tcp_wrappers_7.6/ 744 465 506 0 6322311617 7154 tcp_wrappers_7.6/README 444 465 506 136127 6314551412 10173 @(#) README 1.30 97/03/21 19:27:21 This is the 7.6 version of the TCP/IP daemon wrapper package. Thank you for using this program. If you like it, send me a postcard. My postal address is at the bottom of this file. Read the BLURB file for a brief summary of what is new. The CHANGES file gives a complete account of differences with respect to previous releases. Announcements of new releases of this software are posted to Usenet (comp.security.unix, comp.unix.admin), to the cert-tools mailing list, and to a dedicated mailing list. You can subscribe to the dedicated mailing list by sending an email message to majordomo@wzv.win.tue.nl with in the body (not subject): subscribe tcp-wrappers-announce. Table of contents ----------------- 1 - Introduction 2 - Disclaimer 3 - Tutorials 3.1 - How it works 3.2 - Where the logging information goes 4 - Features 4.1 - Access control 4.2 - Host name spoofing 4.3 - Host address spoofing 4.4 - Client username lookups 4.5 - Language extensions 4.6 - Multiple ftp/gopher/www archives on one host 4.7 - Banner messages 4.8 - Sequence number guessing 5 - Other works 5.1 - Related documents 5.2 - Related software 6 - Limitations 6.1 - Known wrapper limitations 6.2 - Known system software bugs 7 - Configuration and installation 7.1 - Easy configuration and installation 7.2 - Advanced configuration and installation 7.3 - Daemons with arbitrary path names 7.4 - Building and testing the access control rules 7.5 - Other applications 8 - Acknowledgements 1 - Introduction ---------------- With this package you can monitor and filter incoming requests for the SYSTAT, FINGER, FTP, TELNET, RLOGIN, RSH, EXEC, TFTP, TALK, and other network services. It supports both 4.3BSD-style sockets and System V.4-style TLI. Praise yourself lucky if you don't know what that means. The package provides tiny daemon wrapper programs that can be installed without any changes to existing software or to existing configuration files. The wrappers report the name of the client host and of the requested service; the wrappers do not exchange information with the client or server applications, and impose no overhead on the actual conversation between the client and server applications. Optional features are: access control to restrict what systems can connect to what network daemons; client user name lookups with the RFC 931 etc. protocol; additional protection against hosts that pretend to have someone elses host name; additional protection against hosts that pretend to have someone elses host address. The programs are very portable. Build procedures are provided for many common (and not so common) environments, and guidelines are provided in case your environment is not among them. Requirements are that network daemons are spawned by a super server such as the inetd; a 4.3BSD-style socket programming interface and/or System V.4-style TLI programming interface; and the availability of a syslog(3) library and of a syslogd(8) daemon. The wrappers should run without modification on any system that satisfies these requirements. Workarounds have been implemented for several common bugs in systems software. What to do if this is your first encounter with the wrapper programs: 1) read the tutorial sections for an introduction to the relevant concepts and terminology; 2) glance over the security feature sections in this document; 3) follow the installation instructions (easy or advanced). I recommend that you first use the default security feature settings. Run the wrappers for a few days to become familiar with their logs, before doing anything drastic such as cutting off access or installing booby traps. 2 - Disclaimer -------------- The wrapper programs rely on source address information obtained from network packets. This information is provided by the client host. It is not 100 percent reliable, although the wrappers do their best to expose forgeries. In the absence of cryptographic protection of message contents, and of cryptographic authentication of message originators, all data from the network should be treated with sound scepticism. THIS RESTRICTION IS BY NO MEANS SPECIFIC TO THE TCP/IP PROTOCOLS. 3 - Tutorials ------------- The tutorial sections give a gentle introduction to the operation of the wrapper programs, and introduce some of the terminology that is used in the remainder of the document: client, server, the inetd and syslogd daemons, and their configuration files. 3.1 - How it works ------------------ Almost every application of the TCP/IP protocols is based on a client- server model. For example, when a user invokes the telnet command to connect to one of your systems, a telnet server process is executed on the target host. The telnet server process connects the user to a login process. A few examples of client and server programs are shown in the table below: client server application -------------------------------- telnet telnetd remote login ftp ftpd file transfer finger fingerd show users The usual approach is to run one single daemon process that waits for all kinds of incoming network connections. Whenever a connection is established, this daemon (usually called inetd) runs the appropriate server program and goes back to sleep, waiting for other connections. The wrapper programs rely on a simple, but powerful mechanism. Instead of directly running the desired server program, the inetd is tricked into running a small wrapper program. The wrapper logs the client host name or address and performs some additional checks. When all is well, the wrapper executes the desired server program and goes away. The wrapper programs have no interaction with the client user (or with the client process). Nor do the wrappers interact with the server application. This has two major advantages: 1) the wrappers are application-independent, so that the same program can protect many kinds of network services; 2) no interaction also means that the wrappers are invisible from outside (at least for authorized users). Another important property is that the wrapper programs are active only when the initial contact between client and server is established. Once a wrapper has done its work there is no overhead on the client-server conversation. The simple mechanism has one major drawback: the wrappers go away after the initial contact between client and server processes, so the wrappers are of little use with network daemons that service more than one client. The wrappers would only see the first client attempt to contact such a server. The NFS mount daemon is a typical example of a daemon that services requests from multiple clients. See the section on related software for ways to deal with such server programs. There are two ways to use the wrapper programs: 1) The easy way: move network daemons to some other directory and fill the resulting holes with copies of the wrapper programs. This approach involves no changes to system configuration files, so there is very little risk of breaking things. 2) The advanced way: leave the network daemons alone and modify the inetd configuration file. For example, an entry such as: tftp dgram udp wait root /usr/etc/tcpd in.tftpd -s /tftpboot When a tftp request arrives, inetd will run the wrapper program (tcpd) with a process name `in.tftpd'. This is the name that the wrapper will use when logging the request and when scanning the optional access control tables. `in.tftpd' is also the name of the server program that the wrapper will attempt to run when all is well. Any arguments (`-s /tftpboot' in this particular example) are transparently passed on to the server program. For an account of the history of the wrapper programs, with real-life examples, see the section below on related documents. 3.2 - Where the logging information goes ---------------------------------------- The wrapper programs send their logging information to the syslog daemon (syslogd). The disposition of the wrapper logs is determined by the syslog configuration file (usually /etc/syslog.conf). Messages are written to files, to the console, or are forwarded to a @loghost. Some syslogd versions can even forward messages down a |pipeline. Older syslog implementations (still found on Ultrix systems) only support priority levels ranging from 9 (debug-level messages) to 0 (alerts). All logging information of the specified priority level or more urgent is written to the same destination. In the syslog.conf file, priority levels are specified in numerical form. For example, 8/usr/spool/mqueue/syslog causes all messages with priority 8 (informational messages), and anything that is more urgent, to be appended to the file /usr/spool/mqueue/syslog. Newer syslog implementations support message classes in addition to priority levels. Examples of message classes are: mail, daemon, auth and news. In the syslog.conf file, priority levels are specified with symbolic names: debug, info, notice, ..., emerg. For example, mail.debug /var/log/syslog causes all messages of class mail with priority debug (or more urgent) to be appended to the /var/log/syslog file. By default, the wrapper logs go to the same place as the transaction logs of the sendmail daemon. The disposition can be changed by editing the Makefile and/or the syslog.conf file. Send a `kill -HUP' to the syslogd after changing its configuration file. Remember that syslogd, just like sendmail, insists on one or more TABs between the left-hand side and the right-hand side expressions in its configuration file. Solaris 2.x note: the syslog daemon depends on the m4 macro processor. The m4 program is installed as part of the software developer packages. Trouble shooting note: when the syslogging does not work as expected, run the program by hand (`syslogd -d') and see what really happens. 4 - Features ------------ 4.1 - Access control -------------------- When compiled with -DHOSTS_ACCESS, the wrapper programs support a simple form of access control. Access can be controlled per host, per service, or combinations thereof. The software provides hooks for the execution of shell commands when an access control rule fires; this feature may be used to install "booby traps". For details, see the hosts_access.5 manual page, which is in `nroff -man' format. A later section describes how you can test your access control rules. Access control can also be used to connect clients to the "right" service. What is right may depend on the requested service, the origin of the request, and what host address the client connects to. Examples: (1) A gopher or www database speaks native language when contacted from within the country, otherwise it speaks English. (2) A service provider offers different ftp, gopher or www services with different internet hostnames from one host (section 4.6). Access control is enabled by default. It can be turned off by editing the Makefile, or by providing no access control tables. The install instructions below describe the Makefile editing process. The hosts_options.5 manual page (`nroff -man' format) documents an extended version of the access control language. The extensions are disabled by default. See the section below on language extensions. Later System V implementations provide the Transport Level Interface (TLI), a network programming interface that performs functions similar to the Berkeley socket programming interface. Like Berkeley sockets, TLI was designed to cover multiple protocols, not just Internet. When the wrapper discovers that the TLI interface sits on top of a TCP/IP or UDP/IP conversation it uses this knowledge to provide the same functions as with traditional socket-based applications. When some other protocol is used underneath TLI, the host address will be some universal magic cookie that may not even be usable for access control purposes. 4.2 - Host name spoofing ------------------------ With some network applications, such as RSH or RLOGIN, the client host name plays an important role in the authentication process. Host name information can be reliable when lookups are done from a _local_ hosts table, provided that the client IP address can be trusted. With _distributed_ name services, authentication schemes that rely on host names become more problematic. The security of your system now may depend on some far-away DNS (domain name server) outside your own control. The wrapper programs verify the client host name that is returned by the address->name DNS server, by asking for a second opinion. To this end, the programs look at the name and addresses that are returned by the name->address DNS server, which may be an entirely different host. If any name or address discrepancies are found, or if the second DNS opinion is not available, the wrappers assume that one of the two name servers is lying, and assume that the client host pretends to have someone elses host name. When compiled with -DPARANOID, the wrappers will always attempt to look up and double check the client host name, and will always refuse service in case of a host name/address discrepancy. This is a reasonable policy for most systems. When compiled without -DPARANOID, the wrappers by default still perform hostname lookup. You can match hosts with a name/address discrepancy with the PARANOID wildcard and decide whether or not to grant service. Automatic hostname verification is enabled by default. Automatic hostname lookups and verification can be turned off by editing the Makefile. The configuration and installation section below describes the Makefile editing process. 4.3 - Host address spoofing --------------------------- While host name spoofing can be found out by asking a second opinion, it is much harder to find out that a host claims to have someone elses network address. And since host names are deduced from network addresses, address spoofing is at least as effective as name spoofing. The wrapper programs can give additional protection against hosts that claim to have an address that lies outside their own network. For example, some far-away host that claims to be a trusted host within your own network. Such things are possible even while the impersonated system is up and running. This additional protection is not an invention of my own; it has been present for at least five years in the BSD rsh and rlogin daemons. Unfortunately, that feature was added *after* 4.3 BSD came out, so that very few, if any, UNIX vendors have adopted it. Our site, and many other ones, has been running these enhanced daemons for several years, and without any ill effects. When the wrapper programs are compiled with -DKILL_IP_OPTIONS, the programs refuse to service TCP connections with IP source routing options. -DKILL_IP_OPTIONS is not needed on modern UNIX systems that can stop source-routed traffic in the kernel. Examples are 4.4BSD derivatives, Solaris 2.x, and Linux. See your system manuals for details. If you are going to use this feature on SunOS 4.1.x you should apply patch 100804-03+ or 101790-something depending on your SunOS version. Otherwise you may experience "BAD TRAP" and "Data fault" panics when the getsockopt() system call is executed after a TCP RESET has been received. This is a kernel bug, it is not the fault of the wrappers. The feature is disabled by default. It can be turned on by editing the Makefile. The configuration and installation section below describes the Makefile editing process. UDP services do not benefit from this additional protection. With UDP, all you can be certain of is the network packet's destination address. 4.4 - Client username lookups ----------------------------- The protocol proposed in RFC 931 provides a means to obtain the client user name from the client host. The requirement is that the client host runs an RFC 931-compliant daemon. The information provided by such a daemon is not intended to be used for authentication purposes, but it can provide additional information about the owner of a TCP connection. The RFC 931 protocol has diverged into different directions (IDENT, TAP, RFC 1413). To add to the confusion, they all use the same network port. The daemon wrappers implement a common subset of the protocols. There are some limitations: the number of hosts that run an RFC 931 (or compatible) daemon is limited (but growing); client user name lookups do not work for datagram (UDP) services. More seriously, client user name lookups can cause noticeable delays with connections from non-UNIX PCs. Recent PC software seem to have fixed this (for example NCSA telnet). The wrappers use a 10-second timeout for RFC931 lookups, to accommodate slow networks and slow hosts. By default, the wrappers will do username lookup only when the access control rules require them to do so (via user@host client patterns, see the hosts_access.5 manual page) or when the username is needed for % expansions. You can configure the wrappers to always perform client username lookups, by editing the Makefile. The client username lookup timeout period (10 seconds default) can be changed by editing the Makefile. The installation sections below describe the Makefile editing process. On System V with TLI-based network services, client username lookups will be possible only when the underlying network protocol is TCP/IP. 4.5 - Language extensions ------------------------- The wrappers sport only a limited number of features. This is for a good reason: programs that run at high privilege levels must be easy to verify. And the smaller a program, the easier to verify. There is, however, a provision to add features. The options.c module provides a framework for language extensions. Quite a few extensions have already been implemented; they are documented in the hosts_options.5 document, which is in `nroff -man' format. Examples: changing the severity level at which a request for service is logged; "allow" and "deny" keywords; running a customized server instead of the standard one; many others. The language extensions are not enabled by default because they introduce an incompatible change to the access control language syntax. Instructions to enable the extensions are given in the Makefile. 4.6 - Multiple ftp/gopher/www archives on one host -------------------------------------------------- Imagine one host with multiple internet addresses. These addresses do not need to have the same internet hostname. Thus, it is possible to offer services with different internet hostnames from just one host. Service providers can use this to offer organizations a presence on the "net" with their own internet hostname, even when those organizations aren't connected to the Internet at all. To the end user it makes no difference, because applications use internet hostnames. There are several ways to assign multiple addresses to one machine. The nice way is to take an existing network interface and to assign additional internet addresses with the `ifconfig' command. Examples: Solaris 2: ifconfig le0:1
netmask up 4.4 BSD: ifconfig en0 alias
netmask On other systems one has to increase the number of network interfaces: either with hardware interfaces, or with pseudo interfaces like SLIP or PPP. The interfaces do not need to be attached to anything. They just need to be up and to be assigned a suitable internet address and mask. With the wrapper software, `daemon@host' access control patterns can be used to distinguish requests by the network address that they are aimed at. Judicious use of the `twist' option (see the hosts_options.5 file, `nroff -man' format) can guide the requests to the right server. These can be servers that live in separate chroot areas, or servers modified to take additional context from the command line, or a combination. Another way is to modify gopher or www listeners so that they bind to only one specific network address. Multiple gopher or www servers can then be run side by side, each taking requests sent to its respective network address. 4.7 - Banner messages --------------------- Some sites are required to present an informational message to users before they attempt to login. Banner messages can also be useful when denying service: instead of simply dropping the connection a polite explanation is given first. Finally, banners can be used to give your system a more personal touch. The wrapper software provides easy-to-use tools to generate pre-login banners for ftp, telnet, rlogin etc. from a single prototype banner textfile. Details on banners and on-the-fly % expansions are given in the hosts_options.5 manual page (`nroff -man' format). An example is given in the file Banners.Makefile. In order to support banner messages the wrappers have to be built with language extensions enabled. See the section on language extensions. 4.8 - Sequence number guessing ------------------------------ Recently, systems came under attack from intruders that exploited a well-known weakness in TCP/IP sequence number generators. This weakness allows intruders to impersonate trusted hosts. Break-ins have been reported via the rsh service. In fact, any network service can be exploited that trusts the client host name or address. A long-term solution is to stop using network services that trust the client host name or address, and to use data encryption instead. A short-term solution, as outlined in in CERT advisory CA-95:01, is to configure network routers so that they discard datagrams from "outside" with an "inside" source address. This approach is most fruitful when you do not trust any hosts outside your local network. The IDENT (RFC931 etc.) client username lookup protocol can help to detect host impersonation attacks. Before accepting a client request, the wrappers can query the client's IDENT server and find out that the client never sent that request. When the client host provides IDENT service, a negative IDENT lookup result (the client matches `UNKNOWN@host') is strong evidence of a host impersonation attack. A positive IDENT lookup result (the client matches `KNOWN@host') is less trustworthy. It is possible for an attacker to spoof both the client request and the IDENT lookup connection, although doing so should be much harder than spoofing just a client request. Another possibility is that the client's IDENT server is lying. Client username lookups are described in more detail in a previous section. Pointers to IDENT daemon software are described in the section on related software. 5 - Other works --------------- 5.1 - Related documents ----------------------- The war story behind the tcp wrapper tools is described in: W.Z. Venema, "TCP WRAPPER, network monitoring, access control and booby traps", UNIX Security Symposium III Proceedings (Baltimore), September 1992. ftp.win.tue.nl:/pub/security/tcp_wrapper.ps.Z (postscript) ftp.win.tue.nl:/pub/security/tcp_wrapper.txt.Z (flat text) The same cracker is also described in: W.R. Cheswick, "An Evening with Berferd, In Which a Cracker is Lured, Endured, and Studied", Proceedings of the Winter USENIX Conference (San Francisco), January 1992. research.att.com:/dist/internet_security/berferd.ps An updated version of the latter paper appeared in: W.R. Cheswick, S.M. Bellovin, "Firewalls and Internet Security", Addison-Wesley, 1994. Discussions on internet firewalls are archived on ftp.greatcircle.com. Subscribe to the mailing list by sending a message to majordomo@greatcircle.com With in the body (not subject): subscribe firewalls. 5.2 - Related software ---------------------- Network daemons etc. with enhanced logging capabilities can generate massive amounts of information: our 150+ workstations generate several hundred kbytes each day. egrep-based filters can help to suppress some of the noise. A more powerful tool is the Swatch monitoring system by Stephen E. Hansen and E. Todd Atkins. Swatch can process log files in real time and can associate arbitrary actions with patterns; its applications are by no means restricted to security. Swatch is available ftp.stanford.edu, directory /general/security-tools/swatch. Socks, described in the UNIX Security III proceedings, can be used to control network traffic from hosts on an internal network, through a firewall host, to the outer world. Socks consists of a daemon that is run on the firewall host, and of a library with routines that redirect application socket calls through the firewall daemon. Socks is available from s1.gov in /pub/firewalls/socks.tar.Z. For a modified Socks version by Ying-Da Lee (ylee@syl.dl.nec.com) try ftp.nec.com, directory /pub/security/socks.cstc. Tcpr is a set of perl scripts by Paul Ziemba that enable you to run ftp and telnet commands across a firewall. Unlike socks it can be used with unmodified client software. Available from ftp.alantec.com, /pub/tcpr. The TIS firewall toolkit provides a multitude of tools to build your own internet firewall system. ftp.tis.com, directory /pub/firewalls. Versions of rshd and rlogind, modified to report the client user name in addition to the client host name, are available for anonymous ftp (ftp.win.tue.nl:/pub/security/logdaemon-XX.tar.Z). These programs are drop-in replacements for SunOS 4.x, Ultrix 4.x, SunOS 5.x and HP-UX 9.x. This archive also contains ftpd/rexecd/login versions that support S/Key or SecureNet one-time passwords in addition to traditional UNIX reusable passwords. The securelib shared library by William LeFebvre can be used to control access to network daemons that are not run under control of the inetd or that serve more than one client, such as the NFS mount daemon that runs until the machine goes down. Available from eecs.nwu.edu, file /pub/securelib.tar. xinetd (posted to comp.sources.unix) is an inetd replacement that provides, among others, logging, username lookup and access control. However, it does not support the System V TLI services, and involves much more source code than the daemon wrapper programs. Available from ftp.uu.net, directory /usenet/comp.sources.unix. netlog from Texas A&M relies on the SunOS 4.x /dev/nit interface to passively watch all TCP and UDP network traffic on a network. The current version is on net.tamu.edu in /pub/security/TAMU. Where shared libraries or router-based packet filtering are not an option, an alternative portmap daemon can help to prevent hackers from mounting your NFS file systems using the proxy RPC facility. ftp.win.tue.nl:/pub/security/portmap-X.shar.Z was tested with SunOS 4.1.X Ultrix 3.0 and Ultrix 4.x, HP-UX 8.x and some version of AIX. The protection is less effective than that of the securelib library because portmap is mostly a dictionary service. An rpcbind replacement (the Solaris 2.x moral equivalent of portmap) can be found on ftp.win.tue.nl in /pub/security. It prevents hackers from mounting your NFS file systems by using the proxy RPC facility. Source for a portable RFC 931 (TAP, IDENT, RFC 1413) daemon by Peter Eriksson is available from ftp.lysator.liu.se:/pub/ident/servers. Some TCP/IP implementations come without syslog library. Some come with the library but have no syslog daemon. A replacement can be found in ftp.win.tue.nl:/pub/security/surrogate-syslog.tar.Z. The fakesyslog library that comes with the nntp sources reportedly works well, too. 6 - Limitations --------------- 6.1 - Known wrapper limitations ------------------------------- Many UDP (and rpc/udp) daemons linger around for a while after they have serviced a request, just in case another request comes in. In the inetd configuration file these daemons are registered with the `wait' option. Only the request that started such a daemon will be seen by the wrappers. Such daemons are better protected with the securelib shared library (see: Related software). The wrappers do not work with RPC services over TCP. These services are registered as rpc/tcp in the inetd configuration file. The only non- trivial service that is affected by this limitation is rexd, which is used by the on(1) command. This is no great loss. On most systems, rexd is less secure than a wildcard in /etc/hosts.equiv. Some RPC requests (for example: rwall, rup, rusers) appear to come from the server host. What happens is that the client broadcasts its request to all portmap daemons on its network; each portmap daemon forwards the request to a daemon on its own system. As far as the rwall etc. daemons know, the request comes from the local host. Portmap and RPC (e.g. NIS and NFS) (in)security is a topic in itself. See the section in this document on related software. 