.\" .\" Copyright (c) 2020-2024 Jim Warner .\" Copyright (c) 2020-2024 Craig Small .\" .\" This manual is free software; you can redistribute it and/or .\" modify it under the terms of the GNU Lesser General Public .\" License as published by the Free Software Foundation; either .\" version 2.1 of the License, or (at your option) any later version. .\" .\" .TH PROCPS_MISC 3 2024-07-06 procps-ng .\" Please adjust this date whenever revising the manpage. .\" .nh .SH NAME procps_misc \- API for miscellaneous information in the /proc filesystem .SH SYNOPSIS .nf .B #include .PP Platform Particulars .RS 4 .PP .RB "long " procps_cpu_count " (void); .RB "long " procps_hertz_get " (void); .RB "unsigned int " procps_pid_length " (void); .RB "int " procps_linux_version " (void); .RE .PP Runtime Particulars .PP .RS 4 .RI "int \fB procps_container_uptime\fR (double *" uptime_secs ");" .RI "int \fB procps_loadavg\fR (double *" av1 ", double *" av5 ", double *" av15 ");" .RI "int \fB procps_uptime\fR (double *" uptime_secs ", double *" idle_secs ");" .RI "int \fB procps_uptime_snprint\fR ( char *restrict " str ", size_t " size ", double " uptime_secs ", const int " pretty ");" .RB "char *" procps_uptime_sprint " (void);" .RB "char *" procps_uptime_sprint_short " (void);" .RB "int " procps_users " (void);" .RE .PP Namespace Particulars .PP .RS 4 .RI "int \fB procps_ns_get_id\fR (const char *" name ");" .RI "const char\fB *procps_ns_get_name\fR (int " id ");" .RI "int \fB procps_ns_read_pid\fR (int " pid ", struct procps_ns *" nsp ");" .RE .PP Mask Name Resolving .PP .RS 4 .RI "int \fBprocps_sigmask_names\fR (char *const " str ", size_t " size " , const char *" sigmask ");" .RI "int \fBprocps_capmask_names\fR (char *const " str ", size_t " size ", const char *" capmask ");" .RE .PP .P Link with \fI\-lproc2\fP. .SH DESCRIPTION .BR procps_cpu_count () returns the number of CPUs that are currently online as .BI sysconf( _SC_NPROCESSORS_ONLY ) or an assumed \fI1\fR. .P .BR procps_hertz_get () returns the number of clock ticks per second as .BI sysconf( _SC_CLK_TCK ) or an assumed \fI100\fR. Dividing tics by this value yields seconds. .P .BR procps_pid_length () returns the maximum string length for a PID on the system. For example, if the largest possible PID value on was 123, then the length would be 3. If the file \fI/proc/sys/kernel/pid_max\fR is unreadable, the value is assumed to be \fI5\fR. .P .BR procps_linux_version () returns the current Linux version as an encoded integer. On non-Linux systems that have an emulated proc filesystem this function returns the version of the Linux emulation instead. The version consists of three positive integers representing the major, minor and patch levels. The following macros are provided for encoding a given Linux version or separating out the components of the current version. .RS 4 .PP LINUX_VERSION(\ major\ ,\ minor\ ,\ patch\ ) .PP LINUX_VERSION_MAJOR(\ ver\ ) .PP LINUX_VERSION_MINOR(\ ver\ ) .PP LINUX_VERSION_PATCH(\ ver\ ) .RE .P .BR procps_loadavg () fetches the system load average and puts the 1, 5 and 15 minute averages into location(s) specified by any pointer which is not \fINULL\fR. .P .BR procps_container_uptime () returns the uptime of a container into location specified by the pointer if it is not \fINULL\fR. This is currently determined by the elapsed time of PID 1, which could be given with the command: .PP .in +4n .EX $ \fBps -o etimes 1\fR ELAPSED 12345 .EE .in .PP There is no concept of idle time in a container so \fBprocps_container_uptime\fR() doesn't have that parameter. .P .BR procps_uptime () returns uptime and/or idle seconds into location(s) specified by any pointer which is not \fINULL\fR. The \fBsprint\fR varieties return a human-readable string in one of two forms. .RS 4 .PP HH:MM:SS up HH:MM, # users, load average: 1, 5, 15 MM averages .PP up HH, MM .RE .P .BR procps_uptime_snprint () Uses the given buffer \fIstr\fR of \fIsize\fR length to fill in the uptime string instead of a statically allocated buffer. The function requires the \fIuptime_secs\fR value, which can be obtained from .BR procps_uptime () or .BR procps_container_uptime () and the \fIpretty\fR flag determines if the output is standard or pretty/short. .P .BR procps_users () returns the number of users on the system. This value comes from .BR sd_get_sessions (3) and counting the sessions of user, user-early and user-incomplete class or enumerating through .BR getutent (3). .P .BR procps_ns_get_id () returns the integer id (enum namespace_type) of the namespace for the given namespace \fIname\fR. .P .BR procps_ns_get_name () returns the name of the namespace for the given \fIid\fR (enum namespace_type). .P .BR procps_ns_read_pid () returns the inodes for the namespaces of the given process in the procps_ns structure pointed to by \fInsp\fR. Those inodes will appear in the order proscribed by enum namespace_type. .PP .RS 4 .nf enum namespace_type { PROCPS_NS_CGROUP, PROCPS_NS_IPC, PROCPS_NS_MNT, PROCPS_NS_NET, PROCPS_NS_PID, PROCPS_NS_TIME, PROCPS_NS_USER, PROCPS_NS_UTS }; .fi .RE .P .BR procps_sigmask_names () Fills \fIstr\fR with a human-readable text string of up to \fIsize\fR bytes of the signals set in the hexadecimal signal mask \fIsigmask\fR, see .BR signal (7). If the list of signals exceeds \fIsize\fR bytes, the string will be truncated and will end with a '+'. It is up to the calling program to ensure that \fIstr\fR is correctly allocated with at least \fIsize\fR bytes. .P .BR procps_capmask_names () Fills \fIstr\fR with a human-readable text string of up to \fIsize\fR bytes of the capabilites set in the hexadecimal capability mask \fIcapmask\fR, see .BR capabilities (7). If the list of capabilities exceeds \fIsize\fR bytes, the string will be truncated and will end with a '+'. It is up to the calling program to ensure that \fIstr\fR is correctly allocated with at least \fIsize\fR bytes. .P For a process that has no capabilities or all capabilities the string will be "-" and "full" respectively. .SH RETURN VALUE .SS Functions Returning an \[oq]int\[cq] or \[oq]long\[cq] An error will be indicated by a negative number that is always the inverse of some well known errno.h value. .SS Functions Returning an \[oq]address\[cq] An error will be indicated by a NULL return pointer with the reason found in the formal errno value. .SH FILES .TP .I /proc/loadavg The raw values for load average. .TP .I /proc/sys/kernel/osrelease Contains the release version of the Linux kernel or proc filesystem. .TP .I /proc/sys/kernel/pid_max Contains the value at which PIDs wrap around, one greater than the maximum PID value. .TP .I /proc/uptime The raw values for system uptime and idle time. .TP .IR /proc/ PID /ns contains the set of namespaces for a particular .IR PID . .SH SEE ALSO .BR procps (3), .BR procps_pids (3), .BR getutent (3), .BR sd_get_sessions (3), .BR proc (5), .BR capabilities (7), .BR signal (7).