM)
   7 .\" Permission is granted to make and distribute verbatim copies of this
   8 .\" manual provided the copyright notice and this permission notice are
   9 .\" preserved on all copies.
  10 .\"
  11 .\" Permission is granted to copy and distribute modified versions of this
  12 .\" manual under the conditions for verbatim copying, provided that the
  13 .\" entire resulting derived work is distributed under the terms of a
  14 .\" permission notice identical to this one.
  15 .\"
  16 .\" Since the Linux kernel and libraries are constantly changing, this
  17 .\" manual page may be incorrect or out-of-date.  The author(s) assume no
  18 .\" responsibility for errors or omissions, or for damages resulting from
  19 .\" the use of the information contained herein.  The author(s) may not
  20 .\" have taken the same level of care in the production of this manual,
  21 .\" which is licensed free of charge, as they might when working
  22 .\" professionally.
  23 .\"
  24 .\" Formatted or processed versions of this manual, if unaccompanied by
  25 .\" the source, must acknowledge the copyright and authors of this work.
  26 .\" %%%LICENSE_END
  27 .\"
  28 .\" Modified Thu Nov 11 04:19:42 MET 1999, aeb: added PR_GET_PDEATHSIG
  29 .\" Modified 27 Jun 02, Michael Kerrisk
  30 .\"     Added PR_SET_DUMPABLE, PR_GET_DUMPABLE,
  31 .\"     PR_SET_KEEPCAPS, PR_GET_KEEPCAPS
  32 .\" Modified 2006-08-30 Guillem Jover <guillem@hadrons.org>
  33 .\"     Updated Linux versions where the options where introduced.
  34 .\"     Added PR_SET_TIMING, PR_GET_TIMING, PR_SET_NAME, PR_GET_NAME,
  35 .\"     PR_SET_UNALIGN, PR_GET_UNALIGN, PR_SET_FPEMU, PR_GET_FPEMU,
  36 .\"     PR_SET_FPEXC, PR_GET_FPEXC
  37 .\" 2008-04-29 Serge Hallyn, Document PR_CAPBSET_READ and PR_CAPBSET_DROP
  38 .\" 2008-06-13 Erik Bosman, <ejbosman@cs.vu.nl>
  39 .\"     Document PR_GET_TSC and PR_SET_TSC.
  40 .\" 2008-06-15 mtk, Document PR_SET_SECCOMP, PR_GET_SECCOMP
  41 .\" 2009-10-03 Andi Kleen, document PR_MCE_KILL
  42 .\" 2012-04 Cyrill Gorcunov, Document PR_SET_MM
  43 .\" 2012-04-25 Michael Kerrisk, Document PR_TASK_PERF_EVENTS_DISABLE and
  44 .\"                             PR_TASK_PERF_EVENTS_ENABLE
  45 .\" 2012-09-20 Kees Cook, update PR_SET_SECCOMP for mode 2
  46 .\" 2012-09-20 Kees Cook, document PR_SET_NO_NEW_PRIVS, PR_GET_NO_NEW_PRIVS
  47 .\" 2012-10-25 Michael Kerrisk, Document PR_SET_TIMERSLACK and
  48 .\"                             PR_GET_TIMERSLACK
  49 .\" 2013-01-10 Kees Cook, document PR_SET_PTRACER
  50 .\" 2012-02-04 Michael Kerrisk, document PR_{SET,GET}_CHILD_SUBREAPER
  51 .\" 2014-11-10 Dave Hansen, document PR_MPX_{EN,DIS}ABLE_MANAGEMENT
  52 .\"
  53 .\"
  54 .TH PRCTL 2 2016-07-17 "Linux" "Linux Programmer's Manual"
  55 .SH NAME
  56 prctl \- operations on a process
  57 .SH SYNOPSIS
  58 .nf
  59 .B #include <sys/prctl.h>
  60 .sp
  61 .BI "int prctl(int " option ", unsigned long " arg2 ", unsigned long " arg3 ,
  62 .BI "          unsigned long " arg4 ", unsigned long " arg5 );
  63 .fi
  64 .SH DESCRIPTION
  65 .BR prctl ()
  66 is called with a first argument describing what to do
  67 (with values defined in \fI<linux/prctl.h>\fP), and further
  68 arguments with a significance depending on the first one.
  69 The first argument can be:
  70 .\"
  71 .TP
  72 .BR PR_CAP_AMBIENT " (since Linux 4.3)"
  73 .\" commit 58319057b7847667f0c9585b9de0e8932b0fdb08
  74 Reads or changes the ambient capability set, according to the value of
  75 .IR arg2 ,
  76 which must be one of the following:
  77 .RS
  78 .\"
  79 .TP
  80 .B PR_CAP_AMBIENT_RAISE
  81 The capability specified in
  82 .I arg3
  83 is added to the ambient set.
  84 The specified capability must already be present in
  85 both the permitted and the inheritable sets of the process.
  86 This operation is not permitted if the
  87 .B SECBIT_NO_CAP_AMBIENT_RAISE
  88 securebit is set.
  89 .TP
  90 .B PR_CAP_AMBIENT_LOWER
  91 The capability specified in
  92 .I arg3
  93 is removed from the ambient set.
  94 .TP
  95 .B PR_CAP_AMBIENT_IS_SET
  96 The
  97 .BR prctl ()
  98 call returns 1 if the capability in
  99 .I arg3
 100 is in the ambient set and 0 if it is not.
 101 .TP
 102 .BR PR_CAP_AMBIENT_CLEAR_ALL
 103 All capabilities will be removed from the ambient set.
 104 This operation requires setting
 105 .I arg3
 106 to zero.
 107 .RE
 108 .IP
 109 In all of the above operations,
 110 .I arg4
 111 and
 112 .I arg5
 113 must be specified as 0.
 114 .TP
 115 .BR PR_CAPBSET_READ " (since Linux 2.6.25)"
 116 Return (as the function result) 1 if the capability specified in
 117 .I arg2
 118 is in the calling thread's capability bounding set,
 119 or 0 if it is not.
 120 (The capability constants are defined in
 121 .IR <linux/capability.h> .)
 122 The capability bounding set dictates
 123 whether the process can receive the capability through a
 124 file's permitted capability set on a subsequent call to
 125 .BR execve (2).
 126
 127 If the capability specified in
 128 .I arg2
 129 is not valid, then the call fails with the error
 130 .BR EINVAL .
 131 .TP
 132 .BR PR_CAPBSET_DROP " (since Linux 2.6.25)"
 133 If the calling thread has the
 134 .B CAP_SETPCAP
 135 capability within its user namespace, then drop the capability specified by
 136 .I arg2
 137 from the calling thread's capability bounding set.