6.2 - Known system software bugs -------------------------------- Workarounds have been implemented for several bugs in system software. They are described in the Makefile. Unfortunately, some system software bugs cannot be worked around. The result is loss of functionality. IRIX has so many bugs that it has its own README.IRIX file. Older ConvexOS versions come with a broken recvfrom(2) implementation. This makes it impossible for the daemon wrappers to look up the client host address (and hence, the name) in case of UDP requests. A patch is available for ConvexOS 10.1; later releases should be OK. With early Solaris (SunOS 5) versions, the syslog daemon will leave behind zombie processes when writing to logged-in users. Workaround: increase the syslogd threshold for logging to users, or reduce the wrapper's logging severity. On some systems, the optional RFC 931 etc. client username lookups may trigger a kernel bug. When a client host connects to your system, and the RFC 931 connection from your system to that client is rejected by a router, your kernel may drop all connections with that client. This is not a bug in the wrapper programs: complain to your vendor, and don't enable client user name lookups until the bug has been fixed. Reportedly, SunOS 4.1.1, Next 2.0a, ISC 3.0 with TCP 1.3, and AIX 3.2.2 and later are OK. Sony News/OS 4.51, HP-UX 8-something and Ultrix 4.3 still have the bug. Reportedly, a fix for Ultrix is available (CXO-8919). The following procedure can be used (from outside the tue.nl domain) to find out if your kernel has the bug. From the system under test, do: % ftp 131.155.70.19 This command attempts to make an ftp connection to our anonymous ftp server (ftp.win.tue.nl). When the connection has been established, run the following command from the same system under test, while keeping the ftp connection open: % telnet 131.155.70.19 111 Do not forget the `111' at the end of the command. This telnet command attempts to connect to our portmap process. The telnet command should fail with: "host not reachable", or with a timeout error. If your ftp connection gets messed up, you have the bug. If the telnet command does not fail, please let me know a.s.a.p.! For those who care, the bug is that the BSD kernel code was not careful enough with incoming ICMP UNREACHABLE control messages (it ignored the local and remote port numbers, and therefore zapped *all* connections with the remote system). The bug is still present in the BSD NET/1 source release (1989) but apparently has been fixed in BSD NET/2 (1991). 7 - Configuration and installation ---------------------------------- 7.1 - Easy configuration and installation ----------------------------------------- The "easy" recipe requires no changes to existing software or configuration files. Basically, you move the daemons that you want to protect to a different directory and plug the resulting holes with copies of the wrapper programs. If you don't run Ultrix, you won't need the miscd wrapper program. The miscd daemon implements among others the SYSTAT service, which produces the same output as the WHO command. Type `make' and follow the instructions. The Makefile comes with ready-to-use templates for many common UNIX implementations (sun, ultrix, hp-ux, aix, irix,...). IRIX has so many bugs that it has its own README.IRIX file. When the `make' succeeds the result is five executables (six in case of Ultrix). You can use the `tcpdchk' program to identify the most common problems in your wrapper and inetd configuration files. With the `tcpdmatch' program you can examine how the wrapper would react to specific requests for service. The `safe_finger' command should be used when you implement booby traps: it gives better protection against nasty stuff that remote hosts may do in response to your finger probes. The `try-from' program tests the host and username lookup code. Run it from a remote shell command (`rsh host /some/where/try-from') and it should be able to figure out from what system it is being called. The tcpd program can be used to monitor the telnet, finger, ftp, exec, rsh, rlogin, tftp, talk, comsat and other tcp or udp services that have a one-to-one mapping onto executable files. The tcpd program can also be used for services that are marked as rpc/udp in the inetd configuration file, but not for rpc/tcp services such as rexd. You probably do not want to run rexd anyway. On most systems it is even less secure than a wildcard in /etc/hosts.equiv. With System V.4-style systems, the tcpd program can also handle TLI services. When TCP/IP or UDP/IP is used underneath TLI, tcpd provides the same functions as with socket-based applications. When some other protocol is used underneath TLI, functionality will be limited (no client username lookups, weird network address formats). Decide which services you want to monitor. Move the corresponding vendor-provided daemon programs to the location specified by the REAL_DAEMON_DIR constant in the Makefile, and fill the holes with copies of the tcpd program. That is, one copy of (or link to) the tcpd program for each service that you want to monitor. For example, to monitor the use of your finger service: # mkdir REAL_DAEMON_DIR # mv /usr/etc/in.fingerd REAL_DAEMON_DIR # cp tcpd /usr/etc/in.fingerd The example applies to SunOS 4. With other UNIX implementations the network daemons live in /usr/libexec, /usr/sbin or in /etc, or have no "in." prefix to their names, but you get the idea. File protections: the wrapper, all files used by the wrapper, and all directories in the path leading to those files, should be accessible but not writable for unprivileged users (mode 755 or mode 555). Do not install the wrapper set-uid. Ultrix only: If you want to monitor the SYSTAT service, move the vendor-provided miscd daemon to the location specified by the REAL_DAEMON_DIR macro in the Makefile, and install the miscd wrapper at the original miscd location. In the absence of any access-control tables, the daemon wrappers will just maintain a record of network connections made to your system. 7.2 - Advanced configuration and installation --------------------------------------------- The advanced recipe leaves your daemon executables alone, but involves simple modifications to the inetd configuration file. Type `make' and follow the instructions. The Makefile comes with ready-to-use templates for many common UNIX implementations (sun, ultrix, hp-ux, aix, irix, ...). IRIX users should read the warnings in the README.IRIX file first. When the `make' succeeds the result is five executables (six in case of Ultrix). You can use the `tcpdchk' program to identify the most common problems in your wrapper and inetd configuration files. With the `tcpdmatch' program you can examine how the wrapper would react to specific requests for service. The `try-from' program tests the host and username lookup code. Run it from a remote shell command (`rsh host /some/where/try-from') and it should be able to figure out from what system it is being called. The `safe_finger' command should be used when you implement a booby trap: it gives better protection against nasty stuff that remote hosts may do in response to your finger probes. The tcpd program can be used to monitor the telnet, finger, ftp, exec, rsh, rlogin, tftp, talk, comsat and other tcp or udp services that have a one-to-one mapping onto executable files. With System V.4-style systems, the tcpd program can also handle TLI services. When TCP/IP or UDP/IP is used underneath TLI, tcpd provides the same functions as with socket-based applications. When some other protocol is used underneath TLI, functionality will be limited (no client username lookups, weird network address formats). The tcpd program can also be used for services that are marked as rpc/udp in the inetd configuration file, but not for rpc/tcp services such as rexd. You probably do not want to run rexd anyway. On most systems it is even less secure than a wildcard in /etc/hosts.equiv. Install the tcpd command in a suitable place. Apollo UNIX users will want to install it under a different name because the name "tcpd" is already taken; a suitable name would be "frontd". File protections: the wrapper, all files used by the wrapper, and all directories in the path leading to those files, should be accessible but not writable for unprivileged users (mode 755 or mode 555). Do not install the wrapper set-uid. Then perform the following edits on the inetd configuration file (usually /etc/inetd.conf or /etc/inet/inetd.conf): finger stream tcp nowait nobody /usr/etc/in.fingerd in.fingerd ^^^^^^^^^^^^^^^^^^^ becomes: finger stream tcp nowait nobody /usr/etc/tcpd in.fingerd ^^^^^^^^^^^^^ Send a `kill -HUP' to the inetd process to make the change effective. Some IRIX inetd implementations require that you first disable the finger service (comment out the finger service and `kill -HUP' the inetd) before you can turn on the modified version. Sending a HUP twice seems to work just as well for IRIX 5.3, 6.0, 6.0.1 and 6.1. AIX note: you may have to execute the `inetimp' command after changing the inetd configuration file. The example applies to SunOS 4. With other UNIX implementations the network daemons live in /usr/libexec, /usr/sbin, or /etc, the network daemons have no "in." prefix to their names, or the username field in the inetd configuration file may be missing. When the finger service works as expected you can perform similar changes for other network services. Do not forget the `kill -HUP'. The miscd daemon that comes with Ultrix implements several network services. It decides what to do by looking at its process name. One of the services is systat, which is a kind of limited finger service. If you want to monitor the systat service, install the miscd wrapper in a suitable place and update the inetd configuration file: systat stream tcp nowait /suitable/place/miscd systatd Ultrix 4.3 allows you to specify a user id under which the daemon will be executed. This feature is not documented in the manual pages. Thus, the example would become: systat stream tcp nowait nobody /suitable/place/miscd systatd Older Ultrix systems still run all their network daemons as root. In the absence of any access-control tables, the daemon wrappers will just maintain a record of network connections made to your system. 7.3 - Daemons with arbitrary path names --------------------------------------- The above tcpd examples work fine with network daemons that live in a common directory, but sometimes that is not practical. Having soft links all over your file system is not a clean solution, either. Instead you can specify, in the inetd configuration file, an absolute path name for the daemon process name. For example, ntalk dgram udp wait root /usr/etc/tcpd /usr/local/lib/ntalkd When the daemon process name is an absolute path name, tcpd ignores the value of the REAL_DAEMON_DIR constant, and uses the last path component of the daemon process name for logging and for access control. 7.4 - Building and testing the access control rules --------------------------------------------------- In order to support access control the wrappers must be compiled with the -DHOSTS_ACCESS option. The access control policy is given in the form of two tables (default: /etc/hosts.allow and /etc/hosts.deny). Access control is disabled when there are no access control tables, or when the tables are empty. If you haven't used the wrappers before I recommend that you first run them a couple of days without any access control restrictions. The logfile records should give you an idea of the process names and of the host names that you will have to build into your access control rules. The syntax of the access control rules is documented in the file hosts_access.5, which is in `nroff -man' format. This is a lengthy document, and no-one expects you to read it right away from beginning to end. Instead, after reading the introductory section, skip to the examples at the end so that you get a general idea of the language. Then you can appreciate the detailed reference sections near the beginning of the document. The examples in the hosts_access.5 document (`nroff -man' format) show two specific types of access control policy: 1) mostly closed (only permitting access from a limited number of systems) and 2) mostly open (permitting access from everyone except a limited number of trouble makers). You will have to choose what model suits your situation best. Implementing a mixed policy should not be overly difficult either. Optional extensions to the access control language are described in the hosts_options.5 document (`nroff -man' format). The `tcpdchk' program examines all rules in your access control files and reports any problems it can find. `tcpdchk -v' writes to standard output a pretty-printed list of all rules. `tcpdchk -d' examines the hosts.access and hosts.allow files in the current directory. This program is described in the tcpdchk.8 document (`nroff -man' format). The `tcpdmatch' command can be used to try out your local access control files. The command syntax is: tcpdmatch process_name hostname (e.g.: tcpdmatch in.tftpd localhost) tcpdmatch process_name address (e.g.: tcpdmatch in.tftpd 127.0.0.1) This way you can simulate what decisions will be made, and what actions will be taken, when hosts connect to your own system. The program is described in the tcpdmatch.8 document (`nroff -man' format). Note 1: `tcpdmatch -d' will look for hosts.{allow,deny} tables in the current working directory. This is useful for testing new rules without bothering your users. Note 2: you cannot use the `tcpdmatch' command to simulate what happens when the local system connects to other hosts. In order to find out what process name to use, just use the service and watch the process name that shows up in the logfile. Alternatively, you can look up the name from the inetd configuration file. Coming back to the tftp example in the tutorial section above: tftp dgram udp wait root /usr/etc/tcpd in.tftpd -s /tftpboot This entry causes the inetd to run the wrapper program (tcpd) with a process name `in.tftpd'. This is the name that the wrapper will use when scanning the access control tables. Therefore, `in.tftpd' is the process name that should be given to the `tcpdmatch' command. On your system the actual inetd.conf entry may differ (tftpd instead of in.tftpd, and no `root' field), but you get the idea. When you specify a host name, the `tcpdmatch' program will use both the host name and address. This way you can simulate the most common case where the wrappers know both the host address and the host name. The `tcpdmatch' program will iterate over all addresses that it can find for the given host name. When you specify a host address instead of a host name, the `tcpdmatch' program will pretend that the host name is unknown, so that you can simulate what happens when the wrapper is unable to look up the client host name. 7.5 - Other applications ------------------------ The access control routines can easily be integrated with other programs. The hosts_access.3 manual page (`nroff -man' format) describes the external interface of the libwrap.a library. The tcpd program can even be used to control access to the mail service. This can be useful when you suspect that someone is trying out some obscure sendmail bug, or when a remote site is misconfigured and keeps hammering your mail daemon. In that case, sendmail should not be run as a stand-alone network listener, but it should be registered in the inetd configuration file. For example: smtp stream tcp nowait root /usr/etc/tcpd /usr/lib/sendmail -bs You will still need to run one sendmail background process to handle queued-up outgoing mail. A command like: /usr/lib/sendmail -q15m (no `-bd' flag) should take care of that. You cannot really prevent people from posting forged mail this way, because there are many unprotected smtp daemons on the network. 8 - Acknowledgements -------------------- Many people contributed to the evolution of the programs, by asking inspiring questions, by suggesting features or bugfixes, or by submitting source code. Nevertheless, all mistakes and bugs in the wrappers are my own. Thanks to Brendan Kehoe (cs.widener.edu), Heimir Sverrisson (hafro.is) and Dan Bernstein (kramden.acf.nyu.edu) for feedback on an early release of this product. The host name/address check was suggested by John Kimball (src.honeywell.com). Apollo's UNIX environment has some peculiar quirks: Willem-Jan Withagen (eb.ele.tue.nl), Pieter Schoenmakers (es.ele.tue.nl) and Charles S. Fuller (wccs.psc.edu) provided assistance. Hal R. Brand (addvax.llnl.gov) told me how to get the client IP address in case of datagram-oriented services, and suggested the optional shell command feature. Shabbir Safdar (mentor.cc.purdue.edu) provided a first version of a much-needed manual page. Granville Boman Goza, IV (sei.cmu.edu) suggested to use the client IP address even when the host name is available. Casper H.S. Dik (fwi.uva.nl) provided additional insight into DNS spoofing techniques. The bogus daemon feature was inspired by code from Andrew Macpherson (BNR Europe Ltd). Steve Bellovin (research.att.com) confirmed some of my suspicions about the darker sides of TCP/IP insecurity. Risks of automated fingers were pointed out by Borja Marcos (we.lc.ehu.es). Brad Plecs (jhuspo.ca.jhu.edu) was kind enough to try my early TLI code and to work out how DG/UX differs from Solaris. John P. Rouillard (cs.umb.edu) deserves special mention for his persistent, but constructive, nagging about wrong or missing things, and for trying out and discussing embryonic code or ideas. Last but not least, Howard Chu (hanauma.jpl.nasa.gov), Darren Reed (coombs.anu.edu.au), Icarus Sparry (gdr.bath.ac.uk), Scott Schwartz (cs.psu.edu), John A. Kunze (violet.berkeley.edu), Daniel Len Schales (engr.latech.edu), Chris Turbeville (cse.uta.edu), Paul Kranenburg (cs.few.eur.nl), Marc Boucher (cam.org), Dave Mitchell (dcs.shef.ac.uk), Andrew Maffei, Adrian van Bloois, Rop Gonggrijp, John C. Wingenbach, Everett F. Batey and many, many others provided fixes, code fragments, or ideas for improvements. Wietse Venema (wietse@wzv.win.tue.nl) Department of Mathematics and Computing Science Eindhoven University of Technology P.O. Box 513 5600 MB Eindhoven The Netherlands Currently visiting IBM T.J. Watson Research, Hawthorne NY, USA. When you specify a host name, the `tcpdmatch' program will use both the host name and address. This way you can simulate the most common case where the wrappers know both the host address and the host name. The `tcpdmatch' program will iterate over all addresses that it can find for the given host name. When you specify a host address instead of a host name, the `tcpdmatch' program will pretend that the host name is unktcp_wrappers_7.6/miscd.c 444 465 506 6103 6107411333 10501 /* * Front end to the ULTRIX miscd service. The front end logs the remote host * name and then invokes the real miscd daemon. Install as "/usr/etc/miscd", * after renaming the real miscd daemon to the name defined with the * REAL_MISCD macro. * * Connections and diagnostics are logged through syslog(3). * * The Ultrix miscd program implements (among others) the systat service, which * pipes the output from who(1) to stdout. This information is potentially * useful to systems crackers. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) miscd.c 1.10 96/02/11 17:01:30"; #endif /* System libraries. */ #include #include #include #include #include #include #include #ifndef MAXPATHNAMELEN #define MAXPATHNAMELEN BUFSIZ #endif #ifndef STDIN_FILENO #define STDIN_FILENO 0 #endif /* Local stuff. */ #include "patchlevel.h" #include "tcpd.h" int allow_severity = SEVERITY; /* run-time adjustable */ int deny_severity = LOG_WARNING; /* ditto */ main(argc, argv) int argc; char **argv; { struct request_info request; char path[MAXPATHNAMELEN]; /* Attempt to prevent the creation of world-writable files. */ #ifdef DAEMON_UMASK umask(DAEMON_UMASK); #endif /* * Open a channel to the syslog daemon. Older versions of openlog() * require only two arguments. */ #ifdef LOG_MAIL (void) openlog(argv[0], LOG_PID, FACILITY); #else (void) openlog(argv[0], LOG_PID); #endif /* * Find out the endpoint addresses of this conversation. Host name * lookups and double checks will be done on demand. */ request_init(&request, RQ_DAEMON, argv[0], RQ_FILE, STDIN_FILENO, 0); fromhost(&request); /* * Optionally look up and double check the remote host name. Sites * concerned with security may choose to refuse connections from hosts * that pretend to have someone elses host name. */ #ifdef PARANOID if (STR_EQ(eval_hostname(request.client), paranoid)) refuse(&request); #endif /* * The BSD rlogin and rsh daemons that came out after 4.3 BSD disallow * socket options at the IP level. They do so for a good reason. * Unfortunately, we cannot use this with SunOS 4.1.x because the * getsockopt() system call can panic the system. */ #ifdef KILL_IP_OPTIONS fix_options(&request); #endif /* * Check whether this host can access the service in argv[0]. The * access-control code invokes optional shell commands as specified in * the access-control tables. */ #ifdef HOSTS_ACCESS if (!hosts_access(&request)) refuse(&request); #endif /* Report request and invoke the real daemon program. */ syslog(allow_severity, "connect from %s", eval_client(&request)); sprintf(path, "%s/miscd", REAL_DAEMON_DIR); closelog(); (void) execv(path, argv); syslog(LOG_ERR, "error: cannot execute %s: %m", path); clean_exit(&request); /* NOTREACHED */ } he remote host * name and then invokes the real miscd daemon. Install as "/usr/etc/miscd", * after renaming the real miscd daemon to the name defined with the * REAL_MISCD macro. * * Connections and diagnostics are logged through syslog(3). * * The Ultrix miscd program implements (among others) the systat service, which * pipes the output from who(1) to stdout. This information is potentially * useful to systems crackers.tcp_wrappers_7.6/tcpd.c 444 465 506 6451 6107411335 10344 /* * General front end for stream and datagram IP services. This program logs * the remote host name and then invokes the real daemon. For example, * install as /usr/etc/{tftpd,fingerd,telnetd,ftpd,rlogind,rshd,rexecd}, * after saving the real daemons in the directory specified with the * REAL_DAEMON_DIR macro. This arrangement requires that the network daemons * are started by inetd or something similar. Connections and diagnostics * are logged through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) tcpd.c 1.10 96/02/11 17:01:32"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #ifndef MAXPATHNAMELEN #define MAXPATHNAMELEN BUFSIZ #endif #ifndef STDIN_FILENO #define STDIN_FILENO 0 #endif /* Local stuff. */ #include "patchlevel.h" #include "tcpd.h" int allow_severity = SEVERITY; /* run-time adjustable */ int deny_severity = LOG_WARNING; /* ditto */ main(argc, argv) int argc; char **argv; { struct request_info request; char path[MAXPATHNAMELEN]; /* Attempt to prevent the creation of world-writable files. */ #ifdef DAEMON_UMASK umask(DAEMON_UMASK); #endif /* * If argv[0] is an absolute path name, ignore REAL_DAEMON_DIR, and strip * argv[0] to its basename. */ if (argv[0][0] == '/') { strcpy(path, argv[0]); argv[0] = strrchr(argv[0], '/') + 1; } else { sprintf(path, "%s/%s", REAL_DAEMON_DIR, argv[0]); } /* * Open a channel to the syslog daemon. Older versions of openlog() * require only two arguments. */ #ifdef LOG_MAIL (void) openlog(argv[0], LOG_PID, FACILITY); #else (void) openlog(argv[0], LOG_PID); #endif /* * Find out the endpoint addresses of this conversation. Host name * lookups and double checks will be done on demand. */ request_init(&request, RQ_DAEMON, argv[0], RQ_FILE, STDIN_FILENO, 0); fromhost(&request); /* * Optionally look up and double check the remote host name. Sites * concerned with security may choose to refuse connections from hosts * that pretend to have someone elses host name. */ #ifdef PARANOID if (STR_EQ(eval_hostname(request.client), paranoid)) refuse(&request); #endif /* * The BSD rlogin and rsh daemons that came out after 4.3 BSD disallow * socket options at the IP level. They do so for a good reason. * Unfortunately, we cannot use this with SunOS 4.1.x because the * getsockopt() system call can panic the system. */ #ifdef KILL_IP_OPTIONS fix_options(&request); #endif /* * Check whether this host can access the service in argv[0]. The * access-control code invokes optional shell commands as specified in * the access-control tables. */ #ifdef HOSTS_ACCESS if (!hosts_access(&request)) refuse(&request); #endif /* Report request and invoke the real daemon program. */ syslog(allow_severity, "connect from %s", eval_client(&request)); closelog(); (void) execv(path, argv); syslog(LOG_ERR, "error: cannot execute %s: %m", path); clean_exit(&request); /* NOTREACHED */ } * REAL_DAEMON_DIR macro. This arrangement requires that the network daemons * are started by inetd or something similar. Connections and diagnostics * are logged through syslog(3). * * Author: Wietse Venetcp_wrappers_7.6/fromhost.c 444 465 506 2577 5700312560 11256 /* * On socket-only systems, fromhost() is nothing but an alias for the * socket-specific sock_host() function. * * On systems with sockets and TLI, fromhost() determines the type of API * (sockets, TLI), then invokes the appropriate API-specific routines. