Scroll to navigation

prctl(2) System Calls Manual prctl(2)

NAME

prctl - operations on a process or thread

LIBRARY

Standard C library (libc, -lc)

SYNOPSIS

#include <sys/prctl.h>
int prctl(int op, ...
          /* unsigned long arg2, unsigned long arg3,
          unsigned long arg4, unsigned long arg5 */ );

DESCRIPTION

prctl() manipulates various aspects of the behavior of the calling thread or process.

Note that careless use of some prctl() operations can confuse the user-space run-time environment, so these operations should be used with care.

prctl() is called with a first argument describing what to do (with values defined in <linux/prctl.h>), and further arguments with a significance depending on the first one. The first argument can be:

Reads or changes the ambient capability set of the calling thread, according to the value of arg2, which must be one of the following:
The capability specified in arg3 is added to the ambient set. The specified capability must already be present in both the permitted and the inheritable sets of the process. This operation is not permitted if the SECBIT_NO_CAP_AMBIENT_RAISE securebit is set.
The capability specified in arg3 is removed from the ambient set.
The prctl() call returns 1 if the capability in arg3 is in the ambient set and 0 if it is not.
All capabilities will be removed from the ambient set. This operation requires setting arg3 to zero.
In all of the above operations, arg4 and arg5 must be specified as 0.
Higher-level interfaces layered on top of the above operations are provided in the libcap(3) library in the form of cap_get_ambient(3), cap_set_ambient(3), and cap_reset_ambient(3).
Return (as the function result) 1 if the capability specified in arg2 is in the calling thread's capability bounding set, or 0 if it is not. (The capability constants are defined in <linux/capability.h>.) The capability bounding set dictates whether the process can receive the capability through a file's permitted capability set on a subsequent call to execve(2).
If the capability specified in arg2 is not valid, then the call fails with the error EINVAL.
A higher-level interface layered on top of this operation is provided in the libcap(3) library in the form of cap_get_bound(3).
If the calling thread has the CAP_SETPCAP capability within its user namespace, then drop the capability specified by arg2 from the calling thread's capability bounding set. Any children of the calling thread will inherit the newly reduced bounding set.
The call fails with the error: EPERM if the calling thread does not have the CAP_SETPCAP; EINVAL if arg2 does not represent a valid capability; or EINVAL if file capabilities are not enabled in the kernel, in which case bounding sets are not supported.
A higher-level interface layered on top of this operation is provided in the libcap(3) library in the form of cap_drop_bound(3).
If arg2 is nonzero, set the "child subreaper" attribute of the calling process; if arg2 is zero, unset the attribute.
A subreaper fulfills the role of init(1) for its descendant processes. When a process becomes orphaned (i.e., its immediate parent terminates), then that process will be reparented to the nearest still living ancestor subreaper. Subsequently, calls to getppid(2) in the orphaned process will now return the PID of the subreaper process, and when the orphan terminates, it is the subreaper process that will receive a SIGCHLD signal and will be able to wait(2) on the process to discover its termination status.
The setting of the "child subreaper" attribute is not inherited by children created by fork(2) and clone(2). The setting is preserved across execve(2).
Establishing a subreaper process is useful in session management frameworks where a hierarchical group of processes is managed by a subreaper process that needs to be informed when one of the processes—for example, a double-forked daemon—terminates (perhaps so that it can restart that process). Some init(1) frameworks (e.g., systemd(1)) employ a subreaper process for similar reasons.
Return the "child subreaper" setting of the caller, in the location pointed to by (int *) arg2.
Set the state of the "dumpable" attribute, which determines whether core dumps are produced for the calling process upon delivery of a signal whose default behavior is to produce a core dump.
Up to and including Linux 2.6.12, arg2 must be either 0 (SUID_DUMP_DISABLE, process is not dumpable) or 1 (SUID_DUMP_USER, process is dumpable). Between Linux 2.6.13 and Linux 2.6.17, the value 2 was also permitted, which caused any binary which normally would not be dumped to be dumped readable by root only; for security reasons, this feature has been removed. (See also the description of /proc/sys/fs/suid_dumpable in proc(5).)
Normally, the "dumpable" attribute is set to 1. However, it is reset to the current value contained in the file /proc/sys/fs/suid_dumpable (which by default has the value 0), in the following circumstances:
•
The process's effective user or group ID is changed.
•
The process's filesystem user or group ID is changed (see credentials(7)).
•
The process executes (execve(2)) a set-user-ID or set-group-ID program, resulting in a change of either the effective user ID or the effective group ID.
•
The process executes (execve(2)) a program that has file capabilities (see capabilities(7)), but only if the permitted capabilities gained exceed those already permitted for the process.
Processes that are not dumpable can not be attached via ptrace(2) PTRACE_ATTACH; see ptrace(2) for further details.
If a process is not dumpable, the ownership of files in the process's /proc/pid directory is affected as described in proc(5).
Return (as the function result) the current state of the calling process's dumpable attribute.
Set the endian-ness of the calling process to the value given in arg2, which should be one of the following: PR_ENDIAN_BIG, PR_ENDIAN_LITTLE, or PR_ENDIAN_PPC_LITTLE (PowerPC pseudo little endian).
Return the endian-ness of the calling process, in the location pointed to by (int *) arg2.
On the MIPS architecture, user-space code can be built using an ABI which permits linking with code that has more restrictive floating-point (FP) requirements. For example, user-space code may be built to target the O32 FPXX ABI and linked with code built for either one of the more restrictive FP32 or FP64 ABIs. When more restrictive code is linked in, the overall requirement for the process is to use the more restrictive floating-point mode.
Because the kernel has no means of knowing in advance which mode the process should be executed in, and because these restrictions can change over the lifetime of the process, the PR_SET_FP_MODE operation is provided to allow control of the floating-point mode from user space.
The (unsigned int) arg2 argument is a bit mask describing the floating-point mode used:
When this bit is unset (so called FR=0 or FR0 mode), the 32 floating-point registers are 32 bits wide, and 64-bit registers are represented as a pair of registers (even- and odd- numbered, with the even-numbered register containing the lower 32 bits, and the odd-numbered register containing the higher 32 bits).
When this bit is set (on supported hardware), the 32 floating-point registers are 64 bits wide (so called FR=1 or FR1 mode). Note that modern MIPS implementations (MIPS R6 and newer) support FR=1 mode only.