 138 Any children of the calling thread will inherit the newly
 139 reduced bounding set.
 140
 141 The call fails with the error:
 142 .B EPERM
 143 if the calling thread does not have the
 144 .BR CAP_SETPCAP ;
 145 .BR EINVAL
 146 if
 147 .I arg2
 148 does not represent a valid capability; or
 149 .BR EINVAL
 150 if file capabilities are not enabled in the kernel,
 151 in which case bounding sets are not supported.
 152 .TP
 153 .BR PR_SET_CHILD_SUBREAPER " (since Linux 3.4)"
 154 .\" commit ebec18a6d3aa1e7d84aab16225e87fd25170ec2b
 155 If
 156 .I arg2
 157 is nonzero,
 158 set the "child subreaper" attribute of the calling process;
 159 if
 160 .I arg2
 161 is zero, unset the attribute.
 162
 163 When a process is marked as a child subreaper,
 164 all of the children that it creates, and their descendants,
 165 will be marked as having a subreaper.
 166 In effect, a subreaper fulfills the role of
 167 .BR init (1)
 168 for its descendant processes.
 169 Upon termination of a process
 170 that is orphaned (i.e., its immediate parent has already terminated)
 171 and marked as having a subreaper,
 172 the nearest still living ancestor subreaper
 173 will receive a
 174 .BR SIGCHLD
 175 signal and will be able to
 176 .BR wait (2)
 177 on the process to discover its termination status.
 178 .TP
 179 .BR PR_GET_CHILD_SUBREAPER " (since Linux 3.4)"
 180 Return the "child subreaper" setting of the caller,
 181 in the location pointed to by
 182 .IR "(int\ *) arg2" .
 183 .TP
 184 .BR PR_SET_DUMPABLE " (since Linux 2.3.20)"
 185 Set the state of the "dumpable" flag,
 186 which determines whether core dumps are produced for the calling process
 187 upon delivery of a signal whose default behavior is to produce a core dump.
 188
 189 In kernels up to and including 2.6.12,
 190 .I arg2
 191 must be either 0
 192 .RB ( SUID_DUMP_DISABLE ,
 193 process is not dumpable) or 1
 194 .RB ( SUID_DUMP_USER ,
 195 process is dumpable).
 196 Between kernels 2.6.13 and 2.6.17,
 197 .\" commit abf75a5033d4da7b8a7e92321d74021d1fcfb502
 198 the value 2 was also permitted,
 199 which caused any binary which normally would not be dumped
 200 to be dumped readable by root only;
 201 for security reasons, this feature has been removed.
 202 .\" See http://marc.theaimsgroup.com/?l=linux-kernel&m=115270289030630&w=2
 203 .\" Subject:    Fix prctl privilege escalation (CVE-2006-2451)
 204 .\" From:       Marcel Holtmann <marcel () holtmann ! org>
 205 .\" Date:       2006-07-12 11:12:00
 206 (See also the description of
 207 .I /proc/sys/fs/\:suid_dumpable
 208 in
 209 .BR proc (5).)
 210
 211 Normally, this flag is set to 1.
 212 However, it is reset to the current value contained in the file
 213 .IR /proc/sys/fs/\:suid_dumpable
 214 (which by default has the value 0),
 215 if any of the following attributes of the process
 216 are changed by the operations listed below:
 217 .\" See kernel/cred.c::commit_creds() (Linux 3.18 sources)
 218 .RS
 219 .IP * 3
 220 The effective user or group ID is changed.
 221 .IP *
 222 The filesystem user or group ID is changed (see
 223 .BR credentials (7)).
 224 .IP *
 225 The process's set of permitted capabilities (see
 226 .BR capabilities (7))
 227 is changed such that its new set of capabilities is
 228 not a subset of its previous set of capabilities.
 229 .RE
 230 .IP
 231 The operations that may trigger changes to the dumpable flag include:
 232 .\" Look for uses of commit_creds() in the kernel source code
 233 .RS
 234 .IP * 3
 235 execution
 236 .RB ( execve (2))
 237 of a set-user-ID or set-group-ID program,
 238 or a program that has capabilities (see
 239 .BR capabilities (7));
 240 .IP *
 241 .BR capset (2);
 242 and
 243 .IP *
 244 system calls that change process credentials
 245 .RB ( setuid (2)
 246 .BR setgid (2),
 247 .BR setresuid (2),
 248 .BR setresgid (2),
 249 .BR setgroups (2),
 250 and so on).
 251 .\" Also certain namespace operations;
 252 .RE
 253 .IP
 254 Processes that are not dumpable can not be attached via
 255 .BR ptrace (2)
 256 .BR PTRACE_ATTACH .
 257 .TP
 258 .BR PR_GET_DUMPABLE " (since Linux 2.3.20)"
 259 Return (as the function result) the current state of the calling
 260 process's dumpable flag.
 261 .\" Since Linux 2.6.13, the dumpable flag can have the value 2,
 262 .\" but in 2.6.13 PR_GET_DUMPABLE simply returns 1 if the dumpable
 263 .\" flags has a nonzero value.  This was fixed in 2.6.14.
 264 .TP
 265 .BR PR_SET_ENDIAN " (since Linux 2.6.18, PowerPC only)"
 266 Set the endian-ness of the calling process to the value given
 267 in \fIarg2\fP, which should be one of the following:
 268 .\" Respectively 0, 1, 2
 269 .BR PR_ENDIAN_BIG ,
 270 .BR PR_ENDIAN_LITTLE ,
 271 or
 272 .B PR_ENDIAN_PPC_LITTLE
 273 (PowerPC pseudo little endian).
 274 .TP
 275 .BR PR_GET_ENDIAN " (since Linux 2.6.18, PowerPC only)"
 276 Return the endian-ness of the calling process,
 277 in the location pointed to by
 278 .IR "(int\ *) arg2" .
 279 .TP
 280 .BR PR_SET_FPEMU " (since Linux 2.4.18, 2.5.9, only on ia64)"
 281 Set floating-point emulation control bits to \fIarg2\fP.
 282 Pass
 283 .B PR_FPEMU_NOPRINT
 284 to silently emulate floating-point operation accesses, or
 285 .B PR_FPEMU_SIGFPE
 286 to not emulate floating-point operations and send
 287 .B SIGFPE
 288 instead.
 289 .TP
 290 .BR PR_GET_FPEMU " (since Linux 2.4.18, 2.5.9, only on ia64)"
 291 Return floating-point emulation control bits,
 292 in the location pointed to by
 293 .IR "(int\ *) arg2" .