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) fromhost.c 1.17 94/12/28 17:42:23"; #endif #if defined(TLI) || defined(PTX) || defined(TLI_SEQUENT) /* System libraries. */ #include #include #include /* Local stuff. */ #include "tcpd.h" /* fromhost - find out what network API we should use */ void fromhost(request) struct request_info *request; { /* * On systems with streams support the IP network protocol family may be * accessible via more than one programming interface: Berkeley sockets * and the Transport Level Interface (TLI). * * Thus, we must first find out what programming interface to use: sockets * or TLI. On some systems, sockets are not part of the streams system, * so if request->fd is not a stream we simply assume sockets. */ if (ioctl(request->fd, I_FIND, "timod") > 0) { tli_host(request); } else { sock_host(request); } } #endif /* TLI || PTX || TLI_SEQUENT */ s an absolute path name, ignore REAL_DAEMON_DIR, and strip * argv[0] to its basename. */ if (argv[0][0] == '/') { tcp_wrappers_7.6/hosts_access.c 444 465 506 22644 6300214463 12113 /* * This module implements a simple access control language that is based on * host (or domain) names, NIS (host) netgroup names, IP addresses (or * network numbers) and daemon process names. When a match is found the * search is terminated, and depending on whether PROCESS_OPTIONS is defined, * a list of options is executed or an optional shell command is executed. * * Host and user names are looked up on demand, provided that suitable endpoint * information is available as sockaddr_in structures or TLI netbufs. As a * side effect, the pattern matching process may change the contents of * request structure fields. * * Diagnostics are reported through syslog(3). * * Compile with -DNETGROUP if your library provides support for netgroups. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) hosts_access.c 1.21 97/02/12 02:13:22"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #include #include extern char *fgets(); extern int errno; #ifndef INADDR_NONE #define INADDR_NONE (-1) /* XXX should be 0xffffffff */ #endif /* Local stuff. */ #include "tcpd.h" /* Error handling. */ extern jmp_buf tcpd_buf; /* Delimiters for lists of daemons or clients. */ static char sep[] = ", \t\r\n"; /* Constants to be used in assignments only, not in comparisons... */ #define YES 1 #define NO 0 /* * These variables are globally visible so that they can be redirected in * verification mode. */ char *hosts_allow_table = HOSTS_ALLOW; char *hosts_deny_table = HOSTS_DENY; int hosts_access_verbose = 0; /* * In a long-running process, we are not at liberty to just go away. */ int resident = (-1); /* -1, 0: unknown; +1: yes */ /* Forward declarations. */ static int table_match(); static int list_match(); static int server_match(); static int client_match(); static int host_match(); static int string_match(); static int masked_match(); /* Size of logical line buffer. */ #define BUFLEN 2048 /* hosts_access - host access control facility */ int hosts_access(request) struct request_info *request; { int verdict; /* * If the (daemon, client) pair is matched by an entry in the file * /etc/hosts.allow, access is granted. Otherwise, if the (daemon, * client) pair is matched by an entry in the file /etc/hosts.deny, * access is denied. Otherwise, access is granted. A non-existent * access-control file is treated as an empty file. * * After a rule has been matched, the optional language extensions may * decide to grant or refuse service anyway. Or, while a rule is being * processed, a serious error is found, and it seems better to play safe * and deny service. All this is done by jumping back into the * hosts_access() routine, bypassing the regular return from the * table_match() function calls below. */ if (resident <= 0) resident++; verdict = setjmp(tcpd_buf); if (verdict != 0) return (verdict == AC_PERMIT); if (table_match(hosts_allow_table, request)) return (YES); if (table_match(hosts_deny_table, request)) return (NO); return (YES); } /* table_match - match table entries with (daemon, client) pair */ static int table_match(table, request) char *table; struct request_info *request; { FILE *fp; char sv_list[BUFLEN]; /* becomes list of daemons */ char *cl_list; /* becomes list of clients */ char *sh_cmd; /* becomes optional shell command */ int match = NO; struct tcpd_context saved_context; saved_context = tcpd_context; /* stupid compilers */ /* * Between the fopen() and fclose() calls, avoid jumps that may cause * file descriptor leaks. */ if ((fp = fopen(table, "r")) != 0) { tcpd_context.file = table; tcpd_context.line = 0; while (match == NO && xgets(sv_list, sizeof(sv_list), fp) != 0) { if (sv_list[strlen(sv_list) - 1] != '\n') { tcpd_warn("missing newline or line too long"); continue; } if (sv_list[0] == '#' || sv_list[strspn(sv_list, " \t\r\n")] == 0) continue; if ((cl_list = split_at(sv_list, ':')) == 0) { tcpd_warn("missing \":\" separator"); continue; } sh_cmd = split_at(cl_list, ':'); match = list_match(sv_list, request, server_match) && list_match(cl_list, request, client_match); } (void) fclose(fp); } else if (errno != ENOENT) { tcpd_warn("cannot open %s: %m", table); } if (match) { if (hosts_access_verbose > 1) syslog(LOG_DEBUG, "matched: %s line %d", tcpd_context.file, tcpd_context.line); if (sh_cmd) { #ifdef PROCESS_OPTIONS process_options(sh_cmd, request); #else char cmd[BUFSIZ]; shell_cmd(percent_x(cmd, sizeof(cmd), sh_cmd, request)); #endif } } tcpd_context = saved_context; return (match); } /* list_match - match a request against a list of patterns with exceptions */ static int list_match(list, request, match_fn) char *list; struct request_info *request; int (*match_fn) (); { char *tok; /* * Process tokens one at a time. We have exhausted all possible matches * when we reach an "EXCEPT" token or the end of the list. If we do find * a match, look for an "EXCEPT" list and recurse to determine whether * the match is affected by any exceptions. */ for (tok = strtok(list, sep); tok != 0; tok = strtok((char *) 0, sep)) { if (STR_EQ(tok, "EXCEPT")) /* EXCEPT: give up */ return (NO); if (match_fn(tok, request)) { /* YES: look for exceptions */ while ((tok = strtok((char *) 0, sep)) && STR_NE(tok, "EXCEPT")) /* VOID */ ; return (tok == 0 || list_match((char *) 0, request, match_fn) == 0); } } return (NO); } /* server_match - match server information */ static int server_match(tok, request) char *tok; struct request_info *request; { char *host; if ((host = split_at(tok + 1, '@')) == 0) { /* plain daemon */ return (string_match(tok, eval_daemon(request))); } else { /* daemon@host */ return (string_match(tok, eval_daemon(request)) && host_match(host, request->server)); } } /* client_match - match client information */ static int client_match(tok, request) char *tok; struct request_info *request; { char *host; if ((host = split_at(tok + 1, '@')) == 0) { /* plain host */ return (host_match(tok, request->client)); } else { /* user@host */ return (host_match(host, request->client) && string_match(tok, eval_user(request))); } } /* host_match - match host name and/or address against pattern */ static int host_match(tok, host) char *tok; struct host_info *host; { char *mask; /* * This code looks a little hairy because we want to avoid unnecessary * hostname lookups. * * The KNOWN pattern requires that both address AND name be known; some * patterns are specific to host names or to host addresses; all other * patterns are satisfied when either the address OR the name match. */ if (tok[0] == '@') { /* netgroup: look it up */ #ifdef NETGROUP static char *mydomain = 0; if (mydomain == 0) yp_get_default_domain(&mydomain); return (innetgr(tok + 1, eval_hostname(host), (char *) 0, mydomain)); #else tcpd_warn("netgroup support is disabled"); /* not tcpd_jump() */ return (NO); #endif } else if (STR_EQ(tok, "KNOWN")) { /* check address and name */ char *name = eval_hostname(host); return (STR_NE(eval_hostaddr(host), unknown) && HOSTNAME_KNOWN(name)); } else if (STR_EQ(tok, "LOCAL")) { /* local: no dots in name */ char *name = eval_hostname(host); return (strchr(name, '.') == 0 && HOSTNAME_KNOWN(name)); } else if ((mask = split_at(tok, '/')) != 0) { /* net/mask */ return (masked_match(tok, mask, eval_hostaddr(host))); } else { /* anything else */ return (string_match(tok, eval_hostaddr(host)) || (NOT_INADDR(tok) && string_match(tok, eval_hostname(host)))); } } /* string_match - match string against pattern */ static int string_match(tok, string) char *tok; char *string; { int n; if (tok[0] == '.') { /* suffix */ n = strlen(string) - strlen(tok); return (n > 0 && STR_EQ(tok, string + n)); } else if (STR_EQ(tok, "ALL")) { /* all: match any */ return (YES); } else if (STR_EQ(tok, "KNOWN")) { /* not unknown */ return (STR_NE(string, unknown)); } else if (tok[(n = strlen(tok)) - 1] == '.') { /* prefix */ return (STRN_EQ(tok, string, n)); } else { /* exact match */ return (STR_EQ(tok, string)); } } /* masked_match - match address against netnumber/netmask */ static int masked_match(net_tok, mask_tok, string) char *net_tok; char *mask_tok; char *string; { unsigned long net; unsigned long mask; unsigned long addr; /* * Disallow forms other than dotted quad: the treatment that inet_addr() * gives to forms with less than four components is inconsistent with the * access control language. John P. Rouillard . */ if ((addr = dot_quad_addr(string)) == INADDR_NONE) return (NO); if ((net = dot_quad_addr(net_tok)) == INADDR_NONE || (mask = dot_quad_addr(mask_tok)) == INADDR_NONE) { tcpd_warn("bad net/mask expression: %s/%s", net_tok, mask_tok); return (NO); /* not tcpd_jump() */ } return ((addr & mask) == net); } * the match is affected by any exceptions. */ for (tok = strtok(list, sep); tok !=tcp_wrappers_7.6/shell_cmd.c 444 465 506 4111 5700312604 11330 /* * shell_cmd() takes a shell command after % substitutions. The * command is executed by a /bin/sh child process, with standard input, * standard output and standard error connected to /dev/null. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) shell_cmd.c 1.5 94/12/28 17:42:44"; #endif /* System libraries. */ #include #include #include #include #include #include extern void exit(); /* Local stuff. */ #include "tcpd.h" /* Forward declarations. */ static void do_child(); /* shell_cmd - execute shell command */ void shell_cmd(command) char *command; { int child_pid; int wait_pid; /* * Most of the work is done within the child process, to minimize the * risk of damage to the parent. */ switch (child_pid = fork()) { case -1: /* error */ tcpd_warn("cannot fork: %m"); break; case 00: /* child */ do_child(command); /* NOTREACHED */ default: /* parent */ while ((wait_pid = wait((int *) 0)) != -1 && wait_pid != child_pid) /* void */ ; } } /* do_child - exec command with { stdin, stdout, stderr } to /dev/null */ static void do_child(command) char *command; { char *error; int tmp_fd; /* * Systems with POSIX sessions may send a SIGHUP to grandchildren if the * child exits first. This is sick, sessions were invented for terminals. */ signal(SIGHUP, SIG_IGN); /* Set up new stdin, stdout, stderr, and exec the shell command. */ for (tmp_fd = 0; tmp_fd < 3; tmp_fd++) (void) close(tmp_fd); if (open("/dev/null", 2) != 0) { error = "open /dev/null: %m"; } else if (dup(0) != 1 || dup(0) != 2) { error = "dup: %m"; } else { (void) execl("/bin/sh", "sh", "-c", command, (char *) 0); error = "execl /bin/sh: %m"; } /* Something went wrong. We MUST terminate the child process. */ tcpd_warn(error); _exit(0); } her the address OR the name match. */ if (tok[0] == '@') { /* netgroup: look it up */ #ifdef NETGROUP static char *mydomain = 0; if (mydomain == 0) yp_get_default_domain(&mydomain); return (innetgr(tok + 1, eval_hostname(host), (char *) 0, mydomain)); #else tcpd_warn("netgroup support is disabled"); /* not tcpd_jump() */ return (NO); #endif } else if (STR_EQ(tok, "KNOWN")) { /* check address and name */ chatcp_wrappers_7.6/tcpd.h 444 465 506 17302 6123550461 10370 /* * @(#) tcpd.h 1.5 96/03/19 16:22:24 * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ /* Structure to describe one communications endpoint. */ #define STRING_LENGTH 128 /* hosts, users, processes */ struct host_info { char name[STRING_LENGTH]; /* access via eval_hostname(host) */ char addr[STRING_LENGTH]; /* access via eval_hostaddr(host) */ struct sockaddr_in *sin; /* socket address or 0 */ struct t_unitdata *unit; /* TLI transport address or 0 */ struct request_info *request; /* for shared information */ }; /* Structure to describe what we know about a service request. */ struct request_info { int fd; /* socket handle */ char user[STRING_LENGTH]; /* access via eval_user(request) */ char daemon[STRING_LENGTH]; /* access via eval_daemon(request) */ char pid[10]; /* access via eval_pid(request) */ struct host_info client[1]; /* client endpoint info */ struct host_info server[1]; /* server endpoint info */ void (*sink) (); /* datagram sink function or 0 */ void (*hostname) (); /* address to printable hostname */ void (*hostaddr) (); /* address to printable address */ void (*cleanup) (); /* cleanup function or 0 */ struct netconfig *config; /* netdir handle */ }; /* Common string operations. Less clutter should be more readable. */ #define STRN_CPY(d,s,l) { strncpy((d),(s),(l)); (d)[(l)-1] = 0; } #define STRN_EQ(x,y,l) (strncasecmp((x),(y),(l)) == 0) #define STRN_NE(x,y,l) (strncasecmp((x),(y),(l)) != 0) #define STR_EQ(x,y) (strcasecmp((x),(y)) == 0) #define STR_NE(x,y) (strcasecmp((x),(y)) != 0) /* * Initially, all above strings have the empty value. Information that * cannot be determined at runtime is set to "unknown", so that we can * distinguish between `unavailable' and `not yet looked up'. A hostname * that we do not believe in is set to "paranoid". */ #define STRING_UNKNOWN "unknown" /* lookup failed */ #define STRING_PARANOID "paranoid" /* hostname conflict */ extern char unknown[]; extern char paranoid[]; #define HOSTNAME_KNOWN(s) (STR_NE((s),unknown) && STR_NE((s),paranoid)) #define NOT_INADDR(s) (s[strspn(s,"01234567890./")] != 0) /* Global functions. */ #if defined(TLI) || defined(PTX) || defined(TLI_SEQUENT) extern void fromhost(); /* get/validate client host info */ #else #define fromhost sock_host /* no TLI support needed */ #endif extern int hosts_access(); /* access control */ extern void shell_cmd(); /* execute shell command */ extern char *percent_x(); /* do % expansion */ extern void rfc931(); /* client name from RFC 931 daemon */ extern void clean_exit(); /* clean up and exit */ extern void refuse(); /* clean up and exit */ extern char *xgets(); /* fgets() on steroids */ extern char *split_at(); /* strchr() and split */ extern unsigned long dot_quad_addr(); /* restricted inet_addr() */ /* Global variables. */ extern int allow_severity; /* for connection logging */ extern int deny_severity; /* for connection logging */ extern char *hosts_allow_table; /* for verification mode redirection */ extern char *hosts_deny_table; /* for verification mode redirection */ extern int hosts_access_verbose; /* for verbose matching mode */ extern int rfc931_timeout; /* user lookup timeout */ extern int resident; /* > 0 if resident process */ /* * Routines for controlled initialization and update of request structure * attributes. Each attribute has its own key. */ #ifdef __STDC__ extern struct request_info *request_init(struct request_info *,...); extern struct request_info *request_set(struct request_info *,...); #else extern struct request_info *request_init(); /* initialize request */ extern struct request_info *request_set(); /* update request structure */ #endif #define RQ_FILE 1 /* file descriptor */ #define RQ_DAEMON 2 /* server process (argv[0]) */ #define RQ_USER 3 /* client user name */ #define RQ_CLIENT_NAME 4 /* client host name */ #define RQ_CLIENT_ADDR 5 /* client host address */ #define RQ_CLIENT_SIN 6 /* client endpoint (internal) */ #define RQ_SERVER_NAME 7 /* server host name */ #define RQ_SERVER_ADDR 8 /* server host address */ #define RQ_SERVER_SIN 9 /* server endpoint (internal) */ /* * Routines for delayed evaluation of request attributes. Each attribute * type has its own access method. The trivial ones are implemented by * macros. The other ones are wrappers around the transport-specific host * name, address, and client user lookup methods. The request_info and * host_info structures serve as caches for the lookup results. */ extern char *eval_user(); /* client user */ extern char *eval_hostname(); /* printable hostname */ extern char *eval_hostaddr(); /* printable host address */ extern char *eval_hostinfo(); /* host name or address */ extern char *eval_client(); /* whatever is available */ extern char *eval_server(); /* whatever is available */ #define eval_daemon(r) ((r)->daemon) /* daemon process name */ #define eval_pid(r) ((r)->pid) /* process id */ /* Socket-specific methods, including DNS hostname lookups. */ extern void sock_host(); /* look up endpoint addresses */ extern void sock_hostname(); /* translate address to hostname */ extern void sock_hostaddr(); /* address to printable address */ #define sock_methods(r) \ { (r)->hostname = sock_hostname; (r)->hostaddr = sock_hostaddr; } /* The System V Transport-Level Interface (TLI) interface. */ #if defined(TLI) || defined(PTX) || defined(TLI_SEQUENT) extern void tli_host(); /* look up endpoint addresses etc. */ #endif /* * Problem reporting interface. Additional file/line context is reported * when available. The jump buffer (tcpd_buf) is not declared here, or * everyone would have to include . */ #ifdef __STDC__ extern void tcpd_warn(char *, ...); /* report problem and proceed */ extern void tcpd_jump(char *, ...); /* report problem and jump */ #else extern void tcpd_warn(); extern void tcpd_jump(); #endif struct tcpd_context { char *file; /* current file */ int line; /* current line */ }; extern struct tcpd_context tcpd_context; /* * While processing access control rules, error conditions are handled by * jumping back into the hosts_access() routine. This is cleaner than * checking the return value of each and every silly little function. The * (-1) returns are here because zero is already taken by longjmp(). */ #define AC_PERMIT 1 /* permit access */ #define AC_DENY (-1) /* deny_access */ #define AC_ERROR AC_DENY /* XXX */ /* * In verification mode an option function should just say what it would do, * instead of really doing it. An option function that would not return * should clear the dry_run flag to inform the caller of this unusual * behavior. */ extern void process_options(); /* execute options */ extern int dry_run; /* verification flag */ /* Bug workarounds. */ #ifdef INET_ADDR_BUG /* inet_addr() returns struct */ #define inet_addr fix_inet_addr extern long fix_inet_addr(); #endif #ifdef BROKEN_FGETS /* partial reads from sockets */ #define fgets fix_fgets extern char *fix_fgets(); #endif #ifdef RECVFROM_BUG /* no address family info */ #define recvfrom fix_recvfrom extern int fix_recvfrom(); #endif #ifdef GETPEERNAME_BUG /* claims success with UDP */ #define getpeername fix_getpeername extern int fix_getpeername(); #endif #ifdef SOLARIS_24_GETHOSTBYNAME_BUG /* lists addresses as aliases */ #define gethostbyname fix_gethostbyname extern struct hostent *fix_gethostbyname(); #endif #ifdef USE_STRSEP /* libc calls strtok() */ #define strtok fix_strtok extern char *fix_strtok(); #endif #ifdef LIBC_CALLS_STRTOK /* libc calls strtok() */ #define strtok my_strtok extern char *my_strtok(); #endif uct request_info *request_set(); /* update request structure */ #endif #define RQ_FILE 1 /* file descriptor */ #define RQ_DAEMON 2 /* server process (argv[0]) */ #define RQ_USER 3 /* client user name */ #define RQ_CLIENT_NAME 4 /* client host name */ #define RQ_CLIENT_ADDR 5 /* client host address */ #define tcp_wrappers_7.6/tcpdmatch.c 444 465 506 22036 6107411340 11372 /* * tcpdmatch - explain what tcpd would do in a specific case * * usage: tcpdmatch [-d] [-i inet_conf] daemon[@host] [user@]host * * -d: use the access control tables in the current directory. * * -i: location of inetd.conf file. * * All errors are reported to the standard error stream, including the errors * that would normally be reported via the syslog daemon. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) tcpdmatch.c 1.5 96/02/11 17:01:36"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #include #include extern void exit(); extern int optind; extern char *optarg; #ifndef INADDR_NONE #define INADDR_NONE (-1) /* XXX should be 0xffffffff */ #endif #ifndef S_ISDIR #define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) #endif /* Application-specific. */ #include "tcpd.h" #include "inetcf.h" #include "scaffold.h" static void usage(); static void tcpdmatch(); /* The main program */ int main(argc, argv) int argc; char **argv; { struct hostent *hp; char *myname = argv[0]; char *client; char *server; char *addr; char *user; char *daemon; struct request_info request; int ch; char *inetcf = 0; int count; struct sockaddr_in server_sin; struct sockaddr_in client_sin; struct stat st; /* * Show what rule actually matched. */ hosts_access_verbose = 2; /* * Parse the JCL. */ while ((ch = getopt(argc, argv, "di:")) != EOF) { switch (ch) { case 'd': hosts_allow_table = "hosts.allow"; hosts_deny_table = "hosts.deny"; break; case 'i': inetcf = optarg; break; default: usage(myname); /* NOTREACHED */ } } if (argc != optind + 2) usage(myname); /* * When confusion really strikes... */ if (check_path(REAL_DAEMON_DIR, &st) < 0) { tcpd_warn("REAL_DAEMON_DIR %s: %m", REAL_DAEMON_DIR); } else if (!S_ISDIR(st.st_mode)) { tcpd_warn("REAL_DAEMON_DIR %s is not a directory", REAL_DAEMON_DIR); } /* * Default is to specify a daemon process name. When daemon@host is * specified, separate the two parts. */ if ((server = split_at(argv[optind], '@')) == 0) server = unknown; if (argv[optind][0] == '/') { daemon = strrchr(argv[optind], '/') + 1; tcpd_warn("%s: daemon name normalized to: %s", argv[optind], daemon); } else { daemon = argv[optind]; } /* * Default is to specify a client hostname or address. When user@host is * specified, separate the two parts. */ if ((client = split_at(argv[optind + 1], '@')) != 0) { user = argv[optind + 1]; } else { client = argv[optind + 1]; user = unknown; } /* * Analyze the inetd (or tlid) configuration file, so that we can warn * the user about services that may not be wrapped, services that are not * configured, or services that are wrapped in an incorrect manner. Allow * for services that are not run from inetd, or that have tcpd access * control built into them. */ inetcf = inet_cfg(inetcf); inet_set("portmap", WR_NOT); inet_set("rpcbind", WR_NOT); switch (inet_get(daemon)) { case WR_UNKNOWN: tcpd_warn("%s: no such process name in %s", daemon, inetcf); break; case WR_NOT: tcpd_warn("%s: service possibly not wrapped", daemon); break; } /* * Check accessibility of access control files. */ (void) check_path(hosts_allow_table, &st); (void) check_path(hosts_deny_table, &st); /* * Fill in what we have figured out sofar. Use socket and DNS routines * for address and name conversions. We attach stdout to the request so * that banner messages will become visible. */ request_init(&request, RQ_DAEMON, daemon, RQ_USER, user, RQ_FILE, 1, 0); sock_methods(&request); /* * If a server hostname is specified, insist that the name maps to at * most one address. eval_hostname() warns the user about name server * problems, while using the request.server structure as a cache for host * address and name conversion results. */ if (NOT_INADDR(server) == 0 || HOSTNAME_KNOWN(server)) { if ((hp = find_inet_addr(server)) == 0) exit(1); memset((char *) &server_sin, 0, sizeof(server_sin)); server_sin.sin_family = AF_INET; request_set(&request, RQ_SERVER_SIN, &server_sin, 0); for (count = 0; (addr = hp->h_addr_list[count]) != 0; count++) { memcpy((char *) &server_sin.sin_addr, addr, sizeof(server_sin.sin_addr)); /* * Force evaluation of server host name and address. Host name * conflicts will be reported while eval_hostname() does its job. */ request_set(&request, RQ_SERVER_NAME, "", RQ_SERVER_ADDR, "", 0); if (STR_EQ(eval_hostname(request.server), unknown)) tcpd_warn("host address %s->name lookup failed", eval_hostaddr(request.server)); } if (count > 1) { fprintf(stderr, "Error: %s has more than one address\n", server); fprintf(stderr, "Please specify an address instead\n"); exit(1); } free((char *) hp); } else { request_set(&request, RQ_SERVER_NAME, server, 0); } /* * If a client address is specified, we simulate the effect of client * hostname lookup failure. */ if (dot_quad_addr(client) != INADDR_NONE) { request_set(&request, RQ_CLIENT_ADDR, client, 0); tcpdmatch(&request); exit(0); } /* * Perhaps they are testing special client hostname patterns that aren't * really host names at all. */ if (NOT_INADDR(client) && HOSTNAME_KNOWN(client) == 0) { request_set(&request, RQ_CLIENT_NAME, client, 