Applications that use the O32 FP32 ABI can operate only when this bit is unset (FR=0; or they can be used with FRE enabled, see below). Applications that use the O32 FP64 ABI (and the O32 FP64A ABI, which exists to provide the ability to operate with existing FP32 code; see below) can operate only when this bit is set (FR=1). Applications that use the O32 FPXX ABI can operate with either FR=0 or FR=1.
Enable emulation of 32-bit floating-point mode. When this mode is enabled, it emulates 32-bit floating-point operations by raising a reserved-instruction exception on every instruction that uses 32-bit formats and the kernel then handles the instruction in software. (The problem lies in the discrepancy of handling odd-numbered registers which are the high 32 bits of 64-bit registers with even numbers in FR=0 mode and the lower 32-bit parts of odd-numbered 64-bit registers in FR=1 mode.) Enabling this bit is necessary when code with the O32 FP32 ABI should operate with code with compatible the O32 FPXX or O32 FP64A ABIs (which require FR=1 FPU mode) or when it is executed on newer hardware (MIPS R6 onwards) which lacks FR=0 mode support when a binary with the FP32 ABI is used.
Note that this mode makes sense only when the FPU is in 64-bit mode (FR=1).
Note that the use of emulation inherently has a significant performance hit and should be avoided if possible.
In the N32/N64 ABI, 64-bit floating-point mode is always used, so FPU emulation is not required and the FPU always operates in FR=1 mode.
This operation is mainly intended for use by the dynamic linker (ld.so(8)).
The arguments arg3, arg4, and arg5 are ignored.
Return (as the function result) the current floating-point mode (see the description of PR_SET_FP_MODE for details).
On success, the call returns a bit mask which represents the current floating-point mode.
The arguments arg2, arg3, arg4, and arg5 are ignored.
Set floating-point emulation control bits to arg2. Pass PR_FPEMU_NOPRINT to silently emulate floating-point operation accesses, or PR_FPEMU_SIGFPE to not emulate floating-point operations and send SIGFPE instead.
Return floating-point emulation control bits, in the location pointed to by (int *) arg2.
Set floating-point exception mode to arg2. Pass PR_FP_EXC_SW_ENABLE to use FPEXC for FP exception enables, PR_FP_EXC_DIV for floating-point divide by zero, PR_FP_EXC_OVF for floating-point overflow, PR_FP_EXC_UND for floating-point underflow, PR_FP_EXC_RES for floating-point inexact result, PR_FP_EXC_INV for floating-point invalid operation, PR_FP_EXC_DISABLED for FP exceptions disabled, PR_FP_EXC_NONRECOV for async nonrecoverable exception mode, PR_FP_EXC_ASYNC for async recoverable exception mode, PR_FP_EXC_PRECISE for precise exception mode.
Return floating-point exception mode, in the location pointed to by (int *) arg2.
If a user process is involved in the block layer or filesystem I/O path, and can allocate memory while processing I/O requests it must set arg2 to 1. This will put the process in the IO_FLUSHER state, which allows it special treatment to make progress when allocating memory. If arg2 is 0, the process will clear the IO_FLUSHER state, and the default behavior will be used.
The calling process must have the CAP_SYS_RESOURCE capability.
arg3, arg4, and arg5 must be zero.
The IO_FLUSHER state is inherited by a child process created via fork(2) and is preserved across execve(2).
Examples of IO_FLUSHER applications are FUSE daemons, SCSI device emulation daemons, and daemons that perform error handling like multipath path recovery applications.
Return (as the function result) the IO_FLUSHER state of the caller. A value of 1 indicates that the caller is in the IO_FLUSHER state; 0 indicates that the caller is not in the IO_FLUSHER state.
The calling process must have the CAP_SYS_RESOURCE capability.
arg2, arg3, arg4, and arg5 must be zero.
Set the state of the calling thread's "keep capabilities" flag. The effect of this flag is described in capabilities(7). arg2 must be either 0 (clear the flag) or 1 (set the flag). The "keep capabilities" value will be reset to 0 on subsequent calls to execve(2).
Return (as the function result) the current state of the calling thread's "keep capabilities" flag. See capabilities(7) for a description of this flag.
Set the machine check memory corruption kill policy for the calling thread. If arg2 is PR_MCE_KILL_CLEAR, clear the thread memory corruption kill policy and use the system-wide default. (The system-wide default is defined by /proc/sys/vm/memory_failure_early_kill; see proc(5).) If arg2 is PR_MCE_KILL_SET, use a thread-specific memory corruption kill policy. In this case, arg3 defines whether the policy is early kill (PR_MCE_KILL_EARLY), late kill (PR_MCE_KILL_LATE), or the system-wide default (PR_MCE_KILL_DEFAULT). Early kill means that the thread receives a SIGBUS signal as soon as hardware memory corruption is detected inside its address space. In late kill mode, the process is killed only when it accesses a corrupted page. See sigaction(2) for more information on the SIGBUS signal. The policy is inherited by children. The remaining unused prctl() arguments must be zero for future compatibility.
Return (as the function result) the current per-process machine check kill policy. All unused prctl() arguments must be zero.
Modify certain kernel memory map descriptor fields of the calling process. Usually these fields are set by the kernel and dynamic loader (see ld.so(8) for more information) and a regular application should not use this feature. However, there are cases, such as self-modifying programs, where a program might find it useful to change its own memory map.
The calling process must have the CAP_SYS_RESOURCE capability. The value in arg2 is one of the options below, while arg3 provides a new value for the option. The arg4 and arg5 arguments must be zero if unused.
Before Linux 3.10, this feature is available only if the kernel is built with the CONFIG_CHECKPOINT_RESTORE option enabled.
Set the address above which the program text can run. The corresponding memory area must be readable and executable, but not writable or shareable (see mprotect(2) and mmap(2) for more information).
Set the address below which the program text can run. The corresponding memory area must be readable and executable, but not writable or shareable.
Set the address above which initialized and uninitialized (bss) data are placed. The corresponding memory area must be readable and writable, but not executable or shareable.
Set the address below which initialized and uninitialized (bss) data are placed. The corresponding memory area must be readable and writable, but not executable or shareable.
Set the start address of the stack. The corresponding memory area must be readable and writable.
Set the address above which the program heap can be expanded with brk(2) call. The address must be greater than the ending address of the current program data segment. In addition, the combined size of the resulting heap and the size of the data segment can't exceed the RLIMIT_DATA resource limit (see setrlimit(2)).
Set the current brk(2) value. The requirements for the address are the same as for the PR_SET_MM_START_BRK option.