 294 .TP
 295 .BR PR_SET_FPEXC " (since Linux 2.4.21, 2.5.32, only on PowerPC)"
 296 Set floating-point exception mode to \fIarg2\fP.
 297 Pass \fBPR_FP_EXC_SW_ENABLE\fP to use FPEXC for FP exception enables,
 298 \fBPR_FP_EXC_DIV\fP for floating-point divide by zero,
 299 \fBPR_FP_EXC_OVF\fP for floating-point overflow,
 300 \fBPR_FP_EXC_UND\fP for floating-point underflow,
 301 \fBPR_FP_EXC_RES\fP for floating-point inexact result,
 302 \fBPR_FP_EXC_INV\fP for floating-point invalid operation,
 303 \fBPR_FP_EXC_DISABLED\fP for FP exceptions disabled,
 304 \fBPR_FP_EXC_NONRECOV\fP for async nonrecoverable exception mode,
 305 \fBPR_FP_EXC_ASYNC\fP for async recoverable exception mode,
 306 \fBPR_FP_EXC_PRECISE\fP for precise exception mode.
 307 .TP
 308 .BR PR_GET_FPEXC " (since Linux 2.4.21, 2.5.32, only on PowerPC)"
 309 Return floating-point exception mode,
 310 in the location pointed to by
 311 .IR "(int\ *) arg2" .
 312 .TP
 313 .BR PR_SET_KEEPCAPS " (since Linux 2.2.18)"
 314 Set the state of the thread's "keep capabilities" flag,
 315 which determines whether the thread's permitted
 316 capability set is cleared when a change is made to the thread's user IDs
 317 such that the thread's real UID, effective UID, and saved set-user-ID
 318 all become nonzero when at least one of them previously had the value 0.
 319 By default, the permitted capability set is cleared when such a change is made;
 320 setting the "keep capabilities" flag prevents it from being cleared.
 321 .I arg2
 322 must be either 0 (permitted capabilities are cleared)
 323 or 1 (permitted capabilities are kept).
 324 (A thread's
 325 .I effective
 326 capability set is always cleared when such a credential change is made,
 327 regardless of the setting of the "keep capabilities" flag.)
 328 The "keep capabilities" value will be reset to 0 on subsequent calls to
 329 .BR execve (2).
 330 .TP
 331 .BR PR_GET_KEEPCAPS " (since Linux 2.2.18)"
 332 Return (as the function result) the current state of the calling thread's
 333 "keep capabilities" flag.
 334 .TP
 335 .BR PR_MCE_KILL " (since Linux 2.6.32)"
 336 Set the machine check memory corruption kill policy for the current thread.
 337 If
 338 .I arg2
 339 is
 340 .BR PR_MCE_KILL_CLEAR ,
 341 clear the thread memory corruption kill policy and use the system-wide default.
 342 (The system-wide default is defined by
 343 .IR /proc/sys/vm/memory_failure_early_kill ;
 344 see
 345 .BR proc (5).)
 346 If
 347 .I arg2
 348 is
 349 .BR PR_MCE_KILL_SET ,
 350 use a thread-specific memory corruption kill policy.
 351 In this case,
 352 .I arg3
 353 defines whether the policy is
 354 .I early kill
 355 .RB ( PR_MCE_KILL_EARLY ),
 356 .I late kill
 357 .RB ( PR_MCE_KILL_LATE ),
 358 or the system-wide default
 359 .RB ( PR_MCE_KILL_DEFAULT ).
 360 Early kill means that the thread receives a
 361 .B SIGBUS
 362 signal as soon as hardware memory corruption is detected inside
 363 its address space.
 364 In late kill mode, the process is killed only when it accesses a corrupted page.
 365 See
 366 .BR sigaction (2)
 367 for more information on the
 368 .BR SIGBUS
 369 signal.
 370 The policy is inherited by children.
 371 The remaining unused
 372 .BR prctl ()
 373 arguments must be zero for future compatibility.
 374 .TP
 375 .BR PR_MCE_KILL_GET " (since Linux 2.6.32)"
 376 Return the current per-process machine check kill policy.
 377 All unused
 378 .BR prctl ()
 379 arguments must be zero.
 380 .TP
 381 .BR PR_SET_MM " (since Linux 3.3)"
 382 .\" commit 028ee4be34a09a6d48bdf30ab991ae933a7bc036
 383 Modify certain kernel memory map descriptor fields
 384 of the calling process.
 385 Usually these fields are set by the kernel and dynamic loader (see
 386 .BR ld.so (8)
 387 for more information) and a regular application should not use this feature.
 388 However, there are cases, such as self-modifying programs,
 389 where a program might find it useful to change its own memory map.
 390 This feature is available only if the kernel is built with the
 391 .BR CONFIG_CHECKPOINT_RESTORE
 392 option enabled.
 393 The calling process must have the
 394 .BR CAP_SYS_RESOURCE
 395 capability.
 396 The value in
 397 .I arg2
 398 is one of the options below, while
 399 .I arg3
 400 provides a new value for the option.
 401 .RS
 402 .TP
 403 .BR PR_SET_MM_START_CODE
 404 Set the address above which the program text can run.
 405 The corresponding memory area must be readable and executable,
 406 but not writable or sharable (see
 407 .BR mprotect (2)
 408 and
 409 .BR mmap (2)
 410 for more information).
 411 .TP
 412 .BR PR_SET_MM_END_CODE
 413 Set the address below which the program text can run.
 414 The corresponding memory area must be readable and executable,
 415 but not writable or sharable.
 416 .TP
 417 .BR PR_SET_MM_START_DATA
 418 Set the address above which initialized and
 419 uninitialized (bss) data are placed.
 420 The corresponding memory area must be readable and writable,
 421 but not executable or sharable.
 422 .TP
 423 .B PR_SET_MM_END_DATA
 424 Set the address below which initialized and
 425 uninitialized (bss) data are placed.
 426 The corresponding memory area must be readable and writable,
 427 but not executable or sharable.
 428 .TP
 429 .BR PR_SET_MM_START_STACK
 430 Set the start address of the stack.
 431 The corresponding memory area must be readable and writable.
 432 .TP
 433 .BR PR_SET_MM_START_BRK
 434 Set the address above which the program heap can be expanded with
 435 .BR brk (2)
 436 call.
 437 The address must be greater than the ending address of
 438 the current program data segment.
 439 In addition, the combined size of the resulting heap and
 440 the size of the data segment can't exceed the
 441 .BR RLIMIT_DATA
 442 resource limit (see
 443 .BR setrlimit (2)).
 444 .TP
 445 .BR PR_SET_MM_BRK
 446 Set the current
 447 .BR brk (2)
 448 value.
