GETRLIMIT(2) Linux Programmer's Manual GETRLIMIT(2)
NAME
getrlimit, getrusage, setrlimit - get/set(7,n,1 builtins) resource limits and usage
SYNOPSIS
#include <sys/time.h>
#include <sys/resource.h>
#include <unistd.h>
int getrlimit(int resource, struct rlimit *rlim);
int getrusage(int who, struct rusage *usage);
int setrlimit(int resource, const struct rlimit *rlim);
DESCRIPTION
getrlimit and setrlimit get and set(7,n,1 builtins) resource limits respectively. Each
resource has an associated soft and hard limit, as defined by the
rlimit structure (the rlim argument to both getrlimit() and setr-
limit()):
struct rlimit {
rlim_t rlim_cur; /* Soft limit */
rlim_t rlim_max; /* Hard limit (ceiling
for rlim_cur) */
};
The soft limit is the value that the kernel enforces for the corre-
sponding resource. The hard limit acts as a ceiling for the soft
limit: an unprivileged process may only set(7,n,1 builtins) its soft limit to a value
in(1,8) the range from 0 up to the hard limit, and (irreversibly) lower its
hard limit. A privileged process (under Linux: one with the
CAP_SYS_RESOURCE capability) may make arbitrary changes to either limit
value.
The value RLIM_INFINITY denotes no limit on a resource (both in(1,8) the
structure returned by getrlimit() and in(1,8) the structure passed to setr-
limit()).
resource must be one of:
RLIMIT_AS
The maximum size of the process's virtual(5,8) memory (address space)
in(1,8) bytes. This limit affects calls to brk(2), mmap(2) and
mremap(2), which fail with the error(8,n) ENOMEM upon exceeding this
limit. Also automatic stack expansion will fail (and generate a
SIGSEGV that kills the process when no alternate stack has been
made available). Since the value is a long, on machines with a
32-bit long either this limit is at most 2 GiB, or this resource
is unlimited.
RLIMIT_CORE
Maximum size of core file. When 0 no core dump files are cre-
ated. When nonzero, larger dumps are truncated to this size.
RLIMIT_CPU
CPU time(1,2,n) limit in(1,8) seconds. When the process reaches the soft
limit, it is sent a SIGXCPU signal. The default action for this
signal(2,7) is to terminate the process. However, the signal(2,7) can be
caught, and the handler can return control to the main program.
If the process continues to consume CPU time(1,2,n), it will be sent
SIGXCPU once per second until the hard limit is reached, at
which time(1,2,n) it is sent SIGKILL. (This latter point describes
Linux 2.2 and 2.4 behaviour. Implementations vary in(1,8) how they
treat processes which continue to consume CPU time(1,2,n) after reach-
ing the soft limit. Portable applications that need to catch
this signal(2,7) should perform an orderly termination upon first
receipt of SIGXCPU.)
RLIMIT_DATA
The maximum size of the process's data segment (initialized
data, uninitialized data, and heap). This limit affects calls
to brk() and sbrk(), which fail with the error(8,n) ENOMEM upon
encountering the soft limit of this resource.
RLIMIT_FSIZE
The maximum size of files that the process may create. Attempts
to extend a file(1,n) beyond this limit result in(1,8) delivery of a
SIGXFSZ signal. By default, this signal(2,7) terminates a process,
but a process can catch this signal(2,7) instead, in(1,8) which case the
relevant system call (e.g., write(1,2)(), truncate(2,7)()) fails with the
error(8,n) EFBIG.
RLIMIT_LOCKS
A limit on the combined number of flock(1,2)() locks and fcntl()
leases that this process may establish. (Early Linux 2.4 only.)
RLIMIT_MEMLOCK
The maximum number of bytes of virtual(5,8) memory that may be locked
into RAM using mlock() and mlockall().
RLIMIT_NOFILE
Specifies a value one greater than the maximum file(1,n) descriptor
number that can be opened by this process. Attempts (open(2,3,n)(),
pipe(2,8)(), dup(), etc.) to exceed this limit yield the error(8,n)
EMFILE.
RLIMIT_NPROC
The maximum number of processes that can be created for the real
user ID of the calling process. Upon encountering this limit,
fork() fails with the error(8,n) EAGAIN.
RLIMIT_RSS
Specifies the limit (in(1,8) pages) of the process's resident set(7,n,1 builtins)
(the number of virtual(5,8) pages resident in(1,8) RAM). This limit only
has effect in(1,8) Linux 2.4 onwatrds, and there only affects calls
to madvise() specifying MADVISE_WILLNEED.
RLIMIT_STACK
The maximum size of the process stack, in(1,8) bytes. Upon reaching
this limit, a SIGSEGV signal(2,7) is generated. To handle this sig-
nal(2,7), a process must employ an alternate signal(2,7) stack (sigalt-
stack(2)).
RLIMIT_OFILE is the BSD name for RLIMIT_NOFILE.
getrusage returns the current resource usages, for a who of either
RUSAGE_SELF or RUSAGE_CHILDREN. The former asks for resources used by
the current process, the latter for resources used by those of its
children that have terminated and have been waited for.
struct rusage {
struct timeval ru_utime; /* user time(1,2,n) used */
struct timeval ru_stime; /* system time(1,2,n) used */
long ru_maxrss; /* maximum resident set(7,n,1 builtins) size */
long ru_ixrss; /* integral shared memory size */
long ru_idrss; /* integral unshared data size */
long ru_isrss; /* integral unshared stack size */
long ru_minflt; /* page reclaims */
long ru_majflt; /* page faults */
long ru_nswap; /* swaps */
long ru_inblock; /* block input operations */
long ru_oublock; /* block output operations */
long ru_msgsnd; /* messages sent */
long ru_msgrcv; /* messages received */
long ru_nsignals; /* signals received */
long ru_nvcsw; /* voluntary context switches */
long ru_nivcsw; /* involuntary context switches */
};
RETURN VALUE
On success, zero is returned. On error(8,n), -1 is returned, and errno is
set(7,n,1 builtins) appropriately.
ERRORS
EFAULT rlim or usage points outside the accessible address space.
EINVAL getrlimit or setrlimit is called with a bad resource, or
getrusage is called with a bad who.
EPERM An unprivileged process tried to use setrlimit() to increase a
soft or hard limit above the current hard limit; the
CAP_SYS_RESOURCE capability is required to do this. Or, the
process tried to use setrlimit() to increase the soft or hard
RLIMIT_NOFILE limit above the current kernel maximum (NR_OPEN).
CONFORMING TO
SVr4, BSD 4.3
NOTE
Including <sys/time.h> is not required these days, but increases porta-
bility. (Indeed, struct timeval is defined in(1,8) <sys/time.h>.)
On Linux, if(3,n) the disposition of SIGCHLD is set(7,n,1 builtins) to SIG_IGN then the
resource usages of child processes are automatically included in(1,8) the
value returned by RUSAGE_CHILDREN, although POSIX 1003.1-2001 explic-
itly prohibits this.
The above struct was taken from BSD 4.3 Reno. Not all fields are mean-
ingful under Linux. Right now (Linux 2.4, 2.6) only the fields
ru_utime, ru_stime, ru_minflt, ru_majflt, and ru_nswap are maintained.
SEE ALSO
dup(2), fcntl(2), fork(2), mlock(2), mlockall(2), mmap(2), open(2,3,n)(2),
quotactl(2), sbrk(2), wait3(2), wait4(2), malloc(3), ulimit(3), capa-
bilities(7), signal(2,7)(7)
Linux 2004-06-16 GETRLIMIT(2)