The following options are available since Linux 3.5.

Set the address above which the program command line is placed.
Set the address below which the program command line is placed.
Set the address above which the program environment is placed.
Set the address below which the program environment is placed.
The address passed with PR_SET_MM_ARG_START, PR_SET_MM_ARG_END, PR_SET_MM_ENV_START, and PR_SET_MM_ENV_END should belong to a process stack area. Thus, the corresponding memory area must be readable, writable, and (depending on the kernel configuration) have the MAP_GROWSDOWN attribute set (see mmap(2)).
Set a new auxiliary vector. The arg3 argument should provide the address of the vector. The arg4 is the size of the vector.
Supersede the /proc/pid/exe symbolic link with a new one pointing to a new executable file identified by the file descriptor provided in arg3 argument. The file descriptor should be obtained with a regular open(2) call.
To change the symbolic link, one needs to unmap all existing executable memory areas, including those created by the kernel itself (for example the kernel usually creates at least one executable memory area for the ELF .text section).
In Linux 4.9 and earlier, the PR_SET_MM_EXE_FILE operation can be performed only once in a process's lifetime; attempting to perform the operation a second time results in the error EPERM. This restriction was enforced for security reasons that were subsequently deemed specious, and the restriction was removed in Linux 4.10 because some user-space applications needed to perform this operation more than once.

The following options are available since Linux 3.18.

Provides one-shot access to all the addresses by passing in a struct prctl_mm_map (as defined in <linux/prctl.h>). The arg4 argument should provide the size of the struct.
This feature is available only if the kernel is built with the CONFIG_CHECKPOINT_RESTORE option enabled.
Returns the size of the struct prctl_mm_map the kernel expects. This allows user space to find a compatible struct. The arg4 argument should be a pointer to an unsigned int.
This feature is available only if the kernel is built with the CONFIG_CHECKPOINT_RESTORE option enabled.
Sets an attribute specified in arg2 for virtual memory areas starting from the address specified in arg3 and spanning the size specified in arg4. arg5 specifies the value of the attribute to be set.
Note that assigning an attribute to a virtual memory area might prevent it from being merged with adjacent virtual memory areas due to the difference in that attribute's value.
Currently, arg2 must be one of:
Set a name for anonymous virtual memory areas. arg5 should be a pointer to a null-terminated string containing the name. The name length including null byte cannot exceed 80 bytes. If arg5 is NULL, the name of the appropriate anonymous virtual memory areas will be reset. The name can contain only printable ascii characters (including space), except '[', ']', '\', '$', and '`'.
Enable or disable kernel management of Memory Protection eXtensions (MPX) bounds tables. The arg2, arg3, arg4, and arg5 arguments must be zero.
MPX is a hardware-assisted mechanism for performing bounds checking on pointers. It consists of a set of registers storing bounds information and a set of special instruction prefixes that tell the CPU on which instructions it should do bounds enforcement. There is a limited number of these registers and when there are more pointers than registers, their contents must be "spilled" into a set of tables. These tables are called "bounds tables" and the MPX prctl() operations control whether the kernel manages their allocation and freeing.
When management is enabled, the kernel will take over allocation and freeing of the bounds tables. It does this by trapping the #BR exceptions that result at first use of missing bounds tables and instead of delivering the exception to user space, it allocates the table and populates the bounds directory with the location of the new table. For freeing, the kernel checks to see if bounds tables are present for memory which is not allocated, and frees them if so.
Before enabling MPX management using PR_MPX_ENABLE_MANAGEMENT, the application must first have allocated a user-space buffer for the bounds directory and placed the location of that directory in the bndcfgu register.
These calls fail if the CPU or kernel does not support MPX. Kernel support for MPX is enabled via the CONFIG_X86_INTEL_MPX configuration option. You can check whether the CPU supports MPX by looking for the mpx CPUID bit, like with the following command:

cat /proc/cpuinfo | grep ' mpx '

