README file for Solaris::Procfs 0.14 NAME Solaris::Procfs - access Solaris process information from Perl SYNOPSIS (See the EXAMPLES section below for more info.) DESCRIPTION This module is an interface the /proc filesystem on Solaris systems. Each process on a Solaris system has a corresponding directory under /proc, named after the process id. In each of these directories are a series of files and subdirectories, which contain information about each process. The proc(4) manpage gives complete details about these files. Basically, the files contain one or more C structs with data about its corresponding process, maintained by the kernel. This module provides methods which access these files and convert the data structures contained in them into nested hashrefs. This module has been tested on Solaris 2.6 and Solaris 7. It will not work on Solaris 2.5.1 systems (yet). STATUS This is pre-alpha software. It is far from finished. Parts of it need extensive rewriting and testing. However, the core functionality does seem to work properly. Contributions and critiques would be warmly welcomed. This module has been tested on the following systems, using gcc for builds: SunOS 5.7 (Solaris 7) SPARC SunOS 5.7 (Solaris 7) x86 SunOS 5.6 (Solaris 2.6) SPARC It may not even build on other systems. EXAMPLES There are three different ways to invoke the functions in this module: as object methods, as functions, or as a tied hash. As functions: use Solaris::Procfs; my $psinfo = Solaris::Procfs::psinfo( $pid ); As exported functions: use Solaris::Procfs (:procfiles); my $psinfo = psinfo( $pid ); As process objects: use Solaris::Procfs; use Solaris::Procfs::Process; my $p = new Solaris::Procfs::Process $pid; my $data = $p->psinfo; As process objects with tied hashes: use Solaris::Procfs; use Solaris::Procfs::Process; my $p = new Solaris::Procfs::Process $pid; my $data = $p->{psinfo}; As a filesystem object with tied hashes: use Solaris::Procfs; use Solaris::Procfs::Filesystem; my $p = new Solaris::Procfs::Filesystem; my $data = $p->{psinfo}->{$pid}; FUNCTIONS This module defines functions which each correspond to the files available under the directories in /proc. Complete descriptions of these files are available in the proc(4) manpage. Unless otherwise noted, the corresponding function in the Solaris::Procfs module simply returns the contents of the file in the form of a set of nested hashrefs. Exceptions to this are listed below. These functions can also be accessed implcitly as elements in a tied hash. When used this way, each process can be accessed as if it were one giant Perl structure, containing all the data related to that process id in the files under /proc/{pid}. To do this, you must use either the Solaris::Procfs::Process or the Solaris::Procfs::Filesystem modules. If you only use the Solaris::Procfs module, then you can only use the function-oriented interface. as Not yet implemented. The 'as' file contains the address-space image of the process. auxv The 'auxv' file contains the initial values of the process's aux vector in an array of auxv_t structures (see ). ctl Not implemented. The 'ctl' file is a write-only file to which structured messages are written directing the system to change some aspect of the process's state or control its behavior in some way. cwd Returns a string containing the absolute path to the process' current working directory. The 'cwd' file is a symbolic link to the process's current working directory. fd Returns a hash whose keys are the process' open file descriptors, and whose values are the absolute paths to the open files, as far as can be determined. The 'fd' directory contains references to the open files of the process. Each entry is a decimal number corresponding to an open file descriptor in the process. ldt Not yet implemented. The 'ldt' file exists only on x86 based machines. It is non-empty only if the process has established a local descriptor table (LDT). If non-empty, the file contains the array of currently active LDT entries in an array of elements of type struct ssd, defined in , one element for each active LDT entry. lpsinfo The 'lpsinfo' file contains a prheader structure followed by an array of lwpsinfo structures, one for each lwp in the process. lstatus The 'lstatus' file contains a prheader structure followed by an array of lwpstatus structures, one for each lwp in the process. lusage The 'lusage' file contains a prheader structure followed by an array of prusage structures, one for each lwp in the process plus an additional element at the beginning that contains the summation over all defunct lwps. lwp The 'lwp' directory contains entries each of which names an lwp within the process. These entries are themselves directories containing additional files. This function returns the contents of the files 'lwpstatus', 'lwpsinfo', and 'lwpusage', translated into a set of nested hashes. Interfaces to the files 'asrs', 'gwindoes', 'lwpctl' and 'xregs' have not been implemented. map The 'map' file contains information about the virtual address map of the process. The file contains an array of prmap structures, each of which describes a contiguous virtual address region in the address space of the traced process. object Not yet implemented. The 'object' directory containing read-only files with names corresponding to the entries in the map and pagedata files. Opening such a file yields a file descriptor for the underlying mapped file associated with an address-space mapping in the process. pagedata Not yet implemented. Opening the 'pagedata' file enables tracking of address space references and modifications on a per-page basis. prcred The 'prcred' file contains a description of the credentials associated with the process (UID, GID, etc.). psinfo The 'psinfo' file ontains miscellaneous information about the process and the representative lwp needed by the ps(1) command. status The 'status' file ontains state information about the process and the representative lwp. rmap The 'rmap' file contains information about the reserved address ranges of the process. Examples of such reservations include the address ranges reserved for the process stack and the individual thread stacks of a multi-threaded process. root Returns a string containing the absolute path to the process' root directory. The 'root' file is a symbolic link to the process' current working directory. sigact The 'sigact' file contains an array of sigaction structures describing the current dispositions of all signals associated with the traced process (see sigaction(2)). usage The 'usage' file contains process usage information described by a prusage structure. watch Not yet implemented. The 'watch' file contains an array of prwatch structures, one for each watched area established by the PCWATCH control operation. CHANGES * Version 0.14 Separated the Filesystem and Process modules from the main Procfs module. So Procfs.pm now contains no object-oriented code. * Version 0.10 Initial release on CPAN TO DO * Improve the documentation, test scripts and sample scripts. * Finish writing the Perl interface functions in Procfs.pm. * Add functions which can read the 'as' file. * Implement Perl functions which correspond to each of the procutils binaries (under /usr/proc/bin). These are described in the proc(1) manpage. * Add support for Solaris 8, and make sure that this package will build properly on a variety of recent Solaris flavors. * Add support for Solaris 2.5.1. This will require a good bit of work, as the /proc filesystem is rather different under Solaris 2.5.1. This item is low on the priority list. THANKS Much of this code is modeled after code written by Alan Burlison, and I received some helpful and timely advice from Tye McQueen. Thanks to Daniel J. Urist for writing Proc::ProcessTable. I used his method for keeping track of TTY numbers. AUTHOR John Nolan jpnolan@sonic.net 1999, 2000. A copyright statment is contained in the source code itself.