0); tcpdmatch(&request); exit(0); } /* * Otherwise, assume that a client hostname is specified, and insist that * the address can be looked up. The reason for this requirement is that * in real life the client address is available (at least with IP). Let * eval_hostname() figure out if this host is properly registered, while * using the request.client structure as a cache for host name and * address conversion results. */ if ((hp = find_inet_addr(client)) == 0) exit(1); memset((char *) &client_sin, 0, sizeof(client_sin)); client_sin.sin_family = AF_INET; request_set(&request, RQ_CLIENT_SIN, &client_sin, 0); for (count = 0; (addr = hp->h_addr_list[count]) != 0; count++) { memcpy((char *) &client_sin.sin_addr, addr, sizeof(client_sin.sin_addr)); /* * Force evaluation of client host name and address. Host name * conflicts will be reported while eval_hostname() does its job. */ request_set(&request, RQ_CLIENT_NAME, "", RQ_CLIENT_ADDR, "", 0); if (STR_EQ(eval_hostname(request.client), unknown)) tcpd_warn("host address %s->name lookup failed", eval_hostaddr(request.client)); tcpdmatch(&request); if (hp->h_addr_list[count + 1]) printf("\n"); } free((char *) hp); exit(0); } /* Explain how to use this program */ static void usage(myname) char *myname; { fprintf(stderr, "usage: %s [-d] [-i inet_conf] daemon[@host] [user@]host\n", myname); fprintf(stderr, " -d: use allow/deny files in current directory\n"); fprintf(stderr, " -i: location of inetd.conf file\n"); exit(1); } /* Print interesting expansions */ static void expand(text, pattern, request) char *text; char *pattern; struct request_info *request; { char buf[BUFSIZ]; if (STR_NE(percent_x(buf, sizeof(buf), pattern, request), unknown)) printf("%s %s\n", text, buf); } /* Try out a (server,client) pair */ static void tcpdmatch(request) struct request_info *request; { int verdict; /* * Show what we really know. Suppress uninteresting noise. */ expand("client: hostname", "%n", request); expand("client: address ", "%a", request); expand("client: username", "%u", request); expand("server: hostname", "%N", request); expand("server: address ", "%A", request); expand("server: process ", "%d", request); /* * Reset stuff that might be changed by options handlers. In dry-run * mode, extension language routines that would not return should inform * us of their plan, by clearing the dry_run flag. This is a bit clumsy * but we must be able to verify hosts with more than one network * address. */ rfc931_timeout = RFC931_TIMEOUT; allow_severity = SEVERITY; deny_severity = LOG_WARNING; dry_run = 1; /* * When paranoid mode is enabled, access is rejected no matter what the * access control rules say. */ #ifdef PARANOID if (STR_EQ(eval_hostname(request->client), paranoid)) { printf("access: denied (PARANOID mode)\n\n"); return; } #endif /* * Report the access control verdict. */ verdict = hosts_access(request); printf("access: %s\n", dry_run == 0 ? "delegated" : verdict ? "granted" : "denied"); } l_hostaddr(request.server)); } if (count > 1) { fprintf(stderr, "Error: %s has more than one address\n", server); fprintf(stderr, "Please specify an address instead\n"); exit(1); } free((char *) hp); } else { request_set(&request, RQ_SERVER_NAME, server, 0); } /* * If a client address is specified, we simulate the effect of client * hostname lookup failure. */ if (dot_quad_addr(client) != INADDR_NONE) { request_set(&request, tcp_wrappers_7.6/Makefile 444 465 506 77320 6314551411 10731 # @(#) Makefile 1.23 97/03/21 19:27:20 what: @echo @echo "Usage: edit the REAL_DAEMON_DIR definition in the Makefile then:" @echo @echo " make sys-type" @echo @echo "If you are in a hurry you can try instead:" @echo @echo " make REAL_DAEMON_DIR=/foo/bar sys-type" @echo @echo "And for a version with language extensions enabled:" @echo @echo " make REAL_DAEMON_DIR=/foo/bar STYLE=-DPROCESS_OPTIONS sys-type" @echo @echo "This Makefile knows about the following sys-types:" @echo @echo " generic (most bsd-ish systems with sys5 compatibility)" @echo " 386bsd aix alpha apollo bsdos convex-ultranet dell-gcc dgux dgux543" @echo " dynix epix esix freebsd hpux irix4 irix5 irix6 isc iunix" @echo " linux machten mips(untested) ncrsvr4 netbsd next osf power_unix_211" @echo " ptx-2.x ptx-generic pyramid sco sco-nis sco-od2 sco-os5 sinix sunos4" @echo " sunos40 sunos5 sysv4 tandem ultrix unicos7 unicos8 unixware1 unixware2" @echo " uts215 uxp" @echo @echo "If none of these match your environment, edit the system" @echo "dependencies sections in the Makefile and do a 'make other'." @echo ####################################################### # Choice between easy and advanced installation recipe. # # Advanced installation: vendor-provided daemons are left alone, and the # inetd configuration file is edited. In this case, the REAL_DAEMON_DIR # macro should reflect the actual directory with (most of) your # vendor-provided network daemons. These names can be found in the # inetd.conf file. Usually, the telnet, ftp and finger daemons all live # in the same directory. # # Uncomment the appropriate line if you are going to edit inetd.conf. # # Ultrix 4.x SunOS 4.x ConvexOS 10.x Dynix/ptx #REAL_DAEMON_DIR=/usr/etc # # SysV.4 Solaris 2.x OSF AIX #REAL_DAEMON_DIR=/usr/sbin # # BSD 4.4 #REAL_DAEMON_DIR=/usr/libexec # # HP-UX SCO Unicos #REAL_DAEMON_DIR=/etc # Easy installation: vendor-provided network daemons are moved to "some # other" directory, and the tcpd wrapper fills in the "holes". For this # mode of operation, the REAL_DAEMON_DIR macro should be set to the "some # other" directory. The "..." is here for historical reasons only; you # should probably use some other name. # # Uncomment the appropriate line if you are going to move your daemons. # # Ultrix 4.x SunOS 4.x ConvexOS 10.x Dynix/ptx #REAL_DAEMON_DIR=/usr/etc/... # # SysV.4 Solaris 2.x OSF AIX #REAL_DAEMON_DIR=/usr/sbin/... # # BSD 4.4 #REAL_DAEMON_DIR=/usr/libexec/... # # HP-UX SCO Unicos #REAL_DAEMON_DIR=/etc/... # End of mandatory section ########################## ########################################## # Ready-to-use system-dependent templates. # # Ready-to-use templates are available for many systems (see the "echo" # commands at the start of this Makefile). The templates take care of # all system dependencies: after editing the REAL_DAEMON_DIR definition # above, do a "make sunos4" (or whatever system type is appropriate). # # If your system is not listed (or something that comes close enough), you # have to edit the system dependencies section below and do a "make other". # # Send templates for other UNIX versions to wietse@wzv.win.tue.nl. # This is good for many BSD+SYSV hybrids with NIS (formerly YP). generic aix osf alpha dynix: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= all # Ditto, with vsyslog sunos4: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP VSYSLOG= TLI= all # Generic with resolver library. generic-resolver: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lresolv RANLIB=ranlib ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= all # The NeXT loader needs "-m" or it barfs on redefined library functions. next: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-m RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP=-DNETGROUP TLI= all # SunOS for the 386 was frozen at release 4.0.x. sunos40: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ="setenv.o strcasecmp.o" \ NETGROUP=-DNETGROUP VSYSLOG= TLI= all # Ultrix is like aix, next, etc., but has miscd and setenv(). ultrix: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ= \ NETGROUP=-DNETGROUP TLI= all miscd # This works on EP/IX 1.4.3 and will likely work on Mips (reggers@julian.uwo.ca) epix: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP=-DNETGROUP TLI= SYSTYPE="-systype bsd43" all # Freebsd and linux by default have no NIS. 386bsd netbsd bsdos: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ= NETGROUP= TLI= \ EXTRA_CFLAGS=-DSYS_ERRLIST_DEFINED VSYSLOG= all freebsd: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ= NETGROUP= TLI= \ EXTRA_CFLAGS=-DSYS_ERRLIST_DEFINED VSYSLOG= all linux: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP= TLI= EXTRA_CFLAGS="-DBROKEN_SO_LINGER" all # This is good for many SYSV+BSD hybrids with NIS, probably also for HP-UX 7.x. hpux hpux8 hpux9 hpux10: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=echo ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= all # ConvexOS-10.x with UltraNet support (ukkonen@csc.fi). convex-ultranet: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lulsock RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP=-DNETGROUP TLI= all # Generic support for the Dynix/PTX version of TLI. ptx-generic: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -linet -lnsl" RANLIB=echo ARFLAGS=rv \ AUX_OBJ="setenv.o strcasecmp.o ptx.o" NETGROUP= TLI=-DPTX all # With UDP support optimized for PTX 2.x (timw@sequent.com). ptx-2.x: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -linet -lnsl" RANLIB=echo ARFLAGS=rv \ AUX_OBJ="setenv.o strcasecmp.o tli-sequent.o" NETGROUP= \ TLI=-DTLI_SEQUENT all # IRIX 4.0.x has a special ar(1) flag. irix4: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lc -lsun" RANLIB=echo ARFLAGS=rvs AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= all # IRIX 5.2 is SYSV4 with several broken things (such as -lsocket -lnsl). irix5: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lsun RANLIB=echo ARFLAGS=rv VSYSLOG= \ NETGROUP=-DNETGROUP AUX_OBJ=setenv.o TLI= all # IRIX 6.2 (tucker@math.unc.edu). Must find a better value than 200000. irix6: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=echo ARFLAGS=rv VSYSLOG= \ NETGROUP=-DNETGROUP EXTRA_CFLAGS="-DBSD=200000" TLI= all # SunOS 5.x is another SYSV4 variant. sunos5: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl" RANLIB=echo ARFLAGS=rv VSYSLOG= \ NETGROUP=-DNETGROUP AUX_OBJ=setenv.o TLI=-DTLI \ BUGS="$(BUGS) -DSOLARIS_24_GETHOSTBYNAME_BUG" all # Generic SYSV40 esix sysv4: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl" RANLIB=echo ARFLAGS=rv \ NETGROUP=-DNETGROUP AUX_OBJ=setenv.o TLI=-DTLI all # DG/UX 5.4.1 and 5.4.2 have an unusual inet_addr() interface. dgux: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lnsl RANLIB=echo ARFLAGS=rv \ NETGROUP=-DNETGROUP AUX_OBJ=setenv.o TLI=-DTLI \ BUGS="$(BUGS) -DINET_ADDR_BUG" all dgux543: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lnsl RANLIB=echo ARFLAGS=rv \ NETGROUP=-DNETGROUP AUX_OBJ=setenv.o TLI=-DTLI all # NCR UNIX 02.02.01 and 02.03.00 (Alex Chircop, msu@unimt.mt) ncrsvr4: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lresolv -lnsl -lsocket" RANLIB=echo ARFLAGS=rv \ AUX_OBJ="setenv.o strcasecmp.o" NETGROUP= TLI=-DTLI \ EXTRA_CFLAGS="" FROM_OBJ=ncr.o all # Tandem SYSV4 (eqawas@hedgehog.ac.cowan.edu.au) tandem: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl" RANLIB=echo ARFLAGS=rv \ NETGROUP= AUX_OBJ="setenv.o strcasecmp.o" TLI=-DTLI all # Amdahl UTS 2.1.5 (Richard.Richmond@bridge.bst.bls.com) uts215: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket" RANLIB=echo \ ARFLAGS=rv AUX_OBJ=setenv.o NETGROUP=-DNO_NETGROUP TLI= all # UXP/DS System V.4 clone (vic@uida0.uida.es). uxp: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-L/usr/ucblib -lsocket -lnsl -lucb" \ RANLIB=echo ARFLAGS=rv NETGROUP=-DNETGROUP \ AUX_OBJ=setenv.o TLI="-DTLI -DDRS_XTI" all # DELL System V.4 Issue 2.2 using gcc (kim@tac.nyc.ny.us, jurban@norden1.com) dell-gcc: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl" RANLIB=ranlib ARFLAGS=rv CC=gcc \ AUX_OBJ="setenv.o strcasecmp.o" TLI=-DTLI all # SCO 3.2v4.1 no frills (jedwards@sol1.solinet.net). sco: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl_s" RANLIB=echo ARFLAGS=rv \ NETGROUP= AUX_OBJ=setenv.o TLI= all # SCO OpenDesktop 2.0, release 3.2 (peter@midnight.com). Please simplify. sco-od2: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lrpcsvc -lrpc -lyp -lrpc -lrpcsvc -lsocket" \ RANLIB=echo ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= all # SCO 3.2v4.2 with TCP/IP 1.2.1 (Eduard.Vopicka@vse.cz). Please simplify. sco-nis: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lyp -lrpc -lsocket -lyp -lc_s -lc" \ RANLIB=echo ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= EXTRA_CFLAGS="-nointl -DNO_NETGRENT" all # SCO 3.2v5.0.0 OpenServer 5 (bob@odt.handy.com, bill@razorlogic.com) sco-os5: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lrpcsvc -lsocket" RANLIB=echo ARFLAGS=rv VSYSLOG= \ AUX_OBJ=setenv.o NETGROUP=-DNETGROUP TLI= all # sinix 5.42 setjmp workaround (szrzs023@ub3.ub.uni-kiel.de) sinix: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl -L/usr/ccs/lib -lc -L/usr/ucblib -lucb" \ RANLIB=echo ARFLAGS=rv AUX_OBJ=setenv.o TLI=-DTLI all # Domain SR10.4. Build under bsd, run under either sysv3 or bsd43. apollo: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=setenv.o \ NETGROUP=-DNETGROUP TLI= SYSTYPE="-A run,any -A sys,any" all # Pyramid OSx 5.1, using the BSD universe. pyramid: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ="environ.o vfprintf.o" \ STRINGS="-Dstrchr=index -Dstrrchr=rindex -Dmemcmp=bcmp -Dno_memcpy" \ NETGROUP="-DNETGROUP -DUSE_GETDOMAIN" TLI= all # Untested. mips: @echo "Warning: some definitions may be wrong." make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP=-DNETGROUP TLI= SYSTYPE="-sysname bsd43" all # Cray (tested with UNICOS 7.0.4). unicos7: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS=-lnet RANLIB=echo ARFLAGS=rv \ EXTRA_CFLAGS=-DINADDR_NONE="\"((unsigned long) -1)\"" \ AUX_OBJ="setenv.o strcasecmp.o" NETGROUP= TLI= all # Unicos 8.x, Cray-YMP (Bruce Kelly). unicos8: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=echo AR=bld ARFLAGS=rv \ AUX_OBJ= NETGROUP= TLI= all # Power_UNIX 2.1.1 (amantel@lerc.nasa.gov) power_unix_211: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lnsl -lsocket -lgen -lresolv" RANLIB=echo ARFLAGS=rv \ NETGROUP= AUX_OBJ=setenv.o TLI=-DTLI BUGS="$(BUGS)" all # ISC (fc@all.net) isc: make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-linet -lnsl_s -ldbm" RANLIB=echo ARFLAGS=rv \ AUX_OBJ="setenv.o strcasecmp.o" EXTRA_CFLAGS="-DENOTCONN=ENAVAIL" \ NETGROUP= TLI= all # Interactive UNIX R3.2 version 4.0 (Bobby D. Wright). iunix: make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-linet -lnsl_s -ldbm" RANLIB=echo ARFLAGS=rv \ AUX_OBJ=environ.o strcasecmp.o NETGROUP= TLI= all # RTU 6.0 on a Masscomp 5400 (ben@piglet.cr.usgs.gov). When using the # advanced installation, increment argv before actually looking at it. rtu: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP= TLI= all # Unixware sans NIS (mc@telebase.com). Compiler dislikes strcasecmp.c. unixware1: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl -lc -L/usr/ucblib -lucb" RANLIB=echo ARFLAGS=rv \ NETGROUP=$(NETGROUP) AUX_OBJ=environ.o TLI=-DTLI all unixware2: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl -lgen -lc -L/usr/ucblib -lucb" RANLIB=echo \ ARFLAGS=rv NETGROUP=$(NETGROUP) AUX_OBJ=environ.o TLI=-DTLI all u6000: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS="-lsocket -lnsl" RANLIB=echo ARFLAGS=rv \ NETGROUP=-DNETGROUP AUX_OBJ="setenv.o strcasecmp.o" TLI=-DTLI all # MachTen machten: @make REAL_DAEMON_DIR=$(REAL_DAEMON_DIR) STYLE=$(STYLE) \ LIBS= RANLIB=ranlib ARFLAGS=rv AUX_OBJ=environ.o \ NETGROUP= TLI= all ############################################################### # System dependencies: TLI (transport-level interface) support. # # Uncomment the following macro if your system has System V.4-style TLI # support (/usr/include/sys/timod.h, /etc/netconfig, and the netdir(3) # routines). # #TLI = -DTLI ############################################################################### # System dependencies: differences between ranlib(1) and ar(1) implementations. # # Some C compilers (Ultrix 4.x) insist that ranlib(1) be run on an object # library; some don't care as long as the modules are in the right order; # some systems don't even have a ranlib(1) command. Make your choice. RANLIB = ranlib # have ranlib (BSD-ish UNIX) #RANLIB = echo # no ranlib (SYSV-ish UNIX) ARFLAGS = rv # most systems #ARFLAGS= rvs # IRIX 4.0.x AR = ar #AR = bld # Unicos 8.x ############################################################################# # System dependencies: routines that are not present in the system libraries. # # If your system library does not have set/putenv() or strcasecmp(), use # the ones provided with this source distribution. The environ.c module # implements setenv(), getenv(), and putenv(). AUX_OBJ= setenv.o #AUX_OBJ= environ.o #AUX_OBJ= environ.o strcasecmp.o # Uncomment the following if your C library does not provide the # strchr/strrchr/memcmp routines, but comes with index/rindex/bcmp. # #STRINGS= -Dstrchr=index -Dstrrchr=rindex -Dmemcmp=bcmp -Dno_memcpy ################################################################# # System dependencies: selection of non-default object libraries. # # Most System V implementations require that you explicitly specify the # networking libraries. There is no general consensus, though. # #LIBS = -lsocket -lnsl # SysV.4 Solaris 2.x #LIBS = -lsun # IRIX #LIBS = -lsocket -linet -lnsl -lnfs # PTX #LIBS = -linet -lnsl_s -ldbm # ISC #LIBS = -lnet # Unicos 7 #LIBS = -linet -lsyslog -ldbm #LIBS = -lsyslog -lsocket -lnsl ###################################################### # System dependencies: system-specific compiler flags. # # Apollo Domain/OS offers both bsd and sys5 environments, sometimes # on the same machine. If your Apollo is primarily sys5.3 and also # has bsd4.3, uncomment the following to build under bsd and run under # either environment. # #SYSTYPE= -A run,any -A sys,any # For MIPS RISC/os 4_52.p3, uncomment the following definition. # #SYSTYPE= -sysname bsd43 ################################################## # System dependencies: working around system bugs. # # -DGETPEERNAME_BUG works around a getpeername(2) bug in some versions of # Apollo or SYSV.4 UNIX: the wrapper would report that all UDP requests # come from address 0.0.0.0. The workaround does no harm on other systems. # # -DBROKEN_FGETS works around an fgets(3) bug in some System V versions # (IRIX): fgets() gives up too fast when reading from a network socket. # The workaround does no harm on other systems. # # Some UNIX systems (IRIX) make the error of calling the strtok() library # routine from other library routines such as, e.g., gethostbyname/addr(). # The result is that hosts can slip through the wrapper allow/deny filters. # Compile with -DLIBC_CALLS_STRTOK to avoid the vendor's strtok() routine. # The workaround does no harm on other systems. # # DG/UX 5.4.1 comes with an inet_ntoa() function that returns a structure # instead of a long integer. Compile with -DINET_ADDR_BUG to work around # this mutant behavour. Fixed in 5.4R3. # # Solaris 2.4 gethostbyname(), in DNS through NIS mode, puts only one # address in the host address list; all other addresses are treated as # host name aliases. Compile with -DSOLARIS_24_GETHOSTBYNAME_BUG to work # around this. The workaround does no harm on other Solaris versions. BUGS = -DGETPEERNAME_BUG -DBROKEN_FGETS -DLIBC_CALLS_STRTOK #BUGS = -DGETPEERNAME_BUG -DBROKEN_FGETS -DINET_ADDR_BUG #BUGS = -DGETPEERNAME_BUG -DBROKEN_FGETS -DSOLARIS_24_GETHOSTBYNAME_BUG ########################################################################## # System dependencies: whether or not your system has NIS (or YP) support. # # If your system supports NIS or YP-style netgroups, enable the following # macro definition. Netgroups are used only for host access control. # #NETGROUP= -DNETGROUP ############################################################### # System dependencies: whether or not your system has vsyslog() # # If your system supports vsyslog(), comment out the following definition. # If in doubt leave it in, it won't harm. VSYSLOG = -Dvsyslog=myvsyslog # End of the system dependencies. ################################# ############################## # Start of the optional stuff. ########################################### # Optional: Turning on language extensions # # Instead of the default access control language that is documented in # the hosts_access.5 document, the wrappers can be configured to # implement an extensible language documented in the hosts_options.5 # document. This language is implemented by the "options.c" source # module, which also gives hints on how to add your own extensions. # Uncomment the next definition to turn on the language extensions # (examples: allow, deny, banners, twist and spawn). # #STYLE = -DPROCESS_OPTIONS # Enable language extensions. ################################################################ # Optional: Changing the default disposition of logfile records # # By default, logfile entries are written to the same file as used for # sendmail transaction logs. See your /etc/syslog.conf file for actual # path names of logfiles. The tutorial section in the README file # gives a brief introduction to the syslog daemon. # # Change the FACILITY definition below if you disagree with the default # disposition. Some syslog versions (including Ultrix 4.x) do not provide # this flexibility. # # If nothing shows up on your system, it may be that the syslog records # are sent to a dedicated loghost. It may also be that no syslog daemon # is running at all. The README file gives pointers to surrogate syslog # implementations for systems that have no syslog library routines or # no syslog daemons. When changing the syslog.conf file, remember that # there must be TABs between fields. # # The LOG_XXX names below are taken from the /usr/include/syslog.h file. FACILITY= LOG_MAIL # LOG_MAIL is what most sendmail daemons use # The syslog priority at which successful connections are logged. SEVERITY= LOG_INFO # LOG_INFO is normally not logged to the console ########################### # Optional: Reduce DNS load # # When looking up the address for a host.domain name, the typical DNS # code will first append substrings of your own domain, so it tries # host.domain.your.own.domain, then host.domain.own.domain, and then # host.domain. The APPEND_DOT feature stops this waste of cycles. It is # off by default because it causes problems on sites that don't use DNS # and with Solaris < 2.4. APPEND_DOT will not work with hostnames taken # from /etc/hosts or from NIS maps. It does work with DNS through NIS. # # DOT= -DAPPEND_DOT ################################################## # Optional: Always attempt remote username lookups # # By default, the wrappers look up the remote username only when the # access control rules require them to do so. # # Username lookups require that the remote host runs a daemon that # supports an RFC 931 like protocol. Remote user name lookups are not # possible for UDP-based connections, and can cause noticeable delays # with connections from non-UNIX PCs. On some systems, remote username # lookups can trigger a kernel bug, causing loss of service. The README # file describes how to find out if your UNIX kernel has that problem. # # Uncomment the following definition if the wrappers should always # attempt to get the remote user name. If this is not enabled you can # still do selective username lookups as documented in the hosts_access.5 # and hosts_options.5 manual pages (`nroff -man' format). # #AUTH = -DALWAYS_RFC931 # # The default username lookup timeout is 10 seconds. This may not be long # enough for slow hosts or networks, but is enough to irritate PC users. RFC931_TIMEOUT = 10 ###################################################### # Optional: Changing the default file protection mask # # On many systems, network daemons and other system processes are started # with a zero umask value, so that world-writable files may be produced. # It is a good idea to edit your /etc/rc* files so that they begin with # an explicit umask setting. On our site we use `umask 022' because it # does not break anything yet gives adequate protection against tampering. # # The following macro specifies the default umask for processes run under # control of the daemon wrappers. Comment it out only if you are certain # that inetd and its children are started with a safe umask value. UMASK = -DDAEMON_UMASK=022 ####################################### # Optional: Turning off access control # # By default, host access control is enabled. To disable host access # control, comment out the following definition. Host access control # can also be turned off at runtime by providing no or empty access # control tables. ACCESS = -DHOSTS_ACCESS ######################################################## # Optional: Changing the access control table pathnames # # The HOSTS_ALLOW and HOSTS_DENY macros define where the programs will # look for access control information. Watch out for the quotes and # backslashes when you make changes. TABLES = -DHOSTS_DENY=\"/etc/hosts.deny\" -DHOSTS_ALLOW=\"/etc/hosts.allow\" #################################################### # Optional: dealing with host name/address conflicts # # By default, the software tries to protect against hosts that claim to # have someone elses host name. This is relevant for network services # whose authentication depends on host names, such as rsh and rlogin. # # With paranoid mode on, connections will be rejected when the host name # does not match the host address. Connections will also be rejected when # the host name is available but cannot be verified. # # Comment out the following definition if you want more control over such # requests. When paranoid mode is off and a host name double check fails, # the client can be matched with the PARANOID access control pattern. # # Paranoid mode implies hostname lookup. In order to disable hostname # lookups altogether, see the next section. PARANOID= -DPARANOID ######################################## # Optional: turning off hostname lookups # # By default, the software always attempts to look up the client # hostname. With selective hostname lookups, the client hostname # lookup is postponed until the name is required by an access control # rule or by a %letter expansion. # # In order to perform selective hostname lookups, disable paranoid # mode (see previous section) and comment out the following definition. HOSTNAME= -DALWAYS_HOSTNAME ############################################# # Optional: Turning on host ADDRESS checking # # Optionally, the software tries to protect against hosts that pretend to # have someone elses host address. This is relevant for network services # whose authentication depends on host names, such as rsh and rlogin, # because the network address is used to look up the remote host name. # # The protection is to refuse TCP connections with IP source routing # options. # # This feature cannot be used with SunOS 4.x because of a kernel bug in # the implementation of the getsockopt() system call. Kernel panics have # been observed for SunOS 4.1.