A thread may not switch in or out of long (64-bit) mode while MPX is enabled.
All threads in a process are affected by these calls.
The child of a fork(2) inherits the state of MPX management. During execve(2), MPX management is reset to a state as if PR_MPX_DISABLE_MANAGEMENT had been called.
For further information on Intel MPX, see the kernel source file Documentation/x86/intel_mpx.txt.
Due to a lack of toolchain support, PR_MPX_ENABLE_MANAGEMENT and PR_MPX_DISABLE_MANAGEMENT are not supported in Linux 5.4 and later.
Set the name of the calling thread, using the value in the location pointed to by (char *) arg2. The name can be up to 16 bytes long, including the terminating null byte. (If the length of the string, including the terminating null byte, exceeds 16 bytes, the string is silently truncated.) This is the same attribute that can be set via pthread_setname_np(3) and retrieved using pthread_getname_np(3). The attribute is likewise accessible via /proc/self/task/tid/comm (see proc(5)), where tid is the thread ID of the calling thread, as returned by gettid(2).
Return the name of the calling thread, in the buffer pointed to by (char *) arg2. The buffer should allow space for up to 16 bytes; the returned string will be null-terminated.
Set the calling thread's no_new_privs attribute to the value in arg2. With no_new_privs set to 1, execve(2) promises not to grant privileges to do anything that could not have been done without the execve(2) call (for example, rendering the set-user-ID and set-group-ID mode bits, and file capabilities non-functional). Once set, the no_new_privs attribute cannot be unset. The setting of this attribute is inherited by children created by fork(2) and clone(2), and preserved across execve(2).
Since Linux 4.10, the value of a thread's no_new_privs attribute can be viewed via the NoNewPrivs field in the /proc/pid/status file.
For more information, see the kernel source file Documentation/userspace-api/no_new_privs.rst (or Documentation/prctl/no_new_privs.txt before Linux 4.13). See also seccomp(2).
Return (as the function result) the value of the no_new_privs attribute for the calling thread. A value of 0 indicates the regular execve(2) behavior. A value of 1 indicates execve(2) will operate in the privilege-restricting mode described above.
Securely reset the thread's pointer authentication keys to fresh random values generated by the kernel.
The set of keys to be reset is specified by arg2, which must be a logical OR of zero or more of the following:
instruction authentication key A
instruction authentication key B
data authentication key A
data authentication key B
generic authentication “A” key.
(Yes folks, there really is no generic B key.)
As a special case, if arg2 is zero, then all the keys are reset. Since new keys could be added in future, this is the recommended way to completely wipe the existing keys when establishing a clean execution context. Note that there is no need to use PR_PAC_RESET_KEYS in preparation for calling execve(2), since execve(2) resets all the pointer authentication keys.
The remaining arguments arg3, arg4, and arg5 must all be zero.
If the arguments are invalid, and in particular if arg2 contains set bits that are unrecognized or that correspond to a key not available on this platform, then the call fails with error EINVAL.
Warning: Because the compiler or run-time environment may be using some or all of the keys, a successful PR_PAC_RESET_KEYS may crash the calling process. The conditions for using it safely are complex and system-dependent. Don't use it unless you know what you are doing.
For more information, see the kernel source file Documentation/arm64/pointer-authentication.rst (or Documentation/arm64/pointer-authentication.txt before Linux 5.3).
Set the parent-death signal of the calling process to arg2 (either a signal value in the range [1, NSIG - 1], or 0 to clear). This is the signal that the calling process will get when its parent dies.
Warning: the "parent" in this case is considered to be the thread that created this process. In other words, the signal will be sent when that thread terminates (via, for example, pthread_exit(3)), rather than after all of the threads in the parent process terminate.
The parent-death signal is sent upon subsequent termination of the parent thread and also upon termination of each subreaper process (see the description of PR_SET_CHILD_SUBREAPER above) to which the caller is subsequently reparented. If the parent thread and all ancestor subreapers have already terminated by the time of the PR_SET_PDEATHSIG operation, then no parent-death signal is sent to the caller.
The parent-death signal is process-directed (see signal(7)) and, if the child installs a handler using the sigaction(2) SA_SIGINFO flag, the si_pid field of the siginfo_t argument of the handler contains the PID of the terminating parent process.
The parent-death signal setting is cleared for the child of a fork(2). It is also (since Linux 2.4.36 / 2.6.23) cleared when executing a set-user-ID or set-group-ID binary, or a binary that has associated capabilities (see capabilities(7)); otherwise, this value is preserved across execve(2). The parent-death signal setting is also cleared upon changes to any of the following thread credentials: effective user ID, effective group ID, filesystem user ID, or filesystem group ID.
Return the current value of the parent process death signal, in the location pointed to by (int *) arg2.
This is meaningful only when the Yama LSM is enabled and in mode 1 ("restricted ptrace", visible via /proc/sys/kernel/yama/ptrace_scope). When a "ptracer process ID" is passed in arg2, the caller is declaring that the ptracer process can ptrace(2) the calling process as if it were a direct process ancestor. Each PR_SET_PTRACER operation replaces the previous "ptracer process ID". Employing PR_SET_PTRACER with arg2 set to 0 clears the caller's "ptracer process ID". If arg2 is PR_SET_PTRACER_ANY, the ptrace restrictions introduced by Yama are effectively disabled for the calling process.
For further information, see the kernel source file Documentation/admin-guide/LSM/Yama.rst (or Documentation/security/Yama.txt before Linux 4.13).
Set the secure computing (seccomp) mode for the calling thread, to limit the available system calls. The more recent seccomp(2) system call provides a superset of the functionality of PR_SET_SECCOMP, and is the preferred interface for new applications.