 449 The requirements for the address are the same as for the
 450 .BR PR_SET_MM_START_BRK
 451 option.
 452 .P
 453 The following options are available since Linux 3.5.
 454 .\" commit fe8c7f5cbf91124987106faa3bdf0c8b955c4cf7
 455 .TP
 456 .BR PR_SET_MM_ARG_START
 457 Set the address above which the program command line is placed.
 458 .TP
 459 .BR PR_SET_MM_ARG_END
 460 Set the address below which the program command line is placed.
 461 .TP
 462 .BR PR_SET_MM_ENV_START
 463 Set the address above which the program environment is placed.
 464 .TP
 465 .BR PR_SET_MM_ENV_END
 466 Set the address below which the program environment is placed.
 467 .IP
 468 The address passed with
 469 .BR PR_SET_MM_ARG_START ,
 470 .BR PR_SET_MM_ARG_END ,
 471 .BR PR_SET_MM_ENV_START ,
 472 and
 473 .BR PR_SET_MM_ENV_END
 474 should belong to a process stack area.
 475 Thus, the corresponding memory area must be readable, writable, and
 476 (depending on the kernel configuration) have the
 477 .BR MAP_GROWSDOWN
 478 attribute set (see
 479 .BR mmap (2)).
 480 .TP
 481 .BR PR_SET_MM_AUXV
 482 Set a new auxiliary vector.
 483 The
 484 .I arg3
 485 argument should provide the address of the vector.
 486 The
 487 .I arg4
 488 is the size of the vector.
 489 .TP
 490 .BR PR_SET_MM_EXE_FILE
 491 .\" commit b32dfe377102ce668775f8b6b1461f7ad428f8b6
 492 Supersede the
 493 .IR /proc/pid/exe
 494 symbolic link with a new one pointing to a new executable file
 495 identified by the file descriptor provided in
 496 .I arg3
 497 argument.
 498 The file descriptor should be obtained with a regular
 499 .BR open (2)
 500 call.
 501 .IP
 502 To change the symbolic link, one needs to unmap all existing
 503 executable memory areas, including those created by the kernel itself
 504 (for example the kernel usually creates at least one executable
 505 memory area for the ELF
 506 .IR \.text
 507 section).
 508 .IP
 509 The second limitation is that such transitions can be done only once
 510 in a process life time.
 511 Any further attempts will be rejected.
 512 This should help system administrators monitor unusual
 513 symbolic-link transitions over all processes running on a system.
 514 .RE
 515 .TP
 516 .BR PR_MPX_ENABLE_MANAGEMENT ", " PR_MPX_DISABLE_MANAGEMENT " (since Linux 3.19) "
 517 .\" commit fe3d197f84319d3bce379a9c0dc17b1f48ad358c
 518 .\" See also http://lwn.net/Articles/582712/
 519 .\" See also https://gcc.gnu.org/wiki/Intel%20MPX%20support%20in%20the%20GCC%20compiler
 520 Enable or disable kernel management of Memory Protection eXtensions (MPX)
 521 bounds tables.
 522 The
 523 .IR arg2 ,
 524 .IR arg3 ,
 525 .IR arg4 ,
 526 and
 527 .IR arg5
 528 .\" commit e9d1b4f3c60997fe197bf0243cb4a41a44387a88
 529 arguments must be zero.
 530
 531 MPX is a hardware-assisted mechanism for performing bounds checking on
 532 pointers.
 533 It consists of a set of registers storing bounds information
 534 and a set of special instruction prefixes that tell the CPU on which
 535 instructions it should do bounds enforcement.
 536 There is a limited number of these registers and
 537 when there are more pointers than registers,
 538 their contents must be "spilled" into a set of tables.
 539 These tables are called "bounds tables" and the MPX
 540 .BR prctl ()
 541 operations control
 542 whether the kernel manages their allocation and freeing.
 543
 544 When management is enabled, the kernel will take over allocation
 545 and freeing of the bounds tables.
 546 It does this by trapping the #BR exceptions that result
 547 at first use of missing bounds tables and
 548 instead of delivering the exception to user space,
 549 it allocates the table and populates the bounds directory
 550 with the location of the new table.
 551 For freeing, the kernel checks to see if bounds tables are
 552 present for memory which is not allocated, and frees them if so.
 553
 554 Before enabling MPX management using
 555 .BR PR_MPX_ENABLE_MANAGEMENT ,
 556 the application must first have allocated a user-space buffer for
 557 the bounds directory and placed the location of that directory in the
 558 .I bndcfgu
 559 register.
 560
 561 These calls will fail if the CPU or kernel does not support MPX.
 562 Kernel support for MPX is enabled via the
 563 .BR CONFIG_X86_INTEL_MPX
 564 configuration option.
 565 You can check whether the CPU supports MPX by looking for the 'mpx'
 566 CPUID bit, like with the following command:
 567
 568         cat /proc/cpuinfo | grep ' mpx '
 569
 570 A thread may not switch in or out of long (64-bit) mode while MPX is
 571 enabled.
 572
 573 All threads in a process are affected by these calls.
 574
 575 The child of a
 576 .BR fork (2)
 577 inherits the state of MPX management.
 578 During
 579 .BR execve (2),
 580 MPX management is reset to a state as if
 581 .BR PR_MPX_DISABLE_MANAGEMENT
 582 had been called.
 583
 584 For further information on Intel MPX, see the kernel source file
 585 .IR Documentation/x86/intel_mpx.txt .
 586 .TP
 587 .BR PR_SET_NAME " (since Linux 2.6.9)"
 588 Set the name of the calling thread,
 589 using the value in the location pointed to by
 590 .IR "(char\ *) arg2" .
 591 The name can be up to 16 bytes long,
 592 .\" TASK_COMM_LEN in include/linux/sched.h
 593 including the terminating null byte.
 594 (If the length of the string, including the terminating null byte,
 595 exceeds 16 bytes, the string is silently truncated.)
 596 This is the same attribute that can be set via
 597 .BR pthread_setname_np (3)
 598 and retrieved using
 599 .BR pthread_getname_np (3).
 600 The attribute is likewise accessible via
 601 .IR /proc/self/task/[tid]/comm ,
 602 where
 603 .I tid
 604 is the name of the calling thread.
 605 .TP
 606 .BR PR_GET_NAME " (since Linux 2.6.11)"
 607 Return the name of the calling thread,
 608 in the buffer pointed to by
 609 .IR "(char\ *) arg2" .
 610 The buffer should allow space for up to 16 bytes;
 611 the returned string will be null-terminated.