[1-3]. Symptoms are "BAD TRAP" and "Data # fault" while executing the tcp_ctloutput() kernel function. # # Reportedly, Sun patch 100804-03 or 101790 fixes this for SunOS 4.1.x. # # Uncomment the following macro definition if your getsockopt() is OK. # # -DKILL_IP_OPTIONS is not needed on modern UNIX systems that can stop # source-routed traffic in the kernel. Examples: 4.4BSD derivatives, # Solaris 2.x, and Linux. See your system documentation for details. # # KILL_OPT= -DKILL_IP_OPTIONS ## End configuration options ############################ # Protection against weird shells or weird make programs. SHELL = /bin/sh .c.o:; $(CC) $(CFLAGS) -c $*.c CFLAGS = -O -DFACILITY=$(FACILITY) $(ACCESS) $(PARANOID) $(NETGROUP) \ $(BUGS) $(SYSTYPE) $(AUTH) $(UMASK) \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" $(STYLE) $(KILL_OPT) \ -DSEVERITY=$(SEVERITY) -DRFC931_TIMEOUT=$(RFC931_TIMEOUT) \ $(UCHAR) $(TABLES) $(STRINGS) $(TLI) $(EXTRA_CFLAGS) $(DOT) \ $(VSYSLOG) $(HOSTNAME) LIB_OBJ= hosts_access.o options.o shell_cmd.o rfc931.o eval.o \ hosts_ctl.o refuse.o percent_x.o clean_exit.o $(AUX_OBJ) \ $(FROM_OBJ) fix_options.o socket.o tli.o workarounds.o \ update.o misc.o diag.o percent_m.o myvsyslog.o FROM_OBJ= fromhost.o KIT = README miscd.c tcpd.c fromhost.c hosts_access.c shell_cmd.c \ tcpd.h tcpdmatch.c Makefile hosts_access.5 strcasecmp.c BLURB rfc931.c \ tcpd.8 eval.c hosts_access.3 hosts_ctl.c percent_x.c options.c \ clean_exit.c environ.c patchlevel.h fix_options.c workarounds.c \ socket.c tli.c DISCLAIMER fakelog.c safe_finger.c hosts_options.5 \ CHANGES try-from.c update.c ptx.c vfprintf.c tli-sequent.c \ tli-sequent.h misc.c diag.c ncr.c tcpdchk.c percent_m.c \ myvsyslog.c mystdarg.h printf.ck README.IRIX Banners.Makefile \ refuse.c tcpdchk.8 setenv.c inetcf.c inetcf.h scaffold.c \ scaffold.h tcpdmatch.8 README.NIS LIB = libwrap.a all other: config-check tcpd tcpdmatch try-from safe_finger tcpdchk # Invalidate all object files when the compiler options (CFLAGS) have changed. config-check: @set +e; test -n "$(REAL_DAEMON_DIR)" || { make; exit 1; } @set +e; echo $(CFLAGS) >/tmp/cflags.$$$$ ; \ if cmp cflags /tmp/cflags.$$$$ ; \ then rm /tmp/cflags.$$$$ ; \ else mv /tmp/cflags.$$$$ cflags ; \ fi >/dev/null 2>/dev/null $(LIB): $(LIB_OBJ) rm -f $(LIB) $(AR) $(ARFLAGS) $(LIB) $(LIB_OBJ) -$(RANLIB) $(LIB) tcpd: tcpd.o $(LIB) $(CC) $(CFLAGS) -o $@ tcpd.o $(LIB) $(LIBS) miscd: miscd.o $(LIB) $(CC) $(CFLAGS) -o $@ miscd.o $(LIB) $(LIBS) safe_finger: safe_finger.o $(LIB) $(CC) $(CFLAGS) -o $@ safe_finger.o $(LIB) $(LIBS) TCPDMATCH_OBJ = tcpdmatch.o fakelog.o inetcf.o scaffold.o tcpdmatch: $(TCPDMATCH_OBJ) $(LIB) $(CC) $(CFLAGS) -o $@ $(TCPDMATCH_OBJ) $(LIB) $(LIBS) try-from: try-from.o fakelog.o $(LIB) $(CC) $(CFLAGS) -o $@ try-from.o fakelog.o $(LIB) $(LIBS) TCPDCHK_OBJ = tcpdchk.o fakelog.o inetcf.o scaffold.o tcpdchk: $(TCPDCHK_OBJ) $(LIB) $(CC) $(CFLAGS) -o $@ $(TCPDCHK_OBJ) $(LIB) $(LIBS) shar: $(KIT) @shar $(KIT) kit: $(KIT) @makekit $(KIT) files: @echo $(KIT) archive: $(ARCHIVE) $(KIT) clean: rm -f tcpd miscd safe_finger tcpdmatch tcpdchk try-from *.[oa] core \ cflags tidy: clean chmod -R a+r . chmod 755 . # Enable all bells and whistles for linting. lint: tcpd_lint miscd_lint match_lint chk_lint tcpd_lint: lint -DFACILITY=LOG_MAIL -DHOSTS_ACCESS -DPARANOID -DNETGROUP \ -DGETPEERNAME_BUG -DDAEMON_UMASK=022 -DSEVERITY=$(SEVERITY) \ $(TABLES) -DKILL_IP_OPTIONS -DPROCESS_OPTIONS \ -DRFC931_TIMEOUT=$(RFC931_TIMEOUT) -DALWAYS_RFC931 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" \ -Dvsyslog=myvsyslog \ tcpd.c fromhost.c socket.c tli.c hosts_access.c \ shell_cmd.c refuse.c rfc931.c eval.c percent_x.c clean_exit.c \ options.c setenv.c fix_options.c workarounds.c update.c misc.c \ diag.c myvsyslog.c percent_m.c miscd_lint: lint -DFACILITY=LOG_MAIL -DHOSTS_ACCESS -DPARANOID -DNETGROUP \ -DGETPEERNAME_BUG -DDAEMON_UMASK=022 -DSEVERITY=$(SEVERITY) \ $(TABLES) -DKILL_IP_OPTIONS -DPROCESS_OPTIONS \ -DRFC931_TIMEOUT=$(RFC931_TIMEOUT) -DALWAYS_RFC931 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" \ -Dvsyslog=myvsyslog \ miscd.c fromhost.c socket.c tli.c hosts_access.c \ shell_cmd.c refuse.c rfc931.c eval.c percent_x.c clean_exit.c \ options.c setenv.c fix_options.c workarounds.c update.c misc.c \ diag.c myvsyslog.c percent_m.c match_lint: lint -DFACILITY=LOG_MAIL -DSEVERITY=$(SEVERITY) -DHOSTS_ACCESS \ -DPARANOID $(TABLES) -DNETGROUP -DPROCESS_OPTIONS -DRFC931_TIMEOUT=10 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" \ -Dvsyslog=myvsyslog \ tcpdmatch.c hosts_access.c eval.c percent_x.c options.c workarounds.c \ update.c socket.c misc.c diag.c myvsyslog.c percent_m.c setenv.c \ inetcf.c scaffold.c chk_lint: lint -DFACILITY=LOG_MAIL -DSEVERITY=$(SEVERITY) -DHOSTS_ACCESS \ -DPARANOID $(TABLES) -DNETGROUP -DPROCESS_OPTIONS -DRFC931_TIMEOUT=10 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" \ -Dvsyslog=myvsyslog \ tcpdchk.c eval.c percent_x.c options.c update.c workarounds.c \ setenv.c misc.c diag.c myvsyslog.c percent_m.c inetcf.c scaffold.c printfck: printfck -f printf.ck \ tcpd.c fromhost.c socket.c tli.c hosts_access.c \ shell_cmd.c refuse.c rfc931.c eval.c percent_x.c clean_exit.c \ options.c setenv.c fix_options.c workarounds.c update.c misc.c \ diag.c myvsyslog.c percent_m.c >aap.c lint -DFACILITY=LOG_MAIL -DHOSTS_ACCESS -DPARANOID -DNETGROUP \ -DGETPEERNAME_BUG -DDAEMON_UMASK=022 -DSEVERITY=$(SEVERITY) \ $(TABLES) -DKILL_IP_OPTIONS -DPROCESS_OPTIONS \ -DRFC931_TIMEOUT=$(RFC931_TIMEOUT) -DALWAYS_RFC931 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" -Dvsyslog=myvsyslog aap.c printfck -f printf.ck \ tcpdchk.c eval.c percent_x.c options.c update.c workarounds.c \ setenv.c misc.c diag.c myvsyslog.c percent_m.c inetcf.c scaffold.c \ >aap.c lint -DFACILITY=LOG_MAIL -DSEVERITY=$(SEVERITY) -DHOSTS_ACCESS \ -DPARANOID $(TABLES) -DNETGROUP -DPROCESS_OPTIONS -DRFC931_TIMEOUT=10 \ -Dvsyslog=myvsyslog -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" # Internal compilation dependencies. clean_exit.o: cflags clean_exit.o: tcpd.h diag.o: cflags diag.o: mystdarg.h diag.o: tcpd.h environ.o: cflags eval.o: cflags eval.o: tcpd.h fakelog.o: cflags fakelog.o: mystdarg.h fix_options.o: cflags fix_options.o: tcpd.h fromhost.o: cflags fromhost.o: tcpd.h hosts_access.o: cflags hosts_access.o: tcpd.h hosts_ctl.o: cflags hosts_ctl.o: tcpd.h inetcf.o: cflags inetcf.o: inetcf.h inetcf.o: tcpd.h misc.o: cflags misc.o: tcpd.h miscd.o: cflags miscd.o: patchlevel.h miscd.o: tcpd.h myvsyslog.o: cflags myvsyslog.o: mystdarg.h myvsyslog.o: tcpd.h ncr.o: cflags ncr.o: tcpd.h options.o: cflags options.o: tcpd.h percent_m.o: cflags percent_m.o: mystdarg.h percent_x.o: cflags percent_x.o: tcpd.h ptx.o: cflags ptx.o: tcpd.h refuse.o: cflags refuse.o: tcpd.h rfc931.o: cflags rfc931.o: tcpd.h safe_finger.o: cflags scaffold.o: cflags scaffold.o: scaffold.h scaffold.o: tcpd.h setenv.o: cflags shell_cmd.o: cflags shell_cmd.o: tcpd.h socket.o: cflags socket.o: tcpd.h strcasecmp.o: cflags tcpd.o: cflags tcpd.o: patchlevel.h tcpd.o: tcpd.h tcpdchk.o: cflags tcpdchk.o: inetcf.h tcpdchk.o: scaffold.h tcpdchk.o: tcpd.h tcpdmatch.o: cflags tcpdmatch.o: scaffold.h tcpdmatch.o: tcpd.h tli-sequent.o: cflags tli-sequent.o: tcpd.h tli-sequent.o: tli-sequent.h tli.o: cflags tli.o: tcpd.h try-from.o: cflags try-from.o: tcpd.h update.o: cflags update.o: mystdarg.h update.o: tcpd.h vfprintf.o: cflags workarounds.o: cflags workarounds.o: tcpd.h RNAME_BUG -DDAEMON_UMASK=022 -DSEVERITY=$(SEVERITY) \ $(TABLES) -DKILL_IP_OPTIONS -DPROCESS_OPTIONS \ -DRFC931_TIMEOUT=$(RFC931_TIMEOUT) -DALWAYS_RFC931 \ -DREAL_DAEMON_DIR=\"$(REAL_DAEMON_DIR)\" \ -Dvsyslog=myvsyslog \ tcpd.c fromhost.c socket.c tli.c hosts_access.c \ shell_cmd.c refuse.c rfc931.tcp_wrappers_7.6/hosts_access.5 444 465 506 35571 5713232503 12042 .TH HOSTS_ACCESS 5 .SH NAME hosts_access \- format of host access control files .SH DESCRIPTION This manual page describes a simple access control language that is based on client (host name/address, user name), and server (process name, host name/address) patterns. Examples are given at the end. The impatient reader is encouraged to skip to the EXAMPLES section for a quick introduction. .PP An extended version of the access control language is described in the \fIhosts_options\fR(5) document. The extensions are turned on at program build time by building with -DPROCESS_OPTIONS. .PP In the following text, \fIdaemon\fR is the the process name of a network daemon process, and \fIclient\fR is the name and/or address of a host requesting service. Network daemon process names are specified in the inetd configuration file. .SH ACCESS CONTROL FILES The access control software consults two files. The search stops at the first match: .IP \(bu Access will be granted when a (daemon,client) pair matches an entry in the \fI/etc/hosts.allow\fR file. .IP \(bu Otherwise, access will be denied when a (daemon,client) pair matches an entry in the \fI/etc/hosts.deny\fR file. .IP \(bu Otherwise, access will be granted. .PP A non-existing access control file is treated as if it were an empty file. Thus, access control can be turned off by providing no access control files. .SH ACCESS CONTROL RULES Each access control file consists of zero or more lines of text. These lines are processed in order of appearance. The search terminates when a match is found. .IP \(bu A newline character is ignored when it is preceded by a backslash character. This permits you to break up long lines so that they are easier to edit. .IP \(bu Blank lines or lines that begin with a `#\' character are ignored. This permits you to insert comments and whitespace so that the tables are easier to read. .IP \(bu All other lines should satisfy the following format, things between [] being optional: .sp .ti +3 daemon_list : client_list [ : shell_command ] .PP \fIdaemon_list\fR is a list of one or more daemon process names (argv[0] values) or wildcards (see below). .PP \fIclient_list\fR is a list of one or more host names, host addresses, patterns or wildcards (see below) that will be matched against the client host name or address. .PP The more complex forms \fIdaemon@host\fR and \fIuser@host\fR are explained in the sections on server endpoint patterns and on client username lookups, respectively. .PP List elements should be separated by blanks and/or commas. .PP With the exception of NIS (YP) netgroup lookups, all access control checks are case insensitive. .ne 4 .SH PATTERNS The access control language implements the following patterns: .IP \(bu A string that begins with a `.\' character. A host name is matched if the last components of its name match the specified pattern. For example, the pattern `.tue.nl\' matches the host name `wzv.win.tue.nl\'. .IP \(bu A string that ends with a `.\' character. A host address is matched if its first numeric fields match the given string. For example, the pattern `131.155.\' matches the address of (almost) every host on the Eind\%hoven University network (131.155.x.x). .IP \(bu A string that begins with an `@\' character is treated as an NIS (formerly YP) netgroup name. A host name is matched if it is a host member of the specified netgroup. Netgroup matches are not supported for daemon process names or for client user names. .IP \(bu An expression of the form `n.n.n.n/m.m.m.m\' is interpreted as a `net/mask\' pair. A host address is matched if `net\' is equal to the bitwise AND of the address and the `mask\'. For example, the net/mask pattern `131.155.72.0/255.255.254.0\' matches every address in the range `131.155.72.0\' through `131.155.73.255\'. .SH WILDCARDS The access control language supports explicit wildcards: .IP ALL The universal wildcard, always matches. .IP LOCAL Matches any host whose name does not contain a dot character. .IP UNKNOWN Matches any user whose name is unknown, and matches any host whose name \fIor\fR address are unknown. This pattern should be used with care: host names may be unavailable due to temporary name server problems. A network address will be unavailable when the software cannot figure out what type of network it is talking to. .IP KNOWN Matches any user whose name is known, and matches any host whose name \fIand\fR address are known. This pattern should be used with care: host names may be unavailable due to temporary name server problems. A network address will be unavailable when the software cannot figure out what type of network it is talking to. .IP PARANOID Matches any host whose name does not match its address. When tcpd is built with -DPARANOID (default mode), it drops requests from such clients even before looking at the access control tables. Build without -DPARANOID when you want more control over such requests. .ne 6 .SH OPERATORS .IP EXCEPT Intended use is of the form: `list_1 EXCEPT list_2\'; this construct matches anything that matches \fIlist_1\fR unless it matches \fIlist_2\fR. The EXCEPT operator can be used in daemon_lists and in client_lists. The EXCEPT operator can be nested: if the control language would permit the use of parentheses, `a EXCEPT b EXCEPT c\' would parse as `(a EXCEPT (b EXCEPT c))\'. .br .ne 6 .SH SHELL COMMANDS If the first-matched access control rule contains a shell command, that command is subjected to % substitutions (see next section). The result is executed by a \fI/bin/sh\fR child process with standard input, output and error connected to \fI/dev/null\fR. Specify an `&\' at the end of the command if you do not want to wait until it has completed. .PP Shell commands should not rely on the PATH setting of the inetd. Instead, they should use absolute path names, or they should begin with an explicit PATH=whatever statement. .PP The \fIhosts_options\fR(5) document describes an alternative language that uses the shell command field in a different and incompatible way. .SH % EXPANSIONS The following expansions are available within shell commands: .IP "%a (%A)" The client (server) host address. .IP %c Client information: user@host, user@address, a host name, or just an address, depending on how much information is available. .IP %d The daemon process name (argv[0] value). .IP "%h (%H)" The client (server) host name or address, if the host name is unavailable. .IP "%n (%N)" The client (server) host name (or "unknown" or "paranoid"). .IP %p The daemon process id. .IP %s Server information: daemon@host, daemon@address, or just a daemon name, depending on how much information is available. .IP %u The client user name (or "unknown"). .IP %% Expands to a single `%\' character. .PP Characters in % expansions that may confuse the shell are replaced by underscores. .SH SERVER ENDPOINT PATTERNS In order to distinguish clients by the network address that they connect to, use patterns of the form: .sp .ti +3 process_name@host_pattern : client_list ... .sp Patterns like these can be used when the machine has different internet addresses with different internet hostnames. Service providers can use this facility to offer FTP, GOPHER or WWW archives with internet names that may even belong to different organizations. See also the `twist' option in the hosts_options(5) document. Some systems (Solaris, FreeBSD) can have more than one internet address on one physical interface; with other systems you may have to resort to SLIP or PPP pseudo interfaces that live in a dedicated network address space. .sp The host_pattern obeys the same syntax rules as host names and addresses in client_list context. Usually, server endpoint information is available only with connection-oriented services. .SH CLIENT USERNAME LOOKUP When the client host supports the RFC 931 protocol or one of its descendants (TAP, IDENT, RFC 1413) the wrapper programs can retrieve additional information about the owner of a connection. Client username information, when available, is logged together with the client host name, and can be used to match patterns like: .PP .ti +3 daemon_list : ... user_pattern@host_pattern ... .PP The daemon wrappers can be configured at compile time to perform rule-driven username lookups (default) or to always interrogate the client host. In the case of rule-driven username lookups, the above rule would cause username lookup only when both the \fIdaemon_list\fR and the \fIhost_pattern\fR match. .PP A user pattern has the same syntax as a daemon process pattern, so the same wildcards apply (netgroup membership is not supported). One should not get carried away with username lookups, though. .IP \(bu The client username information cannot be trusted when it is needed most, i.e. when the client system has been compromised. In general, ALL and (UN)KNOWN are the only user name patterns that make sense. .IP \(bu Username lookups are possible only with TCP-based services, and only when the client host runs a suitable daemon; in all other cases the result is "unknown". .IP \(bu A well-known UNIX kernel bug may cause loss of service when username lookups are blocked by a firewall. The wrapper README document describes a procedure to find out if your kernel has this bug. .IP \(bu Username lookups may cause noticeable delays for non-UNIX users. The default timeout for username lookups is 10 seconds: too short to cope with slow networks, but long enough to irritate PC users. .PP Selective username lookups can alleviate the last problem. For example, a rule like: .PP .ti +3 daemon_list : @pcnetgroup ALL@ALL .PP would match members of the pc netgroup without doing username lookups, but would perform username lookups with all other systems. .SH DETECTING ADDRESS SPOOFING ATTACKS A flaw in the sequence number generator of many TCP/IP implementations allows intruders to easily impersonate trusted hosts and to break in via, for example, the remote shell service. The IDENT (RFC931 etc.) service can be used to detect such and other host address spoofing attacks. .PP Before accepting a client request, the wrappers can use the IDENT service to find out that the client did not send the request at all. When the client host provides IDENT service, a negative IDENT lookup result (the client matches `UNKNOWN@host') is strong evidence of a host spoofing attack. .PP A positive IDENT lookup result (the client matches `KNOWN@host') is less trustworthy. It is possible for an intruder to spoof both the client connection and the IDENT lookup, although doing so is much harder than spoofing just a client connection. It may also be that the client\'s IDENT server is lying. .PP Note: IDENT lookups don\'t work with UDP services. .SH EXAMPLES The language is flexible enough that different types of access control policy can be expressed with a minimum of fuss. Although the language uses two access control tables, the most common policies can be implemented with one of the tables being trivial or even empty. .PP When reading the examples below it is important to realize that the allow table is scanned before the deny table, that the search terminates when a match is found, and that access is granted when no match is found at all. .PP The examples use host and domain names. They can be improved by including address and/or network/netmask information, to reduce the impact of temporary name server lookup failures. .SH MOSTLY CLOSED In this case, access is denied by default. Only explicitly authorized hosts are permitted access. .PP The default policy (no access) is implemented with a trivial deny file: .PP .ne 2 /etc/hosts.deny: .in +3 ALL: ALL .PP This denies all service to all hosts, unless they are permitted access by entries in the allow file. .PP The explicitly authorized hosts are listed in the allow file. For example: .PP .ne 2 /etc/hosts.allow: .in +3 ALL: LOCAL @some_netgroup .br ALL: .foobar.edu EXCEPT terminalserver.foobar.edu .PP The first rule permits access from hosts in the local domain (no `.\' in the host name) and from members of the \fIsome_netgroup\fP netgroup. The second rule permits access from all hosts in the \fIfoobar.edu\fP domain (notice the leading dot), with the exception of \fIterminalserver.foobar.edu\fP. .SH MOSTLY OPEN Here, access is granted by default; only explicitly specified hosts are refused service. .PP The default policy (access granted) makes the allow file redundant so that it can be omitted. The explicitly non-authorized hosts are listed in the deny file. For example: .PP /etc/hosts.deny: .in +3 ALL: some.host.name, .some.domain .br ALL EXCEPT in.fingerd: other.host.name, .other.domain .PP The first rule denies some hosts and domains all services; the second rule still permits finger requests from other hosts and domains. .SH BOOBY TRAPS The next example permits tftp requests from hosts in the local domain (notice the leading dot). Requests from any other hosts are denied. Instead of the requested file, a finger probe is sent to the offending host. The result is mailed to the superuser. .PP .ne 2 /etc/hosts.allow: .in +3 .nf in.tftpd: LOCAL, .my.domain .PP .ne 2 /etc/hosts.deny: .in +3 .nf in.tftpd: ALL: (/some/where/safe_finger -l @%h | \\ /usr/ucb/mail -s %d-%h root) & .fi .PP The safe_finger command comes with the tcpd wrapper and should be installed in a suitable place. It limits possible damage from data sent by the remote finger server. It gives better protection than the standard finger command. .PP The expansion of the %h (client host) and %d (service name) sequences is described in the section on shell commands. .PP Warning: do not booby-trap your finger daemon, unless you are prepared for infinite finger loops. .PP On network firewall systems this trick can be carried even further. The typical network firewall only provides a limited set of services to the outer world. All other services can be "bugged" just like the above tftp example. The result is an excellent early-warning system. .br .ne 4 .SH DIAGNOSTICS An error is reported when a syntax error is found in a host access control rule; when the length of an access control rule exceeds the capacity of an internal buffer; when an access control rule is not terminated by a newline character; when the result of % expansion would overflow an internal buffer; when a system call fails that shouldn\'t. All problems are reported via the syslog daemon. .SH FILES .na .nf /etc/hosts.allow, (daemon,client) pairs that are granted access. /etc/hosts.deny, (daemon,client) pairs that are denied access. .ad .fi .SH SEE ALSO .nf tcpd(8) tcp/ip daemon wrapper program. tcpdchk(8), tcpdmatch(8), test programs. .SH BUGS If a name server lookup times out, the host name will not be available to the access control software, even though the host is registered. .PP Domain name server lookups are case insensitive; NIS (formerly YP) netgroup lookups are case sensitive. .SH AUTHOR .na .nf Wietse Venema (wietse@wzv.win.tue.nl) Department of Mathematics and Computing Science Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) hosts_access.5 1.20 95/01/30 19:51:46 is scanned before the deny table, that the search terminates when a match is found, and that access is granted when no match is found atcp_wrappers_7.6/strcasecmp.c 444 465 0 7433 5444416616 11516 /* * Copyright (c) 1987 Regents of the University of California. * All rights reserved. * * Redistribution and use in source and binary forms are permitted * provided that the above copyright notice and this paragraph are * duplicated in all such forms and that any documentation, * advertising materials, and other materials related to such * distribution and use acknowledge that the software was developed * by the University of California, Berkeley. The name of the * University may not be used to endorse or promote products derived * from this software without specific prior written permission. * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED * WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE. */ #if defined(LIBC_SCCS) && !defined(lint) static char sccsid[] = "@(#)strcasecmp.c 5.6 (Berkeley) 6/27/88"; #endif /* LIBC_SCCS and not lint */ /* Some environments don't define u_char -- WZV */ #if 0 #include #else typedef unsigned char u_char; #endif /* * This array is designed for mapping upper and lower case letter * together for a case independent comparison. The mappings are * based upon ascii character sequences. */ static u_char charmap[] = { '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007', '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017', '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027', '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037', '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047', '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057', '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067', '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077', '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147', '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157', '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167', '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137', '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147', '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157', '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167', '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177', '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207', '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217', '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227', '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237', '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247', '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257', '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267', '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277', '\300', '\341', '\342', '\343', '\344', '\345', '\346', '\347', '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357', '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367', '\370', '\371', '\372', '\333', '\334', '\335', '\336', '\337', '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347', '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357', '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367', '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377', }; strcasecmp(s1, s2) char *s1, *s2; { register u_char *cm = charmap, *us1 = (u_char *)s1, *us2 = (u_char *)s2; while (cm[*us1] == cm[*us2++]) if (*us1++ == '\0') return(0); return(cm[*us1] - cm[*--us2]); } strncasecmp(s1, s2, n) char *s1, *s2; register int n; { register u_char *cm = charmap, *us1 = (u_char *)s1, *us2 = (u_char *)s2; while (--n >= 0 && cm[*us1] == cm[*us2++]) if (*us1++ == '\0') return(0); return(n < 0 ? 