The seccomp mode is selected via arg2. (The seccomp constants are defined in <linux/seccomp.h>.) The following values can be specified:
See the description of SECCOMP_SET_MODE_STRICT in seccomp(2).
This operation is available only if the kernel is configured with CONFIG_SECCOMP enabled.
The allowed system calls are defined by a pointer to a Berkeley Packet Filter passed in arg3. This argument is a pointer to struct sock_fprog; it can be designed to filter arbitrary system calls and system call arguments. See the description of SECCOMP_SET_MODE_FILTER in seccomp(2).
This operation is available only if the kernel is configured with CONFIG_SECCOMP_FILTER enabled.
For further details on seccomp filtering, see seccomp(2).
Return (as the function result) the secure computing mode of the calling thread. If the caller is not in secure computing mode, this operation returns 0; if the caller is in strict secure computing mode, then the prctl() call will cause a SIGKILL signal to be sent to the process. If the caller is in filter mode, and this system call is allowed by the seccomp filters, it returns 2; otherwise, the process is killed with a SIGKILL signal.
This operation is available only if the kernel is configured with CONFIG_SECCOMP enabled.
Since Linux 3.8, the Seccomp field of the /proc/pid/status file provides a method of obtaining the same information, without the risk that the process is killed; see proc(5).
Set the "securebits" flags of the calling thread to the value supplied in arg2. See capabilities(7).
Return (as the function result) the "securebits" flags of the calling thread. See capabilities(7).
Return (as the function result) the state of the speculation misfeature specified in arg2. Currently, the only permitted value for this argument is PR_SPEC_STORE_BYPASS (otherwise the call fails with the error ENODEV).
The return value uses bits 0-3 with the following meaning:
Mitigation can be controlled per thread by PR_SET_SPECULATION_CTRL.
The speculation feature is enabled, mitigation is disabled.
The speculation feature is disabled, mitigation is enabled.
Same as PR_SPEC_DISABLE but cannot be undone.
Same as PR_SPEC_DISABLE, but the state will be cleared on execve(2).
If all bits are 0, then the CPU is not affected by the speculation misfeature.
If PR_SPEC_PRCTL is set, then per-thread control of the mitigation is available. If not set, prctl() for the speculation misfeature will fail.
The arg3, arg4, and arg5 arguments must be specified as 0; otherwise the call fails with the error EINVAL.
Sets the state of the speculation misfeature specified in arg2. The speculation-misfeature settings are per-thread attributes.
Currently, arg2 must be one of:
Set the state of the speculative store bypass misfeature.
Set the state of the indirect branch speculation misfeature.
If arg2 does not have one of the above values, then the call fails with the error ENODEV.
The arg3 argument is used to hand in the control value, which is one of the following:
The speculation feature is enabled, mitigation is disabled.
The speculation feature is disabled, mitigation is enabled.
Same as PR_SPEC_DISABLE, but cannot be undone. A subsequent prctl(arg2, PR_SPEC_ENABLE) with the same value for arg2 will fail with the error EPERM.
Same as PR_SPEC_DISABLE, but the state will be cleared on execve(2). Currently only supported for arg2 equal to PR_SPEC_STORE_BYPASS.
Any unsupported value in arg3 will result in the call failing with the error ERANGE.
The arg4 and arg5 arguments must be specified as 0; otherwise the call fails with the error EINVAL.
The speculation feature can also be controlled by the spec_store_bypass_disable boot parameter. This parameter may enforce a read-only policy which will result in the prctl() call failing with the error ENXIO. For further details, see the kernel source file Documentation/admin-guide/kernel-parameters.txt.
Configure the thread's SVE vector length, as specified by (int) arg2. Arguments arg3, arg4, and arg5 are ignored.
The bits of arg2 corresponding to PR_SVE_VL_LEN_MASK must be set to the desired vector length in bytes. This is interpreted as an upper bound: the kernel will select the greatest available vector length that does not exceed the value specified. In particular, specifying SVE_VL_MAX (defined in <asm/sigcontext.h>) for the PR_SVE_VL_LEN_MASK bits requests the maximum supported vector length.
In addition, the other bits of arg2 must be set to one of the following combinations of flags:
0
Perform the change immediately. At the next execve(2) in the thread, the vector length will be reset to the value configured in /proc/sys/abi/sve_default_vector_length.
Perform the change immediately. Subsequent execve(2) calls will preserve the new vector length.
Defer the change, so that it is performed at the next execve(2) in the thread. Further execve(2) calls will reset the vector length to the value configured in /proc/sys/abi/sve_default_vector_length.
PR_SVE_SET_VL_ONEXEC | PR_SVE_VL_INHERIT
Defer the change, so that it is performed at the next execve(2) in the thread. Further execve(2) calls will preserve the new vector length.
In all cases, any previously pending deferred change is canceled.
The call fails with error EINVAL if SVE is not supported on the platform, if arg2 is unrecognized or invalid, or the value in the bits of arg2 corresponding to PR_SVE_VL_LEN_MASK is outside the range SVE_VL_MIN..SVE_VL_MAX or is not a multiple of 16.
On success, a nonnegative value is returned that describes the selected configuration. If PR_SVE_SET_VL_ONEXEC was included in arg2, then the configuration described by the return value will take effect at the next execve(2). Otherwise, the configuration is already in effect when the PR_SVE_SET_VL call returns. In either case, the value is encoded in the same way as the return value of PR_SVE_GET_VL. Note that there is no explicit flag in the return value corresponding to PR_SVE_SET_VL_ONEXEC.
The configuration (including any pending deferred change) is inherited across fork(2) and clone(2).
For more information, see the kernel source file Documentation/arm64/sve.rst (or Documentation/arm64/sve.txt before Linux 5.3).
Warning: Because the compiler or run-time environment may be using SVE, using this call without the PR_SVE_SET_VL_ONEXEC flag may crash the calling process. The conditions for using it safely are complex and system-dependent. Don't use it unless you really know what you are doing.