 612 .TP
 613 .BR PR_SET_NO_NEW_PRIVS " (since Linux 3.5)"
 614 Set the calling process's
 615 .I no_new_privs
 616 bit to the value in
 617 .IR arg2 .
 618 With
 619 .I no_new_privs
 620 set to 1,
 621 .BR execve (2)
 622 promises not to grant privileges to do anything
 623 that could not have been done without the
 624 .BR execve (2)
 625 call (for example,
 626 rendering the set-user-ID and set-group-ID mode bits,
 627 and file capabilities non-functional).
 628 Once set, this bit cannot be unset.
 629 The setting of this bit is inherited by children created by
 630 .BR fork (2)
 631 and
 632 .BR clone (2),
 633 and preserved across
 634 .BR execve (2).
 635
 636 For more information, see the kernel source file
 637 .IR Documentation/prctl/no_new_privs.txt .
 638 .TP
 639 .BR PR_GET_NO_NEW_PRIVS " (since Linux 3.5)"
 640 Return (as the function result) the value of the
 641 .I no_new_privs
 642 bit for the current process.
 643 A value of 0 indicates the regular
 644 .BR execve (2)
 645 behavior.
 646 A value of 1 indicates
 647 .BR execve (2)
 648 will operate in the privilege-restricting mode described above.
 649 .TP
 650 .BR PR_SET_PDEATHSIG " (since Linux 2.1.57)"
 651 Set the parent death signal
 652 of the calling process to \fIarg2\fP (either a signal value
 653 in the range 1..maxsig, or 0 to clear).
 654 This is the signal that the calling process will get when its
 655 parent dies.
 656 This value is cleared for the child of a
 657 .BR fork (2)
 658 and (since Linux 2.4.36 / 2.6.23)
 659 when executing a set-user-ID or set-group-ID binary,
 660 or a binary that has associated capabilities (see
 661 .BR capabilities (7)).
 662 This value is preserved across
 663 .BR execve (2).
 664
 665 .IR Warning :
 666 .\" https://bugzilla.kernel.org/show_bug.cgi?id=43300
 667 the "parent" in this case is considered to be the
 668 .I thread
 669 that created this process.
 670 In other words, the signal will be sent when that thread terminates
 671 (via, for example,
 672 .BR pthread_exit (3)),
 673 rather than after all of the threads in the parent process terminate.
 674 .TP
 675 .BR PR_GET_PDEATHSIG " (since Linux 2.3.15)"
 676 Return the current value of the parent process death signal,
 677 in the location pointed to by
 678 .IR "(int\ *) arg2" .
 679 .TP
 680 .BR PR_SET_PTRACER " (since Linux 3.4)"
 681 .\" commit 2d514487faf188938a4ee4fb3464eeecfbdcf8eb
 682 .\" commit bf06189e4d14641c0148bea16e9dd24943862215
 683 This is meaningful only when the Yama LSM is enabled and in mode 1
 684 ("restricted ptrace", visible via
 685 .IR /proc/sys/kernel/yama/ptrace_scope ).
 686 When a "ptracer process ID" is passed in \fIarg2\fP,
 687 the caller is declaring that the ptracer process can
 688 .BR ptrace (2)
 689 the calling process as if it were a direct process ancestor.
 690 Each
 691 .B PR_SET_PTRACER
 692 operation replaces the previous "ptracer process ID".
 693 Employing
 694 .B PR_SET_PTRACER
 695 with
 696 .I arg2
 697 set to 0 clears the caller's "ptracer process ID".
 698 If
 699 .I arg2
 700 is
 701 .BR PR_SET_PTRACER_ANY ,
 702 the ptrace restrictions introduced by Yama are effectively disabled for the
 703 calling process.
 704
 705 For further information, see the kernel source file
 706 .IR Documentation/security/Yama.txt .
 707 .TP
 708 .BR PR_SET_SECCOMP " (since Linux 2.6.23)"
 709 .\" See http://thread.gmane.org/gmane.linux.kernel/542632
 710 .\" [PATCH 0 of 2] seccomp updates
 711 .\" andrea@cpushare.com
 712 Set the secure computing (seccomp) mode for the calling thread, to limit
 713 the available system calls.
 714 The more recent
 715 .BR seccomp (2)
 716 system call provides a superset of the functionality of
 717 .BR PR_SET_SECCOMP .
 718
 719 The seccomp mode is selected via
 720 .IR arg2 .
 721 (The seccomp constants are defined in
 722 .IR <linux/seccomp.h> .)
 723
 724 With
 725 .IR arg2
 726 set to
 727 .BR SECCOMP_MODE_STRICT ,
 728 the only system calls that the thread is permitted to make are
 729 .BR read (2),
 730 .BR write (2),
 731 .BR _exit (2)
 732 (but not
 733 .BR exit_group (2)),
 734 and
 735 .BR sigreturn (2).
 736 Other system calls result in the delivery of a
 737 .BR SIGKILL
 738 signal.
 739 Strict secure computing mode is useful for number-crunching applications
 740 that may need to execute untrusted byte code,
 741 perhaps obtained by reading from a pipe or socket.
 742 This operation is available only
 743 if the kernel is configured with
 744 .B CONFIG_SECCOMP
 745 enabled.
 746
 747 With
 748 .IR arg2
 749 set to
 750 .BR SECCOMP_MODE_FILTER " (since Linux 3.5),"
 751 the system calls allowed are defined by a pointer
 752 to a Berkeley Packet Filter passed in
 753 .IR arg3 .
 754 This argument is a pointer to
 755 .IR "struct sock_fprog" ;
 756 it can be designed to filter
 757 arbitrary system calls and system call arguments.
 758 This mode is available only if the kernel is configured with
 759 .B CONFIG_SECCOMP_FILTER
 760 enabled.
 761
 762 If
 763 .BR SECCOMP_MODE_FILTER
 764 filters permit
 765 .BR fork (2),
 766 then the seccomp mode is inherited by children created by
 767 .BR fork (2);
 768 if
 769 .BR execve (2)
 770 is permitted, then the seccomp mode is preserved across
 771 .BR execve (2).
 772 If the filters permit
 773 .BR prctl ()
 774 calls, then additional filters can be added;
 775 they are run in order until the first non-allow result is seen.
 776
 777 For further information, see the kernel source file
 778 .IR Documentation/prctl/seccomp_filter.txt .
 779 .TP
 780 .BR PR_GET_SECCOMP " (since Linux 2.6.23)"
 781 Return (as the function result)
 782 the secure computing mode of the calling thread.
 783 If the caller is not in secure computing mode, this operation returns 0;
 784 if the caller is in strict secure computing mode, then the
 785 .BR prctl ()
 786 call will cause a
 787 .B SIGKILL
 788 signal to be sent to the process.