0 : cm[*us1] - cm[*--us2]); } ts are permitted access. .PP The default policy (no access) is implemented with a trivial deny file: .PP .ne 2 /etc/hosts.deny: .in +3 ALL: ALL .PP This denies all service to all hosts, unless they are permitted access by entritcp_wrappers_7.6/BLURB 444 465 506 3310 6314551407 10033 @(#) BLURB 1.28 97/03/21 19:27:18 With this package you can monitor and filter incoming requests for the SYSTAT, FINGER, FTP, TELNET, RLOGIN, RSH, EXEC, TFTP, TALK, and other network services. The package provides tiny daemon wrapper programs that can be installed without any changes to existing software or to existing configuration files. The wrappers report the name of the client host and of the requested service; the wrappers do not exchange information with the client or server applications, and impose no overhead on the actual conversation between the client and server applications. This patch upgrades the tcp wrappers version 7.5 source code to version 7.6. The source-routing protection in version 7.5 was not as strong as it could be. And all this effort was not needed with modern UNIX systems that can already stop source-routed traffic in the kernel. Examples are 4.4BSD derivatives, Solaris 2.x, and Linux. This release does not introduce new features. Do not bother applying this patch when you built your version 7.x tcp wrapper without enabling the KILL_IP_OPTIONS compiler switch; when you can disable IP source routing options in the kernel; when you run a UNIX version that pre-dates 4.4BSD, such as SunOS 4. Such systems are unable to receive source-routed connections and are therefore not vulnerable to IP spoofing attacks with source-routed TCP connections. A complete change log is given in the CHANGES document. As always, problem reports and suggestions for improvement are welcome. Wietse Venema (wietse@wzv.win.tue.nl), Department of Mathematics and Computing Science, Eindhoven University of Technology, The Netherlands. Currently visiting IBM T.J. Watson Research, Hawthorne NY, USA. rsion 7.5 was not as strong as it could be. And all this effort was not needed with modern UNIX systems that can already stop source-routed traffic in the kernel. Examples are 4.4BSD derivatives, Solaris 2.x, and Linux. This release does not introduce new features. Do not bother applying this patch when you butcp_wrappers_7.6/rfc931.c 444 465 506 10546 5702013646 10444 /* * rfc931() speaks a common subset of the RFC 931, AUTH, TAP, IDENT and RFC * 1413 protocols. It queries an RFC 931 etc. compatible daemon on a remote * host to look up the owner of a connection. The information should not be * used for authentication purposes. This routine intercepts alarm signals. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) rfc931.c 1.10 95/01/02 16:11:34"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include /* Local stuff. */ #include "tcpd.h" #define RFC931_PORT 113 /* Semi-well-known port */ #define ANY_PORT 0 /* Any old port will do */ int rfc931_timeout = RFC931_TIMEOUT;/* Global so it can be changed */ static jmp_buf timebuf; /* fsocket - open stdio stream on top of socket */ static FILE *fsocket(domain, type, protocol) int domain; int type; int protocol; { int s; FILE *fp; if ((s = socket(domain, type, protocol)) < 0) { tcpd_warn("socket: %m"); return (0); } else { if ((fp = fdopen(s, "r+")) == 0) { tcpd_warn("fdopen: %m"); close(s); } return (fp); } } /* timeout - handle timeouts */ static void timeout(sig) int sig; { longjmp(timebuf, sig); } /* rfc931 - return remote user name, given socket structures */ void rfc931(rmt_sin, our_sin, dest) struct sockaddr_in *rmt_sin; struct sockaddr_in *our_sin; char *dest; { unsigned rmt_port; unsigned our_port; struct sockaddr_in rmt_query_sin; struct sockaddr_in our_query_sin; char user[256]; /* XXX */ char buffer[512]; /* XXX */ char *cp; char *result = unknown; FILE *fp; /* * Use one unbuffered stdio stream for writing to and for reading from * the RFC931 etc. server. This is done because of a bug in the SunOS * 4.1.x stdio library. The bug may live in other stdio implementations, * too. When we use a single, buffered, bidirectional stdio stream ("r+" * or "w+" mode) we read our own output. Such behaviour would make sense * with resources that support random-access operations, but not with * sockets. */ if ((fp = fsocket(AF_INET, SOCK_STREAM, 0)) != 0) { setbuf(fp, (char *) 0); /* * Set up a timer so we won't get stuck while waiting for the server. */ if (setjmp(timebuf) == 0) { signal(SIGALRM, timeout); alarm(rfc931_timeout); /* * Bind the local and remote ends of the query socket to the same * IP addresses as the connection under investigation. We go * through all this trouble because the local or remote system * might have more than one network address. The RFC931 etc. * client sends only port numbers; the server takes the IP * addresses from the query socket. */ our_query_sin = *our_sin; our_query_sin.sin_port = htons(ANY_PORT); rmt_query_sin = *rmt_sin; rmt_query_sin.sin_port = htons(RFC931_PORT); if (bind(fileno(fp), (struct sockaddr *) & our_query_sin, sizeof(our_query_sin)) >= 0 && connect(fileno(fp), (struct sockaddr *) & rmt_query_sin, sizeof(rmt_query_sin)) >= 0) { /* * Send query to server. Neglect the risk that a 13-byte * write would have to be fragmented by the local system and * cause trouble with buggy System V stdio libraries. */ fprintf(fp, "%u,%u\r\n", ntohs(rmt_sin->sin_port), ntohs(our_sin->sin_port)); fflush(fp); /* * Read response from server. Use fgets()/sscanf() so we can * work around System V stdio libraries that incorrectly * assume EOF when a read from a socket returns less than * requested. */ if (fgets(buffer, sizeof(buffer), fp) != 0 && ferror(fp) == 0 && feof(fp) == 0 && sscanf(buffer, "%u , %u : USERID :%*[^:]:%255s", &rmt_port, &our_port, user) == 3 && ntohs(rmt_sin->sin_port) == rmt_port && ntohs(our_sin->sin_port) == our_port) { /* * Strip trailing carriage return. It is part of the * protocol, not part of the data. */ if (cp = strchr(user, '\r')) *cp = 0; result = user; } } alarm(0); } fclose(fp); } STRN_CPY(dest, result, STRING_LENGTH); } *fp; /* * Use one unbuffered stdio stream for writing to and for reading from * the RFC931 etc. server. This is done because of a bug in tcp_wrappers_7.6/tcpd.8 444 465 506 15556 6112636244 10323 .TH TCPD 8 .SH NAME tcpd \- access control facility for internet services .SH DESCRIPTION .PP The \fItcpd\fR program can be set up to monitor incoming requests for \fItelnet\fR, \fIfinger\fR, \fIftp\fR, \fIexec\fR, \fIrsh\fR, \fIrlogin\fR, \fItftp\fR, \fItalk\fR, \fIcomsat\fR and other services that have a one-to-one mapping onto executable files. .PP The program supports both 4.3BSD-style sockets and System V.4-style TLI. Functionality may be limited when the protocol underneath TLI is not an internet protocol. .PP Operation is as follows: whenever a request for service arrives, the \fIinetd\fP daemon is tricked into running the \fItcpd\fP program instead of the desired server. \fItcpd\fP logs the request and does some additional checks. When all is well, \fItcpd\fP runs the appropriate server program and goes away. .PP Optional features are: pattern-based access control, client username lookups with the RFC 931 etc. protocol, protection against hosts that pretend to have someone elses host name, and protection against hosts that pretend to have someone elses network address. .SH LOGGING Connections that are monitored by .I tcpd are reported through the \fIsyslog\fR(3) facility. Each record contains a time stamp, the client host name and the name of the requested service. The information can be useful to detect unwanted activities, especially when logfile information from several hosts is merged. .PP In order to find out where your logs are going, examine the syslog configuration file, usually /etc/syslog.conf. .SH ACCESS CONTROL Optionally, .I tcpd supports a simple form of access control that is based on pattern matching. The access-control software provides hooks for the execution of shell commands when a pattern fires. For details, see the \fIhosts_access\fR(5) manual page. .SH HOST NAME VERIFICATION The authentication scheme of some protocols (\fIrlogin, rsh\fR) relies on host names. Some implementations believe the host name that they get from any random name server; other implementations are more careful but use a flawed algorithm. .PP .I tcpd verifies the client host name that is returned by the address->name DNS server by looking at the host name and address that are returned by the name->address DNS server. If any discrepancy is detected, .I tcpd concludes that it is dealing with a host that pretends to have someone elses host name. .PP If the sources are compiled with -DPARANOID, .I tcpd will drop the connection in case of a host name/address mismatch. Otherwise, the hostname can be matched with the \fIPARANOID\fR wildcard, after which suitable action can be taken. .SH HOST ADDRESS SPOOFING Optionally, .I tcpd disables source-routing socket options on every connection that it deals with. This will take care of most attacks from hosts that pretend to have an address that belongs to someone elses network. UDP services do not benefit from this protection. This feature must be turned on at compile time. .SH RFC 931 When RFC 931 etc. lookups are enabled (compile-time option) \fItcpd\fR will attempt to establish the name of the client user. This will succeed only if the client host runs an RFC 931-compliant daemon. Client user name lookups will not work for datagram-oriented connections, and may cause noticeable delays in the case of connections from PCs. .SH EXAMPLES The details of using \fItcpd\fR depend on pathname information that was compiled into the program. .SH EXAMPLE 1 This example applies when \fItcpd\fR expects that the original network daemons will be moved to an "other" place. .PP In order to monitor access to the \fIfinger\fR service, move the original finger daemon to the "other" place and install tcpd in the place of the original finger daemon. No changes are required to configuration files. .nf .sp .in +5 # mkdir /other/place # mv /usr/etc/in.fingerd /other/place # cp tcpd /usr/etc/in.fingerd .fi .PP The example assumes that the network daemons live in /usr/etc. On some systems, network daemons live in /usr/sbin or in /usr/libexec, or have no `in.\' prefix to their name. .SH EXAMPLE 2 This example applies when \fItcpd\fR expects that the network daemons are left in their original place. .PP In order to monitor access to the \fIfinger\fR service, perform the following edits on the \fIinetd\fR configuration file (usually \fI/etc/inetd.conf\fR or \fI/etc/inet/inetd.conf\fR): .nf .sp .ti +5 finger stream tcp nowait nobody /usr/etc/in.fingerd in.fingerd .sp becomes: .sp .ti +5 finger stream tcp nowait nobody /some/where/tcpd in.fingerd .sp .fi .PP The example assumes that the network daemons live in /usr/etc. On some systems, network daemons live in /usr/sbin or in /usr/libexec, the daemons have no `in.\' prefix to their name, or there is no userid field in the inetd configuration file. .PP Similar changes will be needed for the other services that are to be covered by \fItcpd\fR. Send a `kill -HUP\' to the \fIinetd\fR(8) process to make the changes effective. AIX users may also have to execute the `inetimp\' command. .SH EXAMPLE 3 In the case of daemons that do not live in a common directory ("secret" or otherwise), edit the \fIinetd\fR configuration file so that it specifies an absolute path name for the process name field. For example: .nf .sp ntalk dgram udp wait root /some/where/tcpd /usr/local/lib/ntalkd .sp .fi .PP Only the last component (ntalkd) of the pathname will be used for access control and logging. .SH BUGS Some UDP (and RPC) daemons linger around for a while after they have finished their work, in case another request comes in. In the inetd configuration file these services are registered with the \fIwait\fR option. Only the request that started such a daemon will be logged. .PP The program does not work with RPC services over TCP. These services are registered as \fIrpc/tcp\fR in the inetd configuration file. The only non-trivial service that is affected by this limitation is \fIrexd\fR, which is used by the \fIon(1)\fR command. This is no great loss. On most systems, \fIrexd\fR is less secure than a wildcard in /etc/hosts.equiv. .PP RPC broadcast requests (for example: \fIrwall, rup, rusers\fR) always appear to come from the responding host. What happens is that the client broadcasts the request to all \fIportmap\fR daemons on its network; each \fIportmap\fR daemon forwards the request to a local daemon. As far as the \fIrwall\fR etc. daemons know, the request comes from the local host. .SH FILES .PP The default locations of the host access control tables are: .PP /etc/hosts.allow .br /etc/hosts.deny .SH SEE ALSO .na .nf hosts_access(5), format of the tcpd access control tables. syslog.conf(5), format of the syslogd control file. inetd.conf(5), format of the inetd control file. .SH AUTHORS .na .nf Wietse Venema (wietse@wzv.win.tue.nl), Department of Mathematics and Computing Science, Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) tcpd.8 1.5 96/02/21 16:39:16 feature must be turned on at compile time. .SH RFC 931 When RFC 931 etc. lookups are enabled (compile-time option) \fItcpd\fR will attempt to esttcp_wrappers_7.6/eval.c 444 465 506 7131 5713232502 10334 /* * Routines for controlled evaluation of host names, user names, and so on. * They are, in fact, wrappers around the functions that are specific for * the sockets or TLI programming interfaces. The request_info and host_info * structures are used for result cacheing. * * These routines allows us to postpone expensive operations until their * results are really needed. Examples are hostname lookups and double * checks, or username lookups. Information that cannot be retrieved is * given the value "unknown" ("paranoid" in case of hostname problems). * * When ALWAYS_HOSTNAME is off, hostname lookup is done only when required by * tcpd paranoid mode, by access control patterns, or by %letter expansions. * * When ALWAYS_RFC931 mode is off, user lookup is done only when required by * access control patterns or %letter expansions. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) eval.c 1.3 95/01/30 19:51:45"; #endif /* System libraries. */ #include #include /* Local stuff. */ #include "tcpd.h" /* * When a string has the value STRING_UNKNOWN, it means: don't bother, I * tried to look up the data but it was unavailable for some reason. When a * host name has the value STRING_PARANOID it means there was a name/address * conflict. */ char unknown[] = STRING_UNKNOWN; char paranoid[] = STRING_PARANOID; /* eval_user - look up user name */ char *eval_user(request) struct request_info *request; { if (request->user[0] == 0) { strcpy(request->user, unknown); if (request->sink == 0 && request->client->sin && request->server->sin) rfc931(request->client->sin, request->server->sin, request->user); } return (request->user); } /* eval_hostaddr - look up printable address */ char *eval_hostaddr(host) struct host_info *host; { if (host->addr[0] == 0) { strcpy(host->addr, unknown); if (host->request->hostaddr != 0) host->request->hostaddr(host); } return (host->addr); } /* eval_hostname - look up host name */ char *eval_hostname(host) struct host_info *host; { if (host->name[0] == 0) { strcpy(host->name, unknown); if (host->request->hostname != 0) host->request->hostname(host); } return (host->name); } /* eval_hostinfo - return string with host name (preferred) or address */ char *eval_hostinfo(host) struct host_info *host; { char *hostname; #ifndef ALWAYS_HOSTNAME /* no implicit host lookups */ if (host->name[0] == 0) return (eval_hostaddr(host)); #endif hostname = eval_hostname(host); if (HOSTNAME_KNOWN(hostname)) { return (host->name); } else { return (eval_hostaddr(host)); } } /* eval_client - return string with as much about the client as we know */ char *eval_client(request) struct request_info *request; { static char both[2 * STRING_LENGTH]; char *hostinfo = eval_hostinfo(request->client); #ifndef ALWAYS_RFC931 /* no implicit user lookups */ if (request->user[0] == 0) return (hostinfo); #endif if (STR_NE(eval_user(request), unknown)) { sprintf(both, "%s@%s", request->user, hostinfo); return (both); } else { return (hostinfo); } } /* eval_server - return string with as much about the server as we know */ char *eval_server(request) struct request_info *request; { static char both[2 * STRING_LENGTH]; char *host = eval_hostinfo(request->server); char *daemon = eval_daemon(request); if (STR_NE(host, unknown)) { sprintf(both, "%s@%s", daemon, host); return (both); } else { return (daemon); } } , unknown); if (request->sink == 0 && request->client->sin && request->server->sin) rfc931(request->client->sin, request->server->sin, request->user); } return (request->user); } /* eval_hostaddr - look up printable address */ char *eval_hostaddr(host) struct host_info *host; { if (host->addr[0] == 0) { strcpy(host->addr, unknown); if (host->request->hostaddr != 0) host->request->hostaddr(hotcp_wrappers_7.6/hosts_access.3 444 465 506 7022 6107411327 12007 .TH HOSTS_ACCESS 3 .SH NAME hosts_access, hosts_ctl, request_init, request_set \- access control library .SH SYNOPSIS .nf #include "tcpd.h" extern int allow_severity; extern int deny_severity; struct request_info *request_init(request, key, value, ..., 0) struct request_info *request; struct request_info *request_set(request, key, value, ..., 0) struct request_info *request; int hosts_access(request) struct request_info *request; int hosts_ctl(daemon, client_name, client_addr, client_user) char *daemon; char *client_name; char *client_addr; char *client_user; .fi .SH DESCRIPTION The routines described in this document are part of the \fIlibwrap.a\fR library. They implement a rule-based access control language with optional shell commands that are executed when a rule fires. .PP request_init() initializes a structure with information about a client request. request_set() updates an already initialized request structure. Both functions take a variable-length list of key-value pairs and return their first argument. The argument lists are terminated with a zero key value. All string-valued arguments are copied. The expected keys (and corresponding value types) are: .IP "RQ_FILE (int)" The file descriptor associated with the request. .IP "RQ_CLIENT_NAME (char *)" The client host name. .IP "RQ_CLIENT_ADDR (char *)" A printable representation of the client network address. .IP "RQ_CLIENT_SIN (struct sockaddr_in *)" An internal representation of the client network address and port. The contents of the structure are not copied. .IP "RQ_SERVER_NAME (char *)" The hostname associated with the server endpoint address. .IP "RQ_SERVER_ADDR (char *)" A printable representation of the server endpoint address. .IP "RQ_SERVER_SIN (struct sockaddr_in *)" An internal representation of the server endpoint address and port. The contents of the structure are not copied. .IP "RQ_DAEMON (char *)" The name of the daemon process running on the server host. .IP "RQ_USER (char *)" The name of the user on whose behalf the client host makes the request. .PP hosts_access() consults the access control tables described in the \fIhosts_access(5)\fR manual page. When internal endpoint information is available, host names and client user names are looked up on demand, using the request structure as a cache. hosts_access() returns zero if access should be denied. .PP hosts_ctl() is a wrapper around the request_init() and hosts_access() routines with a perhaps more convenient interface (though it does not pass on enough information to support automated client username lookups). The client host address, client host name and username arguments should contain valid data or STRING_UNKNOWN. hosts_ctl() returns zero if access should be denied. .PP The \fIallow_severity\fR and \fIdeny_severity\fR variables determine how accepted and rejected requests may be logged. They must be provided by the caller and may be modified by rules in the access control tables. .SH DIAGNOSTICS Problems are reported via the syslog daemon. .SH SEE ALSO hosts_access(5), format of the access control tables. hosts_options(5), optional extensions to the base language. .SH FILES /etc/hosts.allow, /etc/hosts.deny, access control tables. .SH BUGS hosts_access() uses the strtok() library function. This may interfere with other code that relies on strtok(). .SH AUTHOR .na .nf Wietse Venema (wietse@wzv.win.tue.nl) Department of Mathematics and Computing Science Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) hosts_access.3 1.8 96/02/11 17:01:26 he \fIfinger\fR service, move the original finger daemon to the "other" place and install tcpd in the place of the original finger daemon. No changes are required to configuration files. .nf .sp .in +5 # mkdir /other/place # mv /usr/etc/in.fingerd /other/place # cp tcpd /usr/etc/in.fingerd .fi .PP The example assumes that the network daemons live in /usr/etc. On some systems, network daemons live in /usr/sbin or in /usr/libexec, or have no `in.\' prefix to their name. .SH EXAMPLE 2 This extcp_wrappers_7.6/hosts_ctl.c 444 465 506 2117 5700312564 11411 /* * hosts_ctl() combines common applications of the host access control * library routines. It bundles its arguments then calls the hosts_access() * access control checker. The host name and user name arguments should be * empty strings, STRING_UNKNOWN or real data. If a match is found, the * optional shell command is executed. * * Restriction: this interface does not pass enough information to support * selective remote username lookups or selective hostname double checks. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) hosts_ctl.c 1.4 94/12/28 17:42:27"; #endif #include #include "tcpd.h" /* hosts_ctl - limited interface to the hosts_access() routine */ int hosts_ctl(daemon, name, addr, user) char *daemon; char *name; char *addr; char *user; { struct request_info request; return (hosts_access(request_init(&request, RQ_DAEMON, daemon, RQ_CLIENT_NAME, name, RQ_CLIENT_ADDR, addr, RQ_USER, user, 0))); } library routines. It bundles its arguments then calls the hosts_access() * access control checker. The host name and user name arguments should be * empty strings, STRING_UNKNOWN or real data. If a match is found, the * optional shell command is executed. * * Restriction: this interface does not pass enough information to support * selective remote username lookups or selective hostname double checks. * * Authotcp_wrappers_7.6/percent_x.c 444 465 506 4542 5700312576 11405 /* * percent_x() takes a string and performs % expansions. It aborts the * program when the expansion would overflow the output buffer. The result * of % expansion may be passed on to a shell process. For this * reason, characters with a special meaning to shells are replaced by * underscores. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) percent_x.c 1.4 94/12/28 17:42:37"; #endif /* System libraries. */ #include #include #include extern void exit(); /* Local stuff. */ #include "tcpd.h" /* percent_x - do % expansion, abort if result buffer is too small */ char *percent_x(result, result_len, string, request) char *result; int result_len; char *string; struct request_info *request; { char *bp = result; char *end = result + result_len - 1; /* end of result buffer */ char *expansion; int expansion_len; static char ok_chars[] = "1234567890!@%-_=+:,./\ abcdefghijklmnopqrstuvwxyz\ ABCDEFGHIJKLMNOPQRSTUVWXYZ"; char *str = string; char *cp; int ch; /* * Warning: we may be called from a child process or after pattern * matching, so we cannot use clean_exit() or tcpd_jump(). */ while (*str) { if (*str == '%' && (ch = str[1]) != 0) { str += 2; expansion = ch == 'a' ? eval_hostaddr(request->client) : ch == 'A' ? eval_hostaddr(request->server) : ch == 'c' ? eval_client(request) : ch == 'd' ? eval_daemon(request) : ch == 'h' ? eval_hostinfo(request->client) : ch == 'H' ? eval_hostinfo(request->server) : ch == 'n' ? eval_hostname(request->client) : ch == 'N' ? eval_hostname(request->server) : ch == 'p' ? eval_pid(request) : ch == 's' ? eval_server(request) : ch == 'u' ? eval_user(request) : ch == '%' ? "%" : (tcpd_warn("unrecognized %%%c", ch), ""); for (cp = expansion; *(cp += strspn(cp, ok_chars)); /* */ ) *cp = '_'; expansion_len = cp - expansion; } else { expansion = str++; expansion_len = 1; } if (bp + expansion_len >= end) { tcpd_warn("percent_x: expansion too long: %.30s...", result); sleep(5); exit(0); } memcpy(bp, expansion, expansion_len); bp += expansion_len; } *bp = 0; return (result); } daemons live in /usr/etc. On some systems, network daemons live in /usr/sbin or in /usr/libexec, or have no `in.