Get the thread's current SVE vector length configuration.
Arguments arg2, arg3, arg4, and arg5 are ignored.
Provided that the kernel and platform support SVE, this operation always succeeds, returning a nonnegative value that describes the current configuration. The bits corresponding to PR_SVE_VL_LEN_MASK contain the currently configured vector length in bytes. The bit corresponding to PR_SVE_VL_INHERIT indicates whether the vector length will be inherited across execve(2).
Note that there is no way to determine whether there is a pending vector length change that has not yet taken effect.
For more information, see the kernel source file Documentation/arm64/sve.rst (or Documentation/arm64/sve.txt before Linux 5.3).
Configure the Syscall User Dispatch mechanism for the calling thread. This mechanism allows an application to selectively intercept system calls so that they can be handled within the application itself. Interception takes the form of a thread-directed SIGSYS signal that is delivered to the thread when it makes a system call. If intercepted, the system call is not executed by the kernel.
To enable this mechanism, arg2 should be set to PR_SYS_DISPATCH_ON. Once enabled, further system calls will be selectively intercepted, depending on a control variable provided by user space. In this case, arg3 and arg4 respectively identify the offset and length of a single contiguous memory region in the process address space from where system calls are always allowed to be executed, regardless of the control variable. (Typically, this area would include the area of memory containing the C library.)
arg5 points to a char-sized variable that is a fast switch to allow/block system call execution without the overhead of doing another system call to re-configure Syscall User Dispatch. This control variable can either be set to SYSCALL_DISPATCH_FILTER_BLOCK to block system calls from executing or to SYSCALL_DISPATCH_FILTER_ALLOW to temporarily allow them to be executed. This value is checked by the kernel on every system call entry, and any unexpected value will raise an uncatchable SIGSYS at that time, killing the application.
When a system call is intercepted, the kernel sends a thread-directed SIGSYS signal to the triggering thread. Various fields will be set in the siginfo_t structure (see sigaction(2)) associated with the signal:
•
si_signo will contain SIGSYS.
•
si_call_addr will show the address of the system call instruction.
•
si_syscall and si_arch will indicate which system call was attempted.
•
si_code will contain SYS_USER_DISPATCH.
•
si_errno will be set to 0.
The program counter will be as though the system call happened (i.e., the program counter will not point to the system call instruction).
When the signal handler returns to the kernel, the system call completes immediately and returns to the calling thread, without actually being executed. If necessary (i.e., when emulating the system call on user space.), the signal handler should set the system call return value to a sane value, by modifying the register context stored in the ucontext argument of the signal handler. See sigaction(2), sigreturn(2), and getcontext(3) for more information.
If arg2 is set to PR_SYS_DISPATCH_OFF, Syscall User Dispatch is disabled for that thread. the remaining arguments must be set to 0.
The setting is not preserved across fork(2), clone(2), or execve(2).
For more information, see the kernel source file Documentation/admin-guide/syscall-user-dispatch.rst
Controls support for passing tagged user-space addresses to the kernel (i.e., addresses where bits 56—63 are not all zero).
The level of support is selected by arg2, which can be one of the following:
0
Addresses that are passed for the purpose of being dereferenced by the kernel must be untagged.
Addresses that are passed for the purpose of being dereferenced by the kernel may be tagged, with the exceptions summarized below.
The remaining arguments arg3, arg4, and arg5 must all be zero.
On success, the mode specified in arg2 is set for the calling thread and the return value is 0. If the arguments are invalid, the mode specified in arg2 is unrecognized, or if this feature is unsupported by the kernel or disabled via /proc/sys/abi/tagged_addr_disabled, the call fails with the error EINVAL.
In particular, if prctl(PR_SET_TAGGED_ADDR_CTRL, 0, 0, 0, 0) fails with EINVAL, then all addresses passed to the kernel must be untagged.
Irrespective of which mode is set, addresses passed to certain interfaces must always be untagged:
•
brk(2), mmap(2), shmat(2), shmdt(2), and the new_address argument of mremap(2).
(Prior to Linux 5.6 these accepted tagged addresses, but the behaviour may not be what you expect. Don't rely on it.)
•
‘polymorphic’ interfaces that accept pointers to arbitrary types cast to a void * or other generic type, specifically prctl(), ioctl(2), and in general setsockopt(2) (only certain specific setsockopt(2) options allow tagged addresses).
This list of exclusions may shrink when moving from one kernel version to a later kernel version. While the kernel may make some guarantees for backwards compatibility reasons, for the purposes of new software the effect of passing tagged addresses to these interfaces is unspecified.
The mode set by this call is inherited across fork(2) and clone(2). The mode is reset by execve(2) to 0 (i.e., tagged addresses not permitted in the user/kernel ABI).
For more information, see the kernel source file Documentation/arm64/tagged-address-abi.rst.
Warning: This call is primarily intended for use by the run-time environment. A successful PR_SET_TAGGED_ADDR_CTRL call elsewhere may crash the calling process. The conditions for using it safely are complex and system-dependent. Don't use it unless you know what you are doing.
Returns the current tagged address mode for the calling thread.
Arguments arg2, arg3, arg4, and arg5 must all be zero.
If the arguments are invalid or this feature is disabled or unsupported by the kernel, the call fails with EINVAL. In particular, if prctl(PR_GET_TAGGED_ADDR_CTRL, 0, 0, 0, 0) fails with EINVAL, then this feature is definitely either unsupported, or disabled via /proc/sys/abi/tagged_addr_disabled. In this case, all addresses passed to the kernel must be untagged.