 789 If the caller is in filter mode, and this system call is allowed by the
 790 seccomp filters, it returns 2; otherwise, the process is killed with a
 791 .BR SIGKILL
 792 signal.
 793 This operation is available only
 794 if the kernel is configured with
 795 .B CONFIG_SECCOMP
 796 enabled.
 797
 798 Since Linux 3.8, the
 799 .IR Seccomp
 800 field of the
 801 .IR /proc/[pid]/status
 802 file provides a method of obtaining the same information,
 803 without the risk that the process is killed; see
 804 .BR proc (5).
 805 .TP
 806 .BR PR_SET_SECUREBITS " (since Linux 2.6.26)"
 807 Set the "securebits" flags of the calling thread to the value supplied in
 808 .IR arg2 .
 809 See
 810 .BR capabilities (7).
 811 .TP
 812 .BR PR_GET_SECUREBITS " (since Linux 2.6.26)"
 813 Return (as the function result)
 814 the "securebits" flags of the calling thread.
 815 See
 816 .BR capabilities (7).
 817 .TP
 818 .BR PR_SET_THP_DISABLE " (since Linux 3.15)"
 819 .\" commit a0715cc22601e8830ace98366c0c2bd8da52af52
 820 Set the state of the "THP disable" flag for the calling thread.
 821 If
 822 .I arg2
 823 has a nonzero value, the flag is set, otherwise it is cleared.
 824 Setting this flag provides a method
 825 for disabling transparent huge pages
 826 for jobs where the code cannot be modified, and using a malloc hook with
 827 .BR madvise (2)
 828 is not an option (i.e., statically allocated data).
 829 The setting of the "THP disable" flag is inherited by a child created via
 830 .BR fork (2)
 831 and is preserved across
 832 .BR execve (2).
 833 .\"
 834 .TP
 835 .BR PR_TASK_PERF_EVENTS_DISABLE " (since Linux 2.6.31)"
 836 Disable all performance counters attached to the calling process,
 837 regardless of whether the counters were created by
 838 this process or another process.
 839 Performance counters created by the calling process for other
 840 processes are unaffected.
 841 For more information on performance counters, see the Linux kernel source file
 842 .IR tools/perf/design.txt .
 843 .IP
 844 Originally called
 845 .BR PR_TASK_PERF_COUNTERS_DISABLE ;
 846 .\" commit 1d1c7ddbfab358445a542715551301b7fc363e28
 847 renamed (with same numerical value)
 848 in Linux 2.6.32.
 849 .\"
 850 .TP
 851 .BR PR_TASK_PERF_EVENTS_ENABLE " (since Linux 2.6.31)"
 852 The converse of
 853 .BR PR_TASK_PERF_EVENTS_DISABLE ;
 854 enable performance counters attached to the calling process.
 855 .IP
 856 Originally called
 857 .BR PR_TASK_PERF_COUNTERS_ENABLE ;
 858 .\" commit 1d1c7ddbfab358445a542715551301b7fc363e28
 859 renamed
 860 .\" commit cdd6c482c9ff9c55475ee7392ec8f672eddb7be6
 861 in Linux 2.6.32.
 862 .\"
 863 .TP
 864 .BR PR_GET_THP_DISABLE " (since Linux 3.15)"
 865 Return (via the function result) the current setting of the "THP disable"
 866 flag for the calling thread:
 867 either 1, if the flag is set, or 0, if it is not.
 868 .TP
 869 .BR PR_GET_TID_ADDRESS " (since Linux 3.5)"
 870 .\" commit 300f786b2683f8bb1ec0afb6e1851183a479c86d
 871 Retrieve the
 872 .I clear_child_tid
 873 address set by
 874 .BR set_tid_address (2)
 875 and the
 876 .BR clone (2)
 877 .B CLONE_CHILD_CLEARTID
 878 flag, in the location pointed to by
 879 .IR "(int\ **)\ arg2" .
 880 This feature is available only if the kernel is built with the
 881 .BR CONFIG_CHECKPOINT_RESTORE
 882 option enabled.
 883 .TP
 884 .BR PR_SET_TIMERSLACK " (since Linux 2.6.28)"
 885 .\" See https://lwn.net/Articles/369549/
 886 .\" commit 6976675d94042fbd446231d1bd8b7de71a980ada
 887 Each thread has two associated timer slack values:
 888 a "default" value, and a "current" value.
 889 This operation sets the "current" timer slack value for the calling thread.
 890 If the nanosecond value supplied in
 891 .IR arg2
 892 is greater than zero, then the "current" value is set to this value.
 893 If
 894 .I arg2
 895 is less than or equal to zero,
 896 .\" It seems that it's not possible to set the timer slack to zero;
 897 .\" The minimum value is 1? Seems a little strange.
 898 the "current" timer slack is reset to the
 899 thread's "default" timer slack value.
 900
 901 The "current" timer slack is used by the kernel to group timer expirations
 902 for the calling thread that are close to one another;
 903 as a consequence, timer expirations for the thread may be
 904 up to the specified number of nanoseconds late (but will never expire early).
 905 Grouping timer expirations can help reduce system power consumption
 906 by minimizing CPU wake-ups.
 907
 908 The timer expirations affected by timer slack are those set by
 909 .BR select (2),
 910 .BR pselect (2),
 911 .BR poll (2),
 912 .BR ppoll (2),
 913 .BR epoll_wait (2),
 914 .BR epoll_pwait (2),
 915 .BR clock_nanosleep (2),
 916 .BR nanosleep (2),
 917 and
 918 .BR futex (2)
 919 (and thus the library functions implemented via futexes, including
 920 .\" List obtained by grepping for futex usage in glibc source
 921 .BR pthread_cond_timedwait (3),
 922 .BR pthread_mutex_timedlock (3),
 923 .BR pthread_rwlock_timedrdlock (3),
 924 .BR pthread_rwlock_timedwrlock (3),
 925 and
 926 .BR sem_timedwait (3)).
 927
 928 Timer slack is not applied to threads that are scheduled under
 929 a real-time scheduling policy (see
 930 .BR sched_setscheduler (2)).
 931
 932 When a new thread is created,
 933 the two timer slack values are made the same as the "current" value
 934 of the creating thread.
 935 Thereafter, a thread can adjust its "current" timer slack value via
 936 .BR PR_SET_TIMERSLACK .
 937 The "default" value can't be changed.
 938 The timer slack values of
 939 .IR init
 940 (PID 1), the ancestor of all processes,
 941 are 50,000 nanoseconds (50 microseconds).
 942 The timer slack values are preserved across
 943 .BR execve (2).