Otherwise, the call returns a nonnegative value describing the current tagged address mode, encoded in the same way as the arg2 argument of PR_SET_TAGGED_ADDR_CTRL.
For more information, see the kernel source file Documentation/arm64/tagged-address-abi.rst.
Disable all performance counters attached to the calling process, regardless of whether the counters were created by this process or another process. Performance counters created by the calling process for other processes are unaffected. For more information on performance counters, see the Linux kernel source file tools/perf/design.txt.
Originally called PR_TASK_PERF_COUNTERS_DISABLE; renamed (retaining the same numerical value) in Linux 2.6.32.
The converse of PR_TASK_PERF_EVENTS_DISABLE; enable performance counters attached to the calling process.
Originally called PR_TASK_PERF_COUNTERS_ENABLE; renamed in Linux 2.6.32.
Set the state of the "THP disable" flag for the calling thread. If arg2 has a nonzero value, the flag is set, otherwise it is cleared. Setting this flag provides a method for disabling transparent huge pages for jobs where the code cannot be modified, and using a malloc hook with madvise(2) is not an option (i.e., statically allocated data). The setting of the "THP disable" flag is inherited by a child created via fork(2) and is preserved across execve(2).
Return (as the function result) the current setting of the "THP disable" flag for the calling thread: either 1, if the flag is set, or 0, if it is not.
Return the clear_child_tid address set by set_tid_address(2) and the clone(2) CLONE_CHILD_CLEARTID flag, in the location pointed to by (int **) arg2. This feature is available only if the kernel is built with the CONFIG_CHECKPOINT_RESTORE option enabled. Note that since the prctl() system call does not have a compat implementation for the AMD64 x32 and MIPS n32 ABIs, and the kernel writes out a pointer using the kernel's pointer size, this operation expects a user-space buffer of 8 (not 4) bytes on these ABIs.
Each thread has two associated timer slack values: a "default" value, and a "current" value. This operation sets the "current" timer slack value for the calling thread. arg2 is an unsigned long value, then maximum "current" value is ULONG_MAX and the minimum "current" value is 1. If the nanosecond value supplied in arg2 is greater than zero, then the "current" value is set to this value. If arg2 is equal to zero, the "current" timer slack is reset to the thread's "default" timer slack value.
The "current" timer slack is used by the kernel to group timer expirations for the calling thread that are close to one another; as a consequence, timer expirations for the thread may be up to the specified number of nanoseconds late (but will never expire early). Grouping timer expirations can help reduce system power consumption by minimizing CPU wake-ups.
The timer expirations affected by timer slack are those set by select(2), pselect(2), poll(2), ppoll(2), epoll_wait(2), epoll_pwait(2), clock_nanosleep(2), nanosleep(2), and futex(2) (and thus the library functions implemented via futexes, including pthread_cond_timedwait(3), pthread_mutex_timedlock(3), pthread_rwlock_timedrdlock(3), pthread_rwlock_timedwrlock(3), and sem_timedwait(3)).
Timer slack is not applied to threads that are scheduled under a real-time scheduling policy (see sched_setscheduler(2)).
When a new thread is created, the two timer slack values are made the same as the "current" value of the creating thread. Thereafter, a thread can adjust its "current" timer slack value via PR_SET_TIMERSLACK. The "default" value can't be changed. The timer slack values of init (PID 1), the ancestor of all processes, are 50,000 nanoseconds (50 microseconds). The timer slack value is inherited by a child created via fork(2), and is preserved across execve(2).
Since Linux 4.6, the "current" timer slack value of any process can be examined and changed via the file /proc/pid/timerslack_ns. See proc(5).
Return (as the function result) the "current" timer slack value of the calling thread.
Set whether to use (normal, traditional) statistical process timing or accurate timestamp-based process timing, by passing PR_TIMING_STATISTICAL or PR_TIMING_TIMESTAMP to arg2. PR_TIMING_TIMESTAMP is not currently implemented (attempting to set this mode will yield the error EINVAL).
Return (as the function result) which process timing method is currently in use.
Set the state of the flag determining whether the timestamp counter can be read by the process. Pass PR_TSC_ENABLE to arg2 to allow it to be read, or PR_TSC_SIGSEGV to generate a SIGSEGV when the process tries to read the timestamp counter.
Return the state of the flag determining whether the timestamp counter can be read, in the location pointed to by (int *) arg2.
(Only on: ia64, since Linux 2.3.48; parisc, since Linux 2.6.15; PowerPC, since Linux 2.6.18; Alpha, since Linux 2.6.22; sh, since Linux 2.6.34; tile, since Linux 3.12) Set unaligned access control bits to arg2. Pass PR_UNALIGN_NOPRINT to silently fix up unaligned user accesses, or PR_UNALIGN_SIGBUS to generate SIGBUS on unaligned user access. Alpha also supports an additional flag with the value of 4 and no corresponding named constant, which instructs kernel to not fix up unaligned accesses (it is analogous to providing the UAC_NOFIX flag in SSI_NVPAIRS operation of the setsysinfo() system call on Tru64).
(See PR_SET_UNALIGN for information on versions and architectures.) Return unaligned access control bits, in the location pointed to by (unsigned int *) arg2.
Get the auxiliary vector (auxv) into the buffer pointed to by (void *) arg2, whose length is given by arg3. If the buffer is not long enough for the full auxiliary vector, the copy will be truncated. Return (as the function result) the full length of the auxiliary vector. arg4 and arg5 must be 0.
Set the calling process' Memory-Deny-Write-Execute protection mask. Once protection bits are set, they can not be changed. arg2 must be a bit mask of:
New memory mapping protections can't be writable and executable. Non-executable mappings can't become executable.
Do not propagate MDWE protection to child processes on fork(2). Setting this bit requires setting PR_MDWE_REFUSE_EXEC_GAIN too.
Return (as the function result) the Memory-Deny-Write-Execute protection mask of the calling process. (See PR_SET_MDWE for information on the protection mask bits.)