 944
 945 Since Linux 4.6, the "current" timer slack value of any process
 946 can be examined and changed via the file
 947 .IR /proc/[pid]/timerslack_ns .
 948 See
 949 .BR proc (5).
 950 .TP
 951 .BR PR_GET_TIMERSLACK " (since Linux 2.6.28)"
 952 Return (as the function result)
 953 the "current" timer slack value of the calling thread.
 954 .TP
 955 .BR PR_SET_TIMING " (since Linux 2.6.0-test4)"
 956 Set whether to use (normal, traditional) statistical process timing or
 957 accurate timestamp-based process timing, by passing
 958 .B PR_TIMING_STATISTICAL
 959 .\" 0
 960 or
 961 .B PR_TIMING_TIMESTAMP
 962 .\" 1
 963 to \fIarg2\fP.
 964 .B PR_TIMING_TIMESTAMP
 965 is not currently implemented
 966 (attempting to set this mode will yield the error
 967 .BR EINVAL ).
 968 .\" PR_TIMING_TIMESTAMP doesn't do anything in 2.6.26-rc8,
 969 .\" and looking at the patch history, it appears
 970 .\" that it never did anything.
 971 .TP
 972 .BR PR_GET_TIMING " (since Linux 2.6.0-test4)"
 973 Return (as the function result) which process timing method is currently
 974 in use.
 975 .TP
 976 .BR PR_SET_TSC " (since Linux 2.6.26, x86 only)"
 977 Set the state of the flag determining whether the timestamp counter
 978 can be read by the process.
 979 Pass
 980 .B PR_TSC_ENABLE
 981 to
 982 .I arg2
 983 to allow it to be read, or
 984 .B PR_TSC_SIGSEGV
 985 to generate a
 986 .B SIGSEGV
 987 when the process tries to read the timestamp counter.
 988 .TP
 989 .BR PR_GET_TSC " (since Linux 2.6.26, x86 only)"
 990 Return the state of the flag determining whether the timestamp counter
 991 can be read,
 992 in the location pointed to by
 993 .IR "(int\ *) arg2" .
 994 .TP
 995 .B PR_SET_UNALIGN
 996 (Only on: ia64, since Linux 2.3.48; parisc, since Linux 2.6.15;
 997 PowerPC, since Linux 2.6.18; Alpha, since Linux 2.6.22)
 998 Set unaligned access control bits to \fIarg2\fP.
 999 Pass
1000 \fBPR_UNALIGN_NOPRINT\fP to silently fix up unaligned user accesses,
1001 or \fBPR_UNALIGN_SIGBUS\fP to generate
1002 .B SIGBUS
1003 on unaligned user access.
1004 .TP
1005 .B PR_GET_UNALIGN
1006 (see
1007 .B PR_SET_UNALIGN
1008 for information on versions and architectures)
1009 Return unaligned access control bits, in the location pointed to by
1010 .IR "(int\ *) arg2" .
1011 .SH RETURN VALUE
1012 On success,
1013 .BR PR_GET_DUMPABLE ,
1014 .BR PR_GET_KEEPCAPS ,
1015 .BR PR_GET_NO_NEW_PRIVS ,
1016 .BR PR_GET_THP_DISABLE ,
1017 .BR PR_CAPBSET_READ ,
1018 .BR PR_GET_TIMING ,
1019 .BR PR_GET_TIMERSLACK ,
1020 .BR PR_GET_SECUREBITS ,
1021 .BR PR_MCE_KILL_GET ,
1022 .BR PR_CAP_AMBIENT + PR_CAP_AMBIENT_IS_SET ,
1023 and (if it returns)
1024 .BR PR_GET_SECCOMP
1025 return the nonnegative values described above.
1026 All other
1027 .I option
1028 values return 0 on success.
1029 On error, \-1 is returned, and
1030 .I errno
1031 is set appropriately.
1032 .SH ERRORS
1033 .TP
1034 .B EACCES
1035 .I option
1036 is
1037 .BR PR_SET_MM ,
1038 and
1039 .I arg3
1040 is
1041 .BR PR_SET_MM_EXE_FILE ,
1042 the file is not executable.
1043 .TP
1044 .B EBADF
1045 .I option
1046 is
1047 .BR PR_SET_MM ,
1048 .I arg3
1049 is
1050 .BR PR_SET_MM_EXE_FILE ,
1051 and the file descriptor passed in
1052 .I arg4
1053 is not valid.
1054 .TP
1055 .B EBUSY
1056 .I option
1057 is
1058 .BR PR_SET_MM ,
1059 .I arg3
1060 is
1061 .BR PR_SET_MM_EXE_FILE ,
1062 and this the second attempt to change the
1063 .I /proc/pid/exe
1064 symbolic link, which is prohibited.
1065 .TP
1066 .B EFAULT
1067 .I arg2
1068 is an invalid address.
1069 .TP
1070 .B EFAULT
1071 .I option
1072 is
1073 .BR PR_SET_SECCOMP ,
1074 .I arg2
1075 is
1076 .BR SECCOMP_MODE_FILTER ,
1077 the system was built with
1078 .BR CONFIG_SECCOMP_FILTER ,
1079 and
1080 .I arg3
1081 is an invalid address.
1082 .TP
1083 .B EINVAL
1084 The value of
1085 .I option
1086 is not recognized.
1087 .TP
1088 .B EINVAL
1089 .I option
1090 is
1091 .BR PR_MCE_KILL
1092 or
1093 .BR PR_MCE_KILL_GET
1094 or
1095 .BR PR_SET_MM ,
1096 and unused
1097 .BR prctl ()
1098 arguments were not specified as zero.
1099 .TP
1100 .B EINVAL
1101 .I arg2
1102 is not valid value for this
1103 .IR option .
1104 .TP
1105 .B EINVAL
1106 .I option
1107 is
1108 .BR PR_SET_SECCOMP
1109 or
1110 .BR PR_GET_SECCOMP ,
1111 and the kernel was not configured with
1112 .BR CONFIG_SECCOMP .
1113 .TP
1114 .B EINVAL
1115 .I option
1116 is
1117 .BR PR_SET_SECCOMP ,
1118 .I arg2
1119 is
1120 .BR SECCOMP_MODE_FILTER ,
1121 and the kernel was not configured with
1122 .BR CONFIG_SECCOMP_FILTER .