RETURN VALUE

On success, PR_CAP_AMBIENT+PR_CAP_AMBIENT_IS_SET, PR_CAPBSET_READ, PR_GET_DUMPABLE, PR_GET_FP_MODE, PR_GET_IO_FLUSHER, PR_GET_KEEPCAPS, PR_MCE_KILL_GET, PR_GET_NO_NEW_PRIVS, PR_GET_SECUREBITS, PR_GET_SPECULATION_CTRL, PR_SVE_GET_VL, PR_SVE_SET_VL, PR_GET_TAGGED_ADDR_CTRL, PR_GET_THP_DISABLE, PR_GET_TIMING, PR_GET_TIMERSLACK, PR_GET_AUXV, and (if it returns) PR_GET_SECCOMP return the nonnegative values described above. All other op values return 0 on success. On error, -1 is returned, and errno is set to indicate the error.

ERRORS

op is PR_SET_SECCOMP and arg2 is SECCOMP_MODE_FILTER, but the process does not have the CAP_SYS_ADMIN capability or has not set the no_new_privs attribute (see the discussion of PR_SET_NO_NEW_PRIVS above).
op is PR_SET_MM, and arg3 is PR_SET_MM_EXE_FILE, the file is not executable.
op is PR_SET_MM, arg3 is PR_SET_MM_EXE_FILE, and the file descriptor passed in arg4 is not valid.
op is PR_SET_MM, arg3 is PR_SET_MM_EXE_FILE, and this the second attempt to change the /proc/pid/exe symbolic link, which is prohibited.
arg2 is an invalid address.
op is PR_SET_SECCOMP, arg2 is SECCOMP_MODE_FILTER, the system was built with CONFIG_SECCOMP_FILTER, and arg3 is an invalid address.
op is PR_SET_SYSCALL_USER_DISPATCH and arg5 has an invalid address.
The value of op is not recognized, or not supported on this system.
op is PR_MCE_KILL or PR_MCE_KILL_GET or PR_SET_MM, and unused prctl() arguments were not specified as zero.
arg2 is not valid value for this op.
op is PR_SET_SECCOMP or PR_GET_SECCOMP, and the kernel was not configured with CONFIG_SECCOMP.
op is PR_SET_SECCOMP, arg2 is SECCOMP_MODE_FILTER, and the kernel was not configured with CONFIG_SECCOMP_FILTER.
op is PR_SET_MM, and one of the following is true
•
arg4 or arg5 is nonzero;
•
arg3 is greater than TASK_SIZE (the limit on the size of the user address space for this architecture);
•
arg2 is PR_SET_MM_START_CODE, PR_SET_MM_END_CODE, PR_SET_MM_START_DATA, PR_SET_MM_END_DATA, or PR_SET_MM_START_STACK, and the permissions of the corresponding memory area are not as required;
•
arg2 is PR_SET_MM_START_BRK or PR_SET_MM_BRK, and arg3 is less than or equal to the end of the data segment or specifies a value that would cause the RLIMIT_DATA resource limit to be exceeded.
op is PR_SET_PTRACER and arg2 is not 0, PR_SET_PTRACER_ANY, or the PID of an existing process.
op is PR_SET_PDEATHSIG and arg2 is not a valid signal number.
op is PR_SET_DUMPABLE and arg2 is neither SUID_DUMP_DISABLE nor SUID_DUMP_USER.
op is PR_SET_TIMING and arg2 is not PR_TIMING_STATISTICAL.
op is PR_SET_NO_NEW_PRIVS and arg2 is not equal to 1 or arg3, arg4, or arg5 is nonzero.
op is PR_GET_NO_NEW_PRIVS and arg2, arg3, arg4, or arg5 is nonzero.
op is PR_SET_THP_DISABLE and arg3, arg4, or arg5 is nonzero.
op is PR_GET_THP_DISABLE and arg2, arg3, arg4, or arg5 is nonzero.
op is PR_CAP_AMBIENT and an unused argument (arg4, arg5, or, in the case of PR_CAP_AMBIENT_CLEAR_ALL, arg3) is nonzero; or arg2 has an invalid value; or arg2 is PR_CAP_AMBIENT_LOWER, PR_CAP_AMBIENT_RAISE, or PR_CAP_AMBIENT_IS_SET and arg3 does not specify a valid capability.
op was PR_GET_SPECULATION_CTRL or PR_SET_SPECULATION_CTRL and unused arguments to prctl() are not 0.
op is PR_PAC_RESET_KEYS and the arguments are invalid or unsupported. See the description of PR_PAC_RESET_KEYS above for details.
op is PR_SVE_SET_VL and the arguments are invalid or unsupported, or SVE is not available on this platform. See the description of PR_SVE_SET_VL above for details.
op is PR_SVE_GET_VL and SVE is not available on this platform.
op is PR_SET_SYSCALL_USER_DISPATCH and one of the following is true:
•
arg2 is PR_SYS_DISPATCH_OFF and the remaining arguments are not 0;
•
arg2 is PR_SYS_DISPATCH_ON and the memory range specified is outside the address space of the process.
•
arg2 is invalid.
op is PR_SET_TAGGED_ADDR_CTRL and the arguments are invalid or unsupported. See the description of PR_SET_TAGGED_ADDR_CTRL above for details.
op is PR_GET_TAGGED_ADDR_CTRL and the arguments are invalid or unsupported. See the description of PR_GET_TAGGED_ADDR_CTRL above for details.
op was PR_SET_SPECULATION_CTRL the kernel or CPU does not support the requested speculation misfeature.
op was PR_MPX_ENABLE_MANAGEMENT or PR_MPX_DISABLE_MANAGEMENT and the kernel or the CPU does not support MPX management. Check that the kernel and processor have MPX support.
op was PR_SET_SPECULATION_CTRL implies that the control of the selected speculation misfeature is not possible. See PR_GET_SPECULATION_CTRL for the bit fields to determine which option is available.
op is PR_SET_FP_MODE and arg2 has an invalid or unsupported value.
op is PR_SET_SECUREBITS, and the caller does not have the CAP_SETPCAP capability, or tried to unset a "locked" flag, or tried to set a flag whose corresponding locked flag was set (see capabilities(7)).
op is PR_SET_SPECULATION_CTRL wherein the speculation was disabled with PR_SPEC_FORCE_DISABLE and caller tried to enable it again.
op is PR_SET_KEEPCAPS, and the caller's SECBIT_KEEP_CAPS_LOCKED flag is set (see capabilities(7)).
op is PR_CAPBSET_DROP, and the caller does not have the CAP_SETPCAP capability.
op is PR_SET_MM, and the caller does not have the CAP_SYS_RESOURCE capability.
op is PR_CAP_AMBIENT and arg2 is PR_CAP_AMBIENT_RAISE, but either the capability specified in arg3 is not present in the process's permitted and inheritable capability sets, or the PR_CAP_AMBIENT_LOWER securebit has been set.
op was PR_SET_SPECULATION_CTRL and arg3 is not PR_SPEC_ENABLE, PR_SPEC_DISABLE, PR_SPEC_FORCE_DISABLE, nor PR_SPEC_DISABLE_NOEXEC.

VERSIONS

IRIX has a prctl() system call (also introduced in Linux 2.1.44 as irix_prctl on the MIPS architecture), with prototype


ptrdiff_t prctl(int op, int arg2, int arg3);

and operations to get the maximum number of processes per user, get the maximum number of processors the calling process can use, find out whether a specified process is currently blocked, get or set the maximum stack size, and so on.

STANDARDS

Linux.

HISTORY

Linux 2.1.57, glibc 2.0.6

SEE ALSO

signal(2), core(5)

2024-05-02 Linux man-pages 6.8