1123 .TP
1124 .B EINVAL
1125 .I option
1126 is
1127 .BR PR_SET_MM ,
1128 and one of the following is true
1129 .RS
1130 .IP * 3
1131 .I arg4
1132 or
1133 .I arg5
1134 is nonzero;
1135 .IP *
1136 .I arg3
1137 is greater than
1138 .B TASK_SIZE
1139 (the limit on the size of the user address space for this architecture);
1140 .IP *
1141 .I arg2
1142 is
1143 .BR PR_SET_MM_START_CODE ,
1144 .BR PR_SET_MM_END_CODE ,
1145 .BR PR_SET_MM_START_DATA ,
1146 .BR PR_SET_MM_END_DATA ,
1147 or
1148 .BR PR_SET_MM_START_STACK ,
1149 and the permissions of the corresponding memory area are not as required;
1150 .IP *
1151 .I arg2
1152 is
1153 .BR PR_SET_MM_START_BRK
1154 or
1155 .BR PR_SET_MM_BRK ,
1156 and
1157 .I arg3
1158 is less than or equal to the end of the data segment
1159 or specifies a value that would cause the
1160 .B RLIMIT_DATA
1161 resource limit to be exceeded.
1162 .RE
1163 .TP
1164 .B EINVAL
1165 .I option
1166 is
1167 .BR PR_SET_PTRACER
1168 and
1169 .I arg2
1170 is not 0,
1171 .BR PR_SET_PTRACER_ANY ,
1172 or the PID of an existing process.
1173 .TP
1174 .B EINVAL
1175 .I option
1176 is
1177 .B PR_SET_PDEATHSIG
1178 and
1179 .I arg2
1180 is not a valid signal number.
1181 .TP
1182 .B EINVAL
1183 .I option
1184 is
1185 .BR PR_SET_DUMPABLE
1186 and
1187 .I arg2
1188 is neither
1189 .B SUID_DUMP_DISABLE
1190 nor
1191 .BR SUID_DUMP_USER .
1192 .TP
1193 .B EINVAL
1194 .I option
1195 is
1196 .BR PR_SET_TIMING
1197 and
1198 .I arg2
1199 is not
1200 .BR PR_TIMING_STATISTICAL .
1201 .TP
1202 .B EINVAL
1203 .I option
1204 is
1205 .BR PR_SET_NO_NEW_PRIVS
1206 and
1207 .I arg2
1208 is not equal to 1
1209 or
1210 .IR arg3 ,
1211 .IR arg4 ,
1212 or
1213 .IR arg5
1214 is nonzero.
1215 .TP
1216 .B EINVAL
1217 .I option
1218 is
1219 .BR PR_GET_NO_NEW_PRIVS
1220 and
1221 .IR arg2 ,
1222 .IR arg3 ,
1223 .IR arg4 ,
1224 or
1225 .IR arg5
1226 is nonzero.
1227 .TP
1228 .B EINVAL
1229 .I option
1230 is
1231 .BR PR_SET_THP_DISABLE
1232 and
1233 .IR arg3 ,
1234 .IR arg4 ,
1235 or
1236 .IR arg5
1237 is nonzero.
1238 .TP
1239 .B EINVAL
1240 .I option
1241 is
1242 .BR PR_GET_THP_DISABLE
1243 and
1244 .IR arg2 ,
1245 .IR arg3 ,
1246 .IR arg4 ,
1247 or
1248 .IR arg5
1249 is nonzero.
1250 .TP
1251 .B EINVAL
1252 .I option
1253 is
1254 .B PR_CAP_AMBIENT
1255 and an unused argument
1256 .RI ( arg4 ,
1257 .IR arg5 ,
1258 or,
1259 in the case of
1260 .BR PR_CAP_AMBIENT_CLEAR_ALL ,
1261 .IR arg3 )
1262 is nonzero; or
1263 .IR arg2
1264 has an invalid value;
1265 or
1266 .IR arg2
1267 is
1268 .BR PR_CAP_AMBIENT_LOWER ,
1269 .BR PR_CAP_AMBIENT_RAISE ,
1270 or
1271 .BR PR_CAP_AMBIENT_IS_SET
1272 and
1273 .IR arg3
1274 does not specify a valid capability.
1275 .TP
1276 .B ENXIO
1277 .I option
1278 was
1279 .BR PR_MPX_ENABLE_MANAGEMENT
1280 or
1281 .BR PR_MPX_DISABLE_MANAGEMENT
1282 and the kernel or the CPU does not support MPX management.
1283 Check that the kernel and processor have MPX support.
1284 .TP
1285 .B EPERM
1286 .I option
1287 is
1288 .BR PR_SET_SECUREBITS ,
1289 and the caller does not have the
1290 .B CAP_SETPCAP
1291 capability,
1292 or tried to unset a "locked" flag,
1293 or tried to set a flag whose corresponding locked flag was set
1294 (see
1295 .BR capabilities (7)).
1296 .TP
1297 .B EPERM
1298 .I option
1299 is
1300 .BR PR_SET_KEEPCAPS ,
1301 and the caller's
1302 .B SECURE_KEEP_CAPS_LOCKED
1303 flag is set
1304 (see
1305 .BR capabilities (7)).
1306 .TP
1307 .B EPERM
1308 .I option
1309 is
1310 .BR PR_CAPBSET_DROP ,
1311 and the caller does not have the
1312 .B CAP_SETPCAP
1313 capability.
1314 .TP
1315 .B EPERM
1316 .I option
1317 is
1318 .BR PR_SET_MM ,
1319 and the caller does not have the
1320 .B CAP_SYS_RESOURCE
1321 capability.
1322 .TP
1323 .B EPERM
1324 .IR option
1325 is
1326 .BR PR_CAP_AMBIENT
1327 and
1328 .IR arg2
1329 is
1330 .BR PR_CAP_AMBIENT_RAISE ,
1331 but either the capability specified in
1332 .IR arg3
1333 is not present in the process's permitted and inheritable capability sets,
1334 or the
1335 .B PR_CAP_AMBIENT_LOWER
1336 securebit has been set.
1337 .SH VERSIONS
1338 The
1339 .BR prctl ()
1340 system call was introduced in Linux 2.1.57.
1341 .\" The library interface was added in glibc 2.0.6
1342 .SH CONFORMING TO
1343 This call is Linux-specific.
1344 IRIX has a
1345 .BR prctl ()
1346 system call (also introduced in Linux 2.1.44
1347 as irix_prctl on the MIPS architecture),
1348 with prototype
1349 .sp
1350 .BI "ptrdiff_t prctl(int " option ", int " arg2 ", int " arg3 );
1351 .sp
1352 and options to get the maximum number of processes per user,
1353 get the maximum number of processors the calling process can use,
1354 find out whether a specified process is currently blocked,
1355 get or set the maximum stack size, and so on.
1356 .SH SEE ALSO
1357 .BR signal (2),
1358 .BR core (5)