\' prefix to their name. .SH EXAMPLE 2 This extcp_wrappers_7.6/options.c 444 465 506 36211 6107411334 11121 /* * General skeleton for adding options to the access control language. The * features offered by this module are documented in the hosts_options(5) * manual page (source file: hosts_options.5, "nroff -man" format). * * Notes and warnings for those who want to add features: * * In case of errors, abort options processing and deny access. There are too * many irreversible side effects to make error recovery feasible. For * example, it makes no sense to continue after we have already changed the * userid. * * In case of errors, do not terminate the process: the routines might be * called from a long-running daemon that should run forever. Instead, call * tcpd_jump() which does a non-local goto back into the hosts_access() * routine. * * In case of severe errors, use clean_exit() instead of directly calling * exit(), or the inetd may loop on an UDP request. * * In verification mode (for example, with the "tcpdmatch" command) the * "dry_run" flag is set. In this mode, an option function should just "say" * what it is going to do instead of really doing it. * * Some option functions do not return (for example, the twist option passes * control to another program). In verification mode (dry_run flag is set) * such options should clear the "dry_run" flag to inform the caller of this * course of action. */ #ifndef lint static char sccsid[] = "@(#) options.c 1.17 96/02/11 17:01:31"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #ifndef MAXPATHNAMELEN #define MAXPATHNAMELEN BUFSIZ #endif /* Local stuff. */ #include "tcpd.h" /* Options runtime support. */ int dry_run = 0; /* flag set in verification mode */ extern jmp_buf tcpd_buf; /* tcpd_jump() support */ /* Options parser support. */ static char whitespace_eq[] = "= \t\r\n"; #define whitespace (whitespace_eq + 1) static char *get_field(); /* chew :-delimited field off string */ static char *chop_string(); /* strip leading and trailing blanks */ /* List of functions that implement the options. Add yours here. */ static void user_option(); /* execute "user name.group" option */ static void group_option(); /* execute "group name" option */ static void umask_option(); /* execute "umask mask" option */ static void linger_option(); /* execute "linger time" option */ static void keepalive_option(); /* execute "keepalive" option */ static void spawn_option(); /* execute "spawn command" option */ static void twist_option(); /* execute "twist command" option */ static void rfc931_option(); /* execute "rfc931" option */ static void setenv_option(); /* execute "setenv name value" */ static void nice_option(); /* execute "nice" option */ static void severity_option(); /* execute "severity value" */ static void allow_option(); /* execute "allow" option */ static void deny_option(); /* execute "deny" option */ static void banners_option(); /* execute "banners path" option */ /* Structure of the options table. */ struct option { char *name; /* keyword name, case is ignored */ void (*func) (); /* function that does the real work */ int flags; /* see below... */ }; #define NEED_ARG (1<<1) /* option requires argument */ #define USE_LAST (1<<2) /* option must be last */ #define OPT_ARG (1<<3) /* option has optional argument */ #define EXPAND_ARG (1<<4) /* do %x expansion on argument */ #define need_arg(o) ((o)->flags & NEED_ARG) #define opt_arg(o) ((o)->flags & OPT_ARG) #define permit_arg(o) ((o)->flags & (NEED_ARG | OPT_ARG)) #define use_last(o) ((o)->flags & USE_LAST) #define expand_arg(o) ((o)->flags & EXPAND_ARG) /* List of known keywords. Add yours here. */ static struct option option_table[] = { "user", user_option, NEED_ARG, "group", group_option, NEED_ARG, "umask", umask_option, NEED_ARG, "linger", linger_option, NEED_ARG, "keepalive", keepalive_option, 0, "spawn", spawn_option, NEED_ARG | EXPAND_ARG, "twist", twist_option, NEED_ARG | EXPAND_ARG | USE_LAST, "rfc931", rfc931_option, OPT_ARG, "setenv", setenv_option, NEED_ARG | EXPAND_ARG, "nice", nice_option, OPT_ARG, "severity", severity_option, NEED_ARG, "allow", allow_option, USE_LAST, "deny", deny_option, USE_LAST, "banners", banners_option, NEED_ARG, 0, }; /* process_options - process access control options */ void process_options(options, request) char *options; struct request_info *request; { char *key; char *value; char *curr_opt; char *next_opt; struct option *op; char bf[BUFSIZ]; for (curr_opt = get_field(options); curr_opt; curr_opt = next_opt) { next_opt = get_field((char *) 0); /* * Separate the option into name and value parts. For backwards * compatibility we ignore exactly one '=' between name and value. */ curr_opt = chop_string(curr_opt); if (*(value = curr_opt + strcspn(curr_opt, whitespace_eq))) { if (*value != '=') { *value++ = 0; value += strspn(value, whitespace); } if (*value == '=') { *value++ = 0; value += strspn(value, whitespace); } } if (*value == 0) value = 0; key = curr_opt; /* * Disallow missing option names (and empty option fields). */ if (*key == 0) tcpd_jump("missing option name"); /* * Lookup the option-specific info and do some common error checks. * Delegate option-specific processing to the specific functions. */ for (op = option_table; op->name && STR_NE(op->name, key); op++) /* VOID */ ; if (op->name == 0) tcpd_jump("bad option name: \"%s\"", key); if (!value && need_arg(op)) tcpd_jump("option \"%s\" requires value", key); if (value && !permit_arg(op)) tcpd_jump("option \"%s\" requires no value", key); if (next_opt && use_last(op)) tcpd_jump("option \"%s\" must be at end", key); if (value && expand_arg(op)) value = chop_string(percent_x(bf, sizeof(bf), value, request)); if (hosts_access_verbose) syslog(LOG_DEBUG, "option: %s %s", key, value ? value : ""); (*(op->func)) (value, request); } } /* allow_option - grant access */ /* ARGSUSED */ static void allow_option(value, request) char *value; struct request_info *request; { longjmp(tcpd_buf, AC_PERMIT); } /* deny_option - deny access */ /* ARGSUSED */ static void deny_option(value, request) char *value; struct request_info *request; { longjmp(tcpd_buf, AC_DENY); } /* banners_option - expand %, terminate each line with CRLF */ static void banners_option(value, request) char *value; struct request_info *request; { char path[MAXPATHNAMELEN]; char ibuf[BUFSIZ]; char obuf[2 * BUFSIZ]; struct stat st; int ch; FILE *fp; sprintf(path, "%s/%s", value, eval_daemon(request)); if ((fp = fopen(path, "r")) != 0) { while ((ch = fgetc(fp)) == 0) write(request->fd, "", 1); ungetc(ch, fp); while (fgets(ibuf, sizeof(ibuf) - 1, fp)) { if (split_at(ibuf, '\n')) strcat(ibuf, "\r\n"); percent_x(obuf, sizeof(obuf), ibuf, request); write(request->fd, obuf, strlen(obuf)); } fclose(fp); } else if (stat(value, &st) < 0) { tcpd_warn("%s: %m", value); } } /* group_option - switch group id */ /* ARGSUSED */ static void group_option(value, request) char *value; struct request_info *request; { struct group *grp; struct group *getgrnam(); if ((grp = getgrnam(value)) == 0) tcpd_jump("unknown group: \"%s\"", value); endgrent(); if (dry_run == 0 && setgid(grp->gr_gid)) tcpd_jump("setgid(%s): %m", value); } /* user_option - switch user id */ /* ARGSUSED */ static void user_option(value, request) char *value; struct request_info *request; { struct passwd *pwd; struct passwd *getpwnam(); char *group; if ((group = split_at(value, '.')) != 0) group_option(group, request); if ((pwd = getpwnam(value)) == 0) tcpd_jump("unknown user: \"%s\"", value); endpwent(); if (dry_run == 0 && setuid(pwd->pw_uid)) tcpd_jump("setuid(%s): %m", value); } /* umask_option - set file creation mask */ /* ARGSUSED */ static void umask_option(value, request) char *value; struct request_info *request; { unsigned mask; char junk; if (sscanf(value, "%o%c", &mask, &junk) != 1 || (mask & 0777) != mask) tcpd_jump("bad umask value: \"%s\"", value); (void) umask(mask); } /* spawn_option - spawn a shell command and wait */ /* ARGSUSED */ static void spawn_option(value, request) char *value; struct request_info *request; { if (dry_run == 0) shell_cmd(value); } /* linger_option - set the socket linger time (Marc Boucher ) */ /* ARGSUSED */ static void linger_option(value, request) char *value; struct request_info *request; { struct linger linger; char junk; if (sscanf(value, "%d%c", &linger.l_linger, &junk) != 1 || linger.l_linger < 0) tcpd_jump("bad linger value: \"%s\"", value); if (dry_run == 0) { linger.l_onoff = (linger.l_linger != 0); if (setsockopt(request->fd, SOL_SOCKET, SO_LINGER, (char *) &linger, sizeof(linger)) < 0) tcpd_warn("setsockopt SO_LINGER %d: %m", linger.l_linger); } } /* keepalive_option - set the socket keepalive option */ /* ARGSUSED */ static void keepalive_option(value, request) char *value; struct request_info *request; { static int on = 1; if (dry_run == 0 && setsockopt(request->fd, SOL_SOCKET, SO_KEEPALIVE, (char *) &on, sizeof(on)) < 0) tcpd_warn("setsockopt SO_KEEPALIVE: %m"); } /* nice_option - set nice value */ /* ARGSUSED */ static void nice_option(value, request) char *value; struct request_info *request; { int niceval = 10; char junk; if (value != 0 && sscanf(value, "%d%c", &niceval, &junk) != 1) tcpd_jump("bad nice value: \"%s\"", value); if (dry_run == 0 && nice(niceval) < 0) tcpd_warn("nice(%d): %m", niceval); } /* twist_option - replace process by shell command */ static void twist_option(value, request) char *value; struct request_info *request; { char *error; if (dry_run != 0) { dry_run = 0; } else { if (resident > 0) tcpd_jump("twist option in resident process"); syslog(deny_severity, "twist %s to %s", eval_client(request), value); /* Before switching to the shell, set up stdin, stdout and stderr. */ #define maybe_dup2(from, to) ((from == to) ? to : (close(to), dup(from))) if (maybe_dup2(request->fd, 0) != 0 || maybe_dup2(request->fd, 1) != 1 || maybe_dup2(request->fd, 2) != 2) { error = "twist_option: dup: %m"; } else { if (request->fd > 2) close(request->fd); (void) execl("/bin/sh", "sh", "-c", value, (char *) 0); error = "twist_option: /bin/sh: %m"; } /* Something went wrong: we MUST terminate the process. */ tcpd_warn(error); clean_exit(request); } } /* rfc931_option - look up remote user name */ static void rfc931_option(value, request) char *value; struct request_info *request; { int timeout; char junk; if (value != 0) { if (sscanf(value, "%d%c", &timeout, &junk) != 1 || timeout <= 0) tcpd_jump("bad rfc931 timeout: \"%s\"", value); rfc931_timeout = timeout; } (void) eval_user(request); } /* setenv_option - set environment variable */ /* ARGSUSED */ static void setenv_option(value, request) char *value; struct request_info *request; { char *var_value; if (*(var_value = value + strcspn(value, whitespace))) *var_value++ = 0; if (setenv(chop_string(value), chop_string(var_value), 1)) tcpd_jump("memory allocation failure"); } /* * The severity option goes last because it comes with a huge amount of ugly * #ifdefs and tables. */ struct syslog_names { char *name; int value; }; static struct syslog_names log_fac[] = { #ifdef LOG_KERN "kern", LOG_KERN, #endif #ifdef LOG_USER "user", LOG_USER, #endif #ifdef LOG_MAIL "mail", LOG_MAIL, #endif #ifdef LOG_DAEMON "daemon", LOG_DAEMON, #endif #ifdef LOG_AUTH "auth", LOG_AUTH, #endif #ifdef LOG_LPR "lpr", LOG_LPR, #endif #ifdef LOG_NEWS "news", LOG_NEWS, #endif #ifdef LOG_UUCP "uucp", LOG_UUCP, #endif #ifdef LOG_CRON "cron", LOG_CRON, #endif #ifdef LOG_LOCAL0 "local0", LOG_LOCAL0, #endif #ifdef LOG_LOCAL1 "local1", LOG_LOCAL1, #endif #ifdef LOG_LOCAL2 "local2", LOG_LOCAL2, #endif #ifdef LOG_LOCAL3 "local3", LOG_LOCAL3, #endif #ifdef LOG_LOCAL4 "local4", LOG_LOCAL4, #endif #ifdef LOG_LOCAL5 "local5", LOG_LOCAL5, #endif #ifdef LOG_LOCAL6 "local6", LOG_LOCAL6, #endif #ifdef LOG_LOCAL7 "local7", LOG_LOCAL7, #endif 0, }; static struct syslog_names log_sev[] = { #ifdef LOG_EMERG "emerg", LOG_EMERG, #endif #ifdef LOG_ALERT "alert", LOG_ALERT, #endif #ifdef LOG_CRIT "crit", LOG_CRIT, #endif #ifdef LOG_ERR "err", LOG_ERR, #endif #ifdef LOG_WARNING "warning", LOG_WARNING, #endif #ifdef LOG_NOTICE "notice", LOG_NOTICE, #endif #ifdef LOG_INFO "info", LOG_INFO, #endif #ifdef LOG_DEBUG "debug", LOG_DEBUG, #endif 0, }; /* severity_map - lookup facility or severity value */ static int severity_map(table, name) struct syslog_names *table; char *name; { struct syslog_names *t; for (t = table; t->name; t++) if (STR_EQ(t->name, name)) return (t->value); tcpd_jump("bad syslog facility or severity: \"%s\"", name); /* NOTREACHED */ } /* severity_option - change logging severity for this event (Dave Mitchell) */ /* ARGSUSED */ static void severity_option(value, request) char *value; struct request_info *request; { char *level = split_at(value, '.'); allow_severity = deny_severity = level ? severity_map(log_fac, value) | severity_map(log_sev, level) : severity_map(log_sev, value); } /* get_field - return pointer to next field in string */ static char *get_field(string) char *string; { static char *last = ""; char *src; char *dst; char *ret; int ch; /* * This function returns pointers to successive fields within a given * string. ":" is the field separator; warn if the rule ends in one. It * replaces a "\:" sequence by ":", without treating the result of * substitution as field terminator. A null argument means resume search * where the previous call terminated. This function destroys its * argument. * * Work from explicit source or from memory. While processing \: we * overwrite the input. This way we do not have to maintain buffers for * copies of input fields. */ src = dst = ret = (string ? string : last); if (src[0] == 0) return (0); while (ch = *src) { if (ch == ':') { if (*++src == 0) tcpd_warn("rule ends in \":\""); break; } if (ch == '\\' && src[1] == ':') src++; *dst++ = *src++; } last = src; *dst = 0; return (ret); } /* chop_string - strip leading and trailing blanks from string */ static char *chop_string(string) register char *string; { char *start = 0; char *end; char *cp; for (cp = string; *cp; cp++) { if (!isspace(*cp)) { if (start == 0) start = cp; end = cp; } } return (start ? (end[1] = 0, start) : cp); } eout: \"%s\"", value); rfc931_timeout = timeout; } (void) eval_user(request); } /* setenv_option - set environment variable */ /* ARGSUSED */ static void setenv_option(value, request) char *value; struct request_info *request; { char *var_value; if (*(var_value = value + strcspn(value, whitespace))) *var_value++ = 0; if (setenv(chop_string(valtcp_wrappers_7.6/clean_exit.c 444 465 506 2070 5700312554 11517 /* * clean_exit() cleans up and terminates the program. It should be called * instead of exit() when for some reason the real network daemon will not or * cannot be run. Reason: in the case of a datagram-oriented service we must * discard the not-yet received data from the client. Otherwise, inetd will * see the same datagram again and again, and go into a loop. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) clean_exit.c 1.4 94/12/28 17:42:19"; #endif #include extern void exit(); #include "tcpd.h" /* clean_exit - clean up and exit */ void clean_exit(request) struct request_info *request; { /* * In case of unconnected protocols we must eat up the not-yet received * data or inetd will loop. */ if (request->sink) request->sink(request->fd); /* * Be kind to the inetd. We already reported the problem via the syslogd, * and there is no need for additional garbage in the logfile. */ sleep(5); exit(0); } verity value */ static int severity_map(table, name) struct syslog_names *table; char *name; { struct syslog_names *t; for (t = table; t->name; t++) if (STR_EQ(t->name, name)) return (t->value); tcpd_jump("bad syslog facility or severity: \"%s\"", name); /* NOTREACHED */ } /* severity_option - change logging severity for this event (Dave Mitchell) */ /* ARGSUSED */ static void severity_option(value, request) char *value; tcp_wrappers_7.6/environ.c 444 465 506 11401 5544055674 11117 /* * Many systems have putenv() but no setenv(). Other systems have setenv() * but no putenv() (MIPS). Still other systems have neither (NeXT). This is a * re-implementation that hopefully ends all problems. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) environ.c 1.2 94/03/23 16:09:46"; #endif /* System libraries. */ extern char **environ; extern char *strchr(); extern char *strcpy(); extern char *strncpy(); extern char *malloc(); extern char *realloc(); extern int strncmp(); extern void free(); #ifdef no_memcpy #define memcpy(d,s,l) bcopy(s,d,l) #else extern char *memcpy(); #endif /* Local stuff. */ static int addenv(); /* append entry to environment */ static int allocated = 0; /* environ is, or is not, allocated */ #define DO_CLOBBER 1 /* namelength - determine length of name in "name=whatever" */ static int namelength(name) char *name; { char *equal; equal = strchr(name, '='); return ((equal == 0) ? strlen(name) : (equal - name)); } /* findenv - given name, locate name=value */ static char **findenv(name, len) char *name; int len; { char **envp; for (envp = environ; envp && *envp; envp++) if (strncmp(name, *envp, len) == 0 && (*envp)[len] == '=') return (envp); return (0); } /* getenv - given name, locate value */ char *getenv(name) char *name; { int len = namelength(name); char **envp = findenv(name, len); return (envp ? *envp + len + 1 : 0); } /* putenv - update or append environment (name,value) pair */ int putenv(nameval) char *nameval; { char *equal = strchr(nameval, '='); char *value = (equal ? equal : ""); return (setenv(nameval, value, DO_CLOBBER)); } /* unsetenv - remove variable from environment */ void unsetenv(name) char *name; { char **envp; if ((envp = findenv(name, namelength(name))) != 0) while (envp[0] = envp[1]) envp++; } /* setenv - update or append environment (name,value) pair */ int setenv(name, value, clobber) char *name; char *value; int clobber; { char *destination; char **envp; int l_name; /* length of name part */ int l_nameval; /* length of name=value */ /* Permit name= and =value. */ l_name = namelength(name); envp = findenv(name, l_name); if (envp != 0 && clobber == 0) return (0); if (*value == '=') value++; l_nameval = l_name + strlen(value) + 1; /* * Use available memory if the old value is long enough. Never free an * old name=value entry because it may not be allocated. */ destination = (envp != 0 && strlen(*envp) >= l_nameval) ? *envp : malloc(l_nameval + 1); if (destination == 0) return (-1); strncpy(destination, name, l_name); destination[l_name] = '='; strcpy(destination + l_name + 1, value); return ((envp == 0) ? addenv(destination) : (*envp = destination, 0)); } /* cmalloc - malloc and copy block of memory */ static char *cmalloc(new_len, old, old_len) char *old; int old_len; { char *new = malloc(new_len); if (new != 0) memcpy(new, old, old_len); return (new); } /* addenv - append environment entry */ static int addenv(nameval) char *nameval; { char **envp; int n_used; /* number of environment entries */ int l_used; /* bytes used excl. terminator */ int l_need; /* bytes needed incl. terminator */ for (envp = environ; envp && *envp; envp++) /* void */ ; n_used = envp - environ; l_used = n_used * sizeof(*envp); l_need = l_used + 2 * sizeof(*envp); envp = allocated ? (char **) realloc((char *) environ, l_need) : (char **) cmalloc(l_need, (char *) environ, l_used); if (envp == 0) { return (-1); } else { allocated = 1; environ = envp; environ[n_used++] = nameval; /* add new entry */ environ[n_used] = 0; /* terminate list */ return (0); } } #ifdef TEST /* * Stand-alone program for test purposes. */ /* printenv - display environment */ static void printenv() { char **envp; for (envp = environ; envp && *envp; envp++) printf("%s\n", *envp); } int main(argc, argv) int argc; char **argv; { char *cp; int changed = 0; if (argc < 2) { printf("usage: %s name[=value]...\n", argv[0]); return (1); } while (--argc && *++argv) { if (argv[0][0] == '-') { /* unsetenv() test */ unsetenv(argv[0] + 1); changed = 1; } else if (strchr(argv[0], '=') == 0) { /* getenv() test */ cp = getenv(argv[0]); printf("%s: %s\n", argv[0], cp ? cp : "not found"); } else { /* putenv() test */ if (putenv(argv[0])) { perror("putenv"); return (1); } changed = 1; } } if (changed) printenv(); return (0); } #endif /* TEST */ st = src; *dst = 0; return (ret); } /* chop_string - strip leading and trailing blanks from string */ static char *chop_string(string) register char *string; { char *start = 0; char *end; char *cp; for (cp = string; *cp; ctcp_wrappers_7.6/patchlevel.h 444 465 506 130 6314551414 11515 #ifndef lint static char patchlevel[] = "@(#) patchlevel 7.6 97/03/21 19:27:23"; #endif ‚1socket.cä$‚2tli.c.c8@¸ DISCLAIMERkL@Ð fakelog.c.d@á safe_finger.cd|@Ôhosts_options.5Œ‚4CHANGES @î try-from.c´@ïupdate.ccNÄ@Þptx.c.cØA' vfprintf.cað@ë tli-sequent.cð@ì tli-sequent.h@²misc.c.(@Îdiag.cc8@Úncr.cagL€† tcpdchk.c`@Ü percent_m.cttcp_wrappers_7.6/fix_options.c 444 465 506 6460 6322310537 11754 /* * Routine to disable IP-level socket options. This code was taken from 4.4BSD * rlogind and kernel source, but all mistakes in it are my fault. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) fix_options.c 1.6 97/04/08 02:29:19"; #endif #include #include #include #include #include #include #include #include #ifndef IPOPT_OPTVAL #define IPOPT_OPTVAL 0 #define IPOPT_OLEN 1 #endif #include "tcpd.h" #define BUFFER_SIZE 512 /* Was: BUFSIZ */ /* fix_options - get rid of IP-level socket options */ fix_options(request) struct request_info *request; { #ifdef IP_OPTIONS unsigned char optbuf[BUFFER_SIZE / 3], *cp; char lbuf[BUFFER_SIZE], *lp; int optsize = sizeof(optbuf), ipproto; struct protoent *ip; int fd = request->fd; unsigned int opt; int optlen; struct in_addr dummy; if ((ip = getprotobyname("ip")) != 0) ipproto = ip->p_proto; else ipproto = IPPROTO_IP; if (getsockopt(fd, ipproto, IP_OPTIONS, (char *) optbuf, &optsize) == 0 && optsize != 0) { /* * Horror! 4.[34] BSD getsockopt() prepends the first-hop destination * address to the result IP options list when source routing options * are present (see ), but produces no output for * other IP options. Solaris 2.x getsockopt() does produce output for * non-routing IP options, and uses the same format as BSD even when * the space for the destination address is unused. The code below * does the right thing with 4.[34]BSD derivatives and Solaris 2, but * may occasionally miss source routing options on incompatible * systems such as Linux. Their choice. * * Look for source routing options. Drop the connection when one is * found. Just wiping the IP options is insufficient: we would still * help the attacker by providing a real TCP sequence number, and the * attacker would still be able to send packets (blind spoofing). I * discussed this attack with Niels Provos, half a year before the * attack was described in open mailing lists. * * It would be cleaner to just return a yes/no reply and let the caller * decide how to deal with it. Resident servers should not terminate. * However I am not prepared to make changes to internal interfaces * on short notice. */ #define ADDR_LEN sizeof(dummy.s_addr) for (cp = optbuf + ADDR_LEN; cp < optbuf + optsize; cp += optlen) { opt = cp[IPOPT_OPTVAL]; if (opt == IPOPT_LSRR || opt == IPOPT_SSRR) { syslog(LOG_WARNING, "refused connect from %s with IP source routing options", eval_client(request)); shutdown(fd, 2); return; } if (opt == IPOPT_EOL) break; if (opt == IPOPT_NOP) { optlen = 1; } else { optlen = cp[IPOPT_OLEN]; if (optlen <= 0) /* Do not loop! */ break; } } lp = lbuf; for (cp = optbuf; optsize > 0; cp++, optsize--, lp += 3) sprintf(lp, " %2.2x", *cp); syslog(LOG_NOTICE, "connect from %s with IP options (ignored):%s", eval_client(request), lbuf); if (setsockopt(fd, ipproto, IP_OPTIONS, (char *) 0, optsize) != 0) { syslog(LOG_ERR, "setsockopt IP_OPTIONS NULL: %m"); shutdown(fd, 2); } } #endif } etenv_option(value, request) char *value; struct request_info *request; { char *var_value; if (*(var_value = value + strcspn(value, whitespace))) *var_value++ = 0; if (setenv(chop_string(valtcp_wrappers_7.6/workarounds.c 444 465 506 16453 6123550462 12016 /* * Workarounds for known system software bugs. This module provides wrappers * around library functions and system calls that are known to have problems * on some systems. Most of these workarounds won't do any harm on regular * systems. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint char sccsid[] = "@(#) workarounds.c 1.6 96/03/19 16:22:25"; #endif #include #include #include #include #include #include #include #include #include #include extern int errno; #include "tcpd.h" /* * Some AIX versions advertise a too small MAXHOSTNAMELEN value (32). * Result: long hostnames would be truncated, and connections would be * dropped because of host name verification failures. Adrian van Bloois * (A.vanBloois@info.nic.surfnet.nl) figured out what was the problem. */ #if (MAXHOSTNAMELEN < 64) #undef MAXHOSTNAMELEN #endif /* In case not defined in . */ #ifndef MAXHOSTNAMELEN #define MAXHOSTNAMELEN 256 /* storage for host name */ #endif /* * Some DG/UX inet_addr() versions return a struct/union instead of a long. * You have this problem when the compiler complains about illegal lvalues * or something like that. The following code fixes this mutant behaviour. * It should not be enabled on "normal" systems. * * Bug reported by ben@piglet.cr.usgs.gov (Rev. Ben A. Mesander). */ #ifdef INET_ADDR_BUG #undef inet_addr long fix_inet_addr(string) char *string; { return (inet_addr(string).s_addr); } #endif /* INET_ADDR_BUG */ /* * With some System-V versions, the fgets() library function does not * account for partial reads from e.g. sockets. The result is that fgets() * gives up too soon, causing username lookups to fail. Problem first * reported for IRIX 4.0.5, by Steve Kotsopoulos . * The following code works around the problem. It does no harm on "normal" * systems. */ #ifdef BROKEN_FGETS #undef fgets char *fix_fgets(buf, len, fp) char *buf; int len; FILE *fp; { char *cp = buf; int c; /* * Copy until the buffer fills up, until EOF, or until a newline is * found. */ while (len > 1 && (c = getc(fp)) != EOF) { len--; *cp++ = c; if (c == '\n') break; } /* * Return 0 if nothing was read. This is correct even when a silly buffer * length was specified. */ if (cp > buf) { *cp = 0; return (buf); } else { return (0); } } #endif /* BROKEN_FGETS */ /* * With early SunOS 5 versions, recvfrom() does not completely fill in the * source address structure when doing a non-destructive read. The following * code works around the problem. It does no harm on "normal" systems. */ #ifdef RECVFROM_BUG #undef recvfrom int fix_recvfrom(sock, buf, buflen, flags, from, fromlen) int sock; char *buf; int buflen; int flags; struct sockaddr *from; int *fromlen; { int ret; /* Assume that both ends of a socket belong to the same address family. */ if ((ret = recvfrom(sock, buf, buflen, flags, from, fromlen)) >= 0) { if (from->sa_family == 0) { struct sockaddr my_addr; int my_addr_len = sizeof(my_addr); if (getsockname(0, &my_addr, &my_addr_len)) { tcpd_warn("getsockname: %m"); } else { from->sa_family = my_addr.sa_family; } } } return (ret); } #endif /* RECVFROM_BUG */ /* * The Apollo SR10.3 and some SYSV4 getpeername(2) versions do not return an * error in case of a datagram-oriented socket. Instead, they claim that all * UDP requests come from address 0.0.0.0. The following code works around * the problem. It does no harm on "normal" systems. */ #ifdef GETPEERNAME_BUG #undef getpeername int fix_getpeername(sock, sa, len) int sock; struct sockaddr *sa; int *len; { int ret; struct sockaddr_in *sin = (struct sockaddr_in *) sa; if ((ret = getpeername(sock, sa, len)) >= 0 && sa->sa_family == AF_INET && sin->sin_addr.s_addr == 0) { errno = ENOTCONN; return (-1); } else { return (ret); } } #endif /* GETPEERNAME_BUG */ /* * According to Karl Vogel (vogelke@c-17igp.wpafb.af.mil) some Pyramid * versions have no yp_default_domain() function. We use getdomainname() * instead. */ #ifdef USE_GETDOMAIN int yp_get_default_domain(ptr) char **ptr; { static char mydomain[MAXHOSTNAMELEN]; *ptr = mydomain; return (getdomainname(mydomain, MAXHOSTNAMELEN)); } #endif /* USE_GETDOMAIN */ #ifndef INADDR_NONE #define INADDR_NONE 0xffffffff #endif /* * Solaris 2.4 gethostbyname() has problems with multihomed hosts. When * doing DNS through NIS, only one host address ends up in the address list. * All other addresses end up in the hostname alias list, interspersed with * copies of the official host name. This would wreak havoc with tcpd's * hostname double checks. Below is a workaround that should do no harm when * accidentally left in. A side effect of the workaround is that address * list members are no longer properly aligned for structure access. */ #ifdef SOLARIS_24_GETHOSTBYNAME_BUG #undef gethostbyname struct hostent *fix_gethostbyname(name) char *name; { struct hostent *hp; struct in_addr addr; char **o_addr_list; char **o_aliases; char **n_addr_list; int broken_gethostbyname = 0; if ((hp = gethostbyname(name)) && !hp->h_addr_list[1] && hp->h_aliases[1]) { for (o_aliases = n_addr_list = hp->h_aliases; *o_aliases; o_aliases++) { if ((addr.s_addr = inet_addr(*o_aliases)) != INADDR_NONE) { memcpy(*n_addr_list++, (char *) &addr, hp->h_length); broken_gethostbyname = 1; } } if (broken_gethostbyname) { o_addr_list = hp->h_addr_list; memcpy(*n_addr_list++, *o_addr_list, hp->h_length); *n_addr_list = 0; hp->h_addr_list = hp->h_aliases; hp->h_aliases = o_addr_list + 1; } } return (hp); } #endif /* SOLARIS_24_GETHOSTBYNAME_BUG */ /* * Horror! Some FreeBSD 2.0 libc routines call strtok(). Since tcpd depends * heavily on strtok(), strange things may happen. Workaround: use our * private strtok(). This has been fixed in the meantime. */ #ifdef USE_STRSEP char *fix_strtok(buf, sep) char *buf; char *sep; { static char *state; char *result; if (buf) state = buf; while ((result = strsep(&state, sep)) && result[0] == 0) /* void */ ; return (result); } #endif /* USE_STRSEP */ /* * IRIX 5.3 (and possibly earlier versions, too) library routines call the * non-reentrant strtok() library routine, causing hosts to slip through * allow/deny filters. Workaround: don't rely on the vendor and use our own * strtok() function. FreeBSD 2.0 has a similar problem (fixed in 2.0.5). */ #ifdef LIBC_CALLS_STRTOK char *my_strtok(buf, sep) char *buf; char *sep; { static char *state; char *result; if (buf) state = buf; /* * Skip over separator characters and detect end of string. */ if (*(state += strspn(state, sep)) == 0) return (0); /* * Skip over non-separator characters and terminate result. */ result = state; if (*(state += strcspn(state, sep)) != 0) *state++ = 0; return (result); } #endif /* LIBC_CALLS_STRTOK */ LOG_KERN "kern", LOG_KERN, #endif #ifdef LOG_USER "user", LOG_USER, #endif #ifdef LOG_MAIL "mail", LOG_MAIL, #endif #ifdef LOG_DAEMON "daemon", LOG_DAEMON, #endif #ifdef LOG_AUTH "auth", LOG_Atcp_wrappers_7.6/socket.c 444 465 506 15312 6314551415 10722 /* * This module determines the type of socket (datagram, stream), the client * socket address and port, the server socket address and port. In addition, * it provides methods to map a transport address to a printable host name * or address. Socket address information results are in static memory. * * The result from the hostname lookup method is STRING_PARANOID when a host * pretends to have someone elses name, or when a host name is available but * could not be verified. * * When lookup or conversion fails the result is set to STRING_UNKNOWN. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) socket.c 1.15 97/03/21 19:27:24"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include extern char *inet_ntoa(); /* Local stuff. */ #include "tcpd.h" /* Forward declarations. */ static void sock_sink(); #ifdef APPEND_DOT /* * Speed up DNS lookups by terminating the host name with a dot. Should be * done with care. The speedup can give problems with lookups from sources * that lack DNS-style trailing dot magic, such as local files or NIS maps. */ static struct hostent *gethostbyname_dot(name) char *name; { char dot_name[MAXHOSTNAMELEN + 1]; /* * Don't append dots to unqualified names. Such names are likely to come * from local hosts files or from NIS. */ if (strchr(name, '.') == 0 || strlen(name) >= MAXHOSTNAMELEN - 1) { return (gethostbyname(name)); } else { sprintf(dot_name, "%s.", name); return (gethostbyname(dot_name)); } } #define gethostbyname gethostbyname_dot #endif /* sock_host - look up endpoint addresses and install conversion methods */ void sock_host(request) struct request_info *request; { static struct sockaddr_in client; static struct sockaddr_in server; int len; char buf[BUFSIZ]; int fd = request->fd; sock_methods(request); /* * Look up the client host address. Hal R. Brand * suggested how to get the client host info in case of UDP connections: * peek at the first message without actually looking at its contents. We * really should verify that client.sin_family gets the value AF_INET, * but this program has already caused too much grief on systems with * broken library code. */ len = sizeof(client); if (getpeername(fd, (struct sockaddr *) & client, &len) < 0) { request->sink = sock_sink; len = sizeof(client); if (recvfrom(fd, buf, sizeof(buf), MSG_PEEK, (struct sockaddr *) & client, &len) < 0) { tcpd_warn("can't get client address: %m"); return; /* give up */ } #ifdef really_paranoid memset(buf, 0 sizeof(buf)); #endif } request->client->sin = &client; /* * Determine the server binding. This is used for client username * lookups, and for access control rules that trigger on the server * address or name. */ len = sizeof(server); if (getsockname(fd, (struct sockaddr *) & server, &len) < 0) { tcpd_warn("getsockname: %m"); return; } request->server->sin = &server; } /* sock_hostaddr - map endpoint address to printable form */ void sock_hostaddr(host) struct host_info *host; { struct sockaddr_in *sin = host->sin; if (sin != 0) STRN_CPY(host->addr, inet_ntoa(sin->sin_addr), sizeof(host->addr)); } /* sock_hostname - map endpoint address to host name */ void sock_hostname(host) struct host_info *host; { struct sockaddr_in *sin = host->sin; struct hostent *hp; int i; /* * On some systems, for example Solaris 2.3, gethostbyaddr(0.0.0.0) does * not fail. Instead it returns "INADDR_ANY". Unfortunately, this does * not work the other way around: gethostbyname("INADDR_ANY") fails. We * have to special-case 0.0.0.0, in order to avoid false alerts from the * host name/address checking code below. */ if (sin != 0 && sin->sin_addr.s_addr != 0 && (hp = gethostbyaddr((char *) &(sin->sin_addr), sizeof(sin->sin_addr), AF_INET)) != 0) { STRN_CPY(host->name, hp->h_name, sizeof(host->name)); /* * Verify that the address is a member of the address list returned * by gethostbyname(hostname). * * Verify also that gethostbyaddr() and gethostbyname() return the same * hostname, or rshd and rlogind may still end up being spoofed. * * On some sites, gethostbyname("localhost") returns "localhost.domain". * This is a DNS artefact. We treat it as a special case. When we * can't believe the address list from gethostbyname("localhost") * we're in big trouble anyway. */ if ((hp = gethostbyname(host->name)) == 0) { /* * Unable to verify that the host name matches the address. This * may be a transient problem or a botched name server setup. */ tcpd_warn("can't verify hostname: gethostbyname(%s) failed", host->name); } else if (STR_NE(host->name, hp->h_name) && STR_NE(host->name, "localhost")) { /* * The gethostbyaddr() and gethostbyname() calls did not return * the same hostname. This could be a nameserver configuration * problem. It could also be that someone is trying to spoof us. */ tcpd_warn("host name/name mismatch: %s != %.*s", host->name, STRING_LENGTH, hp->h_name); } else { /* * The address should be a member of the address list returned by * gethostbyname(). We should first verify that the h_addrtype * field is AF_INET, but this program has already caused too much * grief on systems with broken library code. */ for (i = 0; hp->h_addr_list[i]; i++) { if (memcmp(hp->h_addr_list[i], (char *) &sin->sin_addr, sizeof(sin->sin_addr)) == 0) return; /* name is good, keep it */ } /* * The host name does not map to the initial address. Perhaps * someone has messed up. Perhaps someone compromised a name * server. */ tcpd_warn("host name/address mismatch: %s != %.*s", inet_ntoa(sin->sin_addr), STRING_LENGTH, hp->h_name); } strcpy(host->name, paranoid); /* name is bad, clobber it */ } } /* sock_sink - absorb unreceived IP datagram */ static void sock_sink(fd) int fd; { char buf[BUFSIZ]; struct sockaddr_in sin; int size = sizeof(sin); /* * Eat up the not-yet received datagram. Some systems insist on a * non-zero source address argument in the recvfrom() call below. */ (void) recvfrom(fd, buf, sizeof(buf), 0, (struct sockaddr *) & sin, &size); } & server, &len) < 0) { tcpd_warn("getsockname: %m"); return; } request->server->sin = &server; } /* sock_hostaddr - map endpoint address to printable form */ void sock_hostaddr(host) struct host_info *host; { struct sockaddr_in *sin = host->sin; if (sin != 0) STRN_CPY(host->addr, inetcp_wrappers_7.6/tli.c 444 465 506 22704 6314551416 10226 /* * tli_host() determines the type of transport (connected, connectionless), * the transport address of a client host, and the transport address of a * server endpoint. In addition, it provides methods to map a transport * address to a printable host name or address. Socket address results are * in static memory; tli structures are allocated from the heap. * * The result from the hostname lookup method is STRING_PARANOID when a host * pretends to have someone elses name, or when a host name is available but * could not be verified. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) tli.c 1.15 97/03/21 19:27:25"; #endif #ifdef TLI /* System libraries. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include extern char *nc_sperror(); extern int errno; extern char *sys_errlist[]; extern int sys_nerr; extern int t_errno; extern char *t_errlist[]; extern int t_nerr; /* Local stuff. */ #include "tcpd.h" /* Forward declarations. */ static void tli_endpoints(); static struct netconfig *tli_transport(); static void tli_hostname(); static void tli_hostaddr(); static void tli_cleanup(); static char *tli_error(); static void tli_sink(); /* tli_host - look up endpoint addresses and install conversion methods */ void tli_host(request) struct request_info *request; { static struct sockaddr_in client; static struct sockaddr_in server; /* * If we discover that we are using an IP transport, pretend we never * were here. Otherwise, use the transport-independent method and stick * to generic network addresses. XXX hard-coded protocol family name. */ tli_endpoints(request); if ((request->config = tli_transport(request->fd)) != 0 && STR_EQ(request->config->nc_protofmly, "inet")) { if (request->client->unit != 0) { client = *(struct sockaddr_in *) request->client->unit->addr.buf; request->client->sin = &client; } if (request->server->unit != 0) { server = *(struct sockaddr_in *) request->server->unit->addr.buf; request->server->sin = &server; } tli_cleanup(request); sock_methods(request); } else { request->hostname = tli_hostname; request->hostaddr = tli_hostaddr; request->cleanup = tli_cleanup; } } /* tli_cleanup - cleanup some dynamically-allocated data structures */ static void tli_cleanup(request) struct request_info *request; { if (request->config != 0) freenetconfigent(request->config); if (request->client->unit != 0) t_free((char *) request->client->unit, T_UNITDATA); if (request->server->unit != 0) t_free((char *) request->server->unit, T_UNITDATA); } /* tli_endpoints - determine TLI client and server endpoint information */ static void tli_endpoints(request) struct request_info *request; { struct t_unitdata *server; struct t_unitdata *client; int fd = request->fd; int flags; /* * Determine the client endpoint address. With unconnected services, peek * at the sender address of the pending protocol data unit without * popping it off the receive queue. This trick works because only the * address member of the unitdata structure has been allocated. * * Beware of successful returns with zero-length netbufs (for example, * Solaris 2.3 with ticlts transport). The netdir(3) routines can't * handle that. Assume connection-less transport when TI_GETPEERNAME * produces no usable result, even when t_rcvudata() is unable to figure * out the peer address. Better to hang than to loop. */ if ((client = (struct t_unitdata *) t_alloc(fd, T_UNITDATA, T_ADDR)) == 0) { tcpd_warn("t_alloc: %s", tli_error()); return; } if (ioctl(fd, TI_GETPEERNAME, &client->addr) < 0 || client->addr.len == 0) { request->sink = tli_sink; if (t_rcvudata(fd, client, &flags) < 0 || client->addr.len == 0) { tcpd_warn("can't get client address: %s", tli_error()); t_free((void *) client, T_UNITDATA); return; } } request->client->unit = client; /* * Look up the server endpoint address. This can be used for filtering on * server address or name, or to look up the client user. */ if ((server = (struct t_unitdata *) t_alloc(fd, T_UNITDATA, T_ADDR)) == 0) { tcpd_warn("t_alloc: %s", tli_error()); return; } if (ioctl(fd, TI_GETMYNAME, &server->addr) < 0) { tcpd_warn("TI_GETMYNAME: %m"); t_free((void *) server, T_UNITDATA); return; } request->server->unit = server; } /* tli_transport - find out TLI transport type */ static struct netconfig *tli_transport(fd) int fd; { struct stat from_client; struct stat from_config; void *handlep; struct netconfig *config; /* * Assuming that the network device is a clone device, we must compare * the major device number of stdin to the minor device number of the * devices listed in the netconfig table. */ if (fstat(fd, &from_client) != 0) { tcpd_warn("fstat(fd %d): %m", fd); return (0); } if ((handlep = setnetconfig()) == 0) { tcpd_warn("setnetconfig: %m"); return (0); } while (config = getnetconfig(handlep)) { if (stat(config->nc_device, &from_config) == 0) { if (minor(from_config.st_rdev) == major(from_client.st_rdev)) break; } } if (config == 0) { tcpd_warn("unable to identify transport protocol"); return (0); } /* * Something else may clobber our getnetconfig() result, so we'd better * acquire our private copy. */ if ((config = getnetconfigent(config->nc_netid)) == 0) { tcpd_warn("getnetconfigent(%s): %s", config->nc_netid, nc_sperror()); return (0); } return (config); } /* tli_hostaddr - map TLI transport address to printable address */ static void tli_hostaddr(host) struct host_info *host; { struct request_info *request = host->request; struct netconfig *config = request->config; struct t_unitdata *unit = host->unit; char *uaddr; if (config != 0 && unit != 0 && (uaddr = taddr2uaddr(config, &unit->addr)) != 0) { STRN_CPY(host->addr, uaddr, sizeof(host->addr)); free(uaddr); } } /* tli_hostname - map TLI transport address to hostname */ static void tli_hostname(host) struct host_info *host; { struct request_info *request = host->request; struct netconfig *config = request->config; struct t_unitdata *unit = host->unit; struct nd_hostservlist *servlist; if (config != 0 && unit != 0 && netdir_getbyaddr(config, &servlist, &unit->addr) == ND_OK) { struct nd_hostserv *service = servlist->h_hostservs; struct nd_addrlist *addr_list; int found = 0; if (netdir_getbyname(config, service, &addr_list) != ND_OK) { /* * Unable to verify that the name matches the address. This may * be a transient problem or a botched name server setup. We * decide to play safe. */ tcpd_warn("can't verify hostname: netdir_getbyname(%.*s) failed", STRING_LENGTH, service->h_host); } else { /* * Look up the host address in the address list we just got. The * comparison is done on the textual representation, because the * transport address is an opaque structure that may have holes * with uninitialized garbage. This approach obviously loses when * the address does not have a textual representation. */ char *uaddr = eval_hostaddr(host); char *ua; int i; for (i = 0; found == 0 && i < addr_list->n_cnt; i++) { if ((ua = taddr2uaddr(config, &(addr_list->n_addrs[i]))) != 0) { found = !strcmp(ua, uaddr); free(ua); } } netdir_free((void *) addr_list, ND_ADDRLIST); /* * When the host name does not map to the initial address, assume * someone has compromised a name server. More likely someone * botched it, but that could be dangerous, too. */ if (found == 0) tcpd_warn("host name/address mismatch: %s != %.*s", host->addr, STRING_LENGTH, service->h_host); } STRN_CPY(host->name, found ? service->h_host : paranoid, sizeof(host->name)); netdir_free((void *) servlist, ND_HOSTSERVLIST); } } /* tli_error - convert tli error number to text */ static char *tli_error() { static char buf[40]; if (t_errno != TSYSERR) { if (t_errno < 0 || t_errno >= t_nerr) { sprintf(buf, "Unknown TLI error %d", t_errno); return (buf); } else { return (t_errlist[t_errno]); } } else { if (errno < 0 || errno >= sys_nerr) { sprintf(buf, "Unknown UNIX error %d", errno); return (buf); } else { return (sys_errlist[errno]); } } } /* tli_sink - absorb unreceived datagram */ static void tli_sink(fd) int fd; { struct t_unitdata *unit; int flags; /* * Something went wrong. Absorb the datagram to keep inetd from looping. * Allocate storage for address, control and data. If that fails, sleep * for a couple of seconds in an attempt to keep inetd from looping too * fast. */ if ((unit = (struct t_unitdata *) t_alloc(fd, T_UNITDATA, T_ALL)) == 0) { tcpd_warn("t_alloc: %s", tli_error()); sleep(5); } else { (void) t_rcvudata(fd, unit, &flags); t_free((void *) unit, T_UNITDATA); } } #endif /* TLI */ fig->nc_device, &from_config) == 0) { if (minor(from_cotcp_wrappers_7.6/DISCLAIMER 444 465 506 1430 6107411323 10572 /************************************************************************ * Copyright 1995 by Wietse Venema. All rights reserved. Some individual * files may be covered by other copyrights. * * This material was originally written and compiled by Wietse Venema at * Eindhoven University of Technology, The Netherlands, in 1990, 1991, * 1992, 1993, 1994 and 1995. * * Redistribution and use in source and binary forms are permitted * provided that this entire copyright notice is duplicated in all such * copies. * * This software is provided "as is" and without any expressed or implied * warranties, including, without limitation, the implied warranties of * merchantibility and fitness for any particular purpose. ************************************************************************/ se { return (sys_errlist[errno]); } } } /* tli_sink - absorb unreceived datagram */ static void tli_sink(fd) int fd; { struct t_unitdata *unit; int flags; /* * Something went wrong. Absorb the dtcp_wrappers_7.6/fakelog.c 444 465 506 1740 5700312556 11021 /* * This module intercepts syslog() library calls and redirects their output * to the standard output stream. For interactive testing. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) fakelog.c 1.3 94/12/28 17:42:21"; #endif #include #include "mystdarg.h" /* openlog - dummy */ /* ARGSUSED */ openlog(name, logopt, facility) char *name; int logopt; int facility; { /* void */ } /* vsyslog - format one record */ vsyslog(severity, fmt, ap) int severity; char *fmt; va_list ap; { char buf[BUFSIZ]; vprintf(percent_m(buf, fmt), ap); printf("\n"); fflush(stdout); } /* syslog - format one record */ /* VARARGS */ VARARGS(syslog, int, severity) { va_list ap; char *fmt; VASTART(ap, int, severity); fmt = va_arg(ap, char *); vsyslog(severity, fmt, ap); VAEND(ap); } /* closelog - dummy */ closelog() { /* void */ } /* void */ } /* vsyslog - ftcp_wrappers_7.6/safe_finger.c 444 465 506 11756 5700312602 11701 /* * safe_finger - finger client wrapper that protects against nasty stuff * from finger servers. Use this program for automatic reverse finger * probes, not the raw finger command. * * Build with: cc -o safe_finger safe_finger.c * * The problem: some programs may react to stuff in the first column. Other * programs may get upset by thrash anywhere on a line. File systems may * fill up as the finger server keeps sending data. Text editors may bomb * out on extremely long lines. The finger server may take forever because * it is somehow wedged. The code below takes care of all this badness. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) safe_finger.c 1.4 94/12/28 17:42:41"; #endif /* System libraries */ #include #include #include #include #include #include extern void exit(); /* Local stuff */ char path[] = "PATH=/bin:/usr/bin:/usr/ucb:/usr/bsd:/etc:/usr/etc:/usr/sbin"; #define TIME_LIMIT 60 /* Do not keep listinging forever */ #define INPUT_LENGTH 100000 /* Do not keep listinging forever */ #define LINE_LENGTH 128 /* Editors can choke on long lines */ #define FINGER_PROGRAM "finger" /* Most, if not all, UNIX systems */ #define UNPRIV_NAME "nobody" /* Preferred privilege level */ #define UNPRIV_UGID 32767 /* Default uid and gid */ int finger_pid; void cleanup(sig) int sig; { kill(finger_pid, SIGKILL); exit(0); } main(argc, argv) int argc; char **argv; { int c; int line_length = 0; int finger_status; int wait_pid; int input_count = 0; struct passwd *pwd; /* * First of all, let's don't run with superuser privileges. */ if (getuid() == 0 || geteuid() == 0) { if ((pwd = getpwnam(UNPRIV_NAME)) && pwd->pw_uid > 0) { setgid(pwd->pw_gid); setuid(pwd->pw_uid); } else { setgid(UNPRIV_UGID); setuid(UNPRIV_UGID); } } /* * Redirect our standard input through the raw finger command. */ if (putenv(path)) { fprintf(stderr, "%s: putenv: out of memory", argv[0]); exit(1); } argv[0] = FINGER_PROGRAM; finger_pid = pipe_stdin(argv); /* * Don't wait forever (Peter Wemm ). */ signal(SIGALRM, cleanup); (void) alarm(TIME_LIMIT); /* * Main filter loop. */ while ((c = getchar()) != EOF) { if (input_count++ >= INPUT_LENGTH) { /* don't listen forever */ fclose(stdin); printf("\n\n Input truncated to %d bytes...\n", input_count - 1); break; } if (c == '\n') { /* good: end of line */ putchar(c); line_length = 0; } else { if (line_length >= LINE_LENGTH) { /* force end of line */ printf("\\\n"); line_length = 0; } if (line_length == 0) { /* protect left margin */ putchar(' '); line_length++; } if (isascii(c) && (isprint(c) || isspace(c))) { /* text */ if (c == '\\') { putchar(c); line_length++; } putchar(c); line_length++; } else { /* quote all other thash */ printf("\\%03o", c & 0377); line_length += 4; } } } /* * Wait until the finger child process has terminated and account for its * exit status. Which will always be zero on most systems. */ while ((wait_pid = wait(&finger_status)) != -1 && wait_pid != finger_pid) /* void */ ; return (wait_pid != finger_pid || finger_status != 0); } /* perror_exit - report system error text and terminate */ void perror_exit(text) char *text; { perror(text); exit(1); } /* pipe_stdin - pipe stdin through program (from my ANSI to OLD C converter) */ int pipe_stdin(argv) char **argv; { int pipefds[2]; int pid; int i; struct stat st; /* * The code that sets up the pipe requires that file descriptors 0,1,2 * are already open. All kinds of mysterious things will happen if that * is not the case. The following loops makes sure that descriptors 0,1,2 * are set up properly. */ for (i = 0; i < 3; i++) { if (fstat(i, &st) == -1 && open("/dev/null", 2) != i) perror_exit("open /dev/null"); } /* * Set up the pipe that interposes the command into our standard input * stream. */ if (pipe(pipefds)) perror_exit("pipe"); switch (pid = fork()) { case -1: /* error */ perror_exit("fork"); /* NOTREACHED */ case 0: /* child */ (void) close(pipefds[0]); /* close reading end */ (void) close(1); /* connect stdout to pipe */ if (dup(pipefds[1]) != 1) perror_exit("dup"); (void) close(pipefds[1]); /* close redundant fd */ (void) execvp(argv[0], argv); perror_exit(argv[0]); /* NOTREACHED */ default: /* parent */ (void) close(pipefds[1]); /* close writing end */ (void) close(0); /* connect stdin to pipe */ if (dup(pipefds[0]) != 0) perror_exit("dup"); (void) close(pipefds[0]); /* close redundant fd */ return (pid); } } ; } } /*tcp_wrappers_7.6/hosts_options.5 444 465 506 14777 5700312565 12304 .TH HOSTS_OPTIONS 5 .SH NAME hosts_options \- host access control language extensions .SH DESCRIPTION This document describes optional extensions to the language described in the hosts_access(5) document. The extensions are enabled at program build time. For example, by editing the Makefile and turning on the PROCESS_OPTIONS compile-time option. .PP The extensible language uses the following format: .sp .ti +3 daemon_list : client_list : option : option ... .PP The first two fields are described in the hosts_access(5) manual page. The remainder of the rules is a list of zero or more options. Any ":" characters within options should be protected with a backslash. .PP An option is of the form "keyword" or "keyword value". Options are processed in the specified order. Some options are subjected to % substitutions. For the sake of backwards compatibility with earlier versions, an "=" is permitted between keyword and value. .SH LOGGING .IP "severity mail.info" .IP "severity notice" Change the severity level at which the event will be logged. Facility names (such as mail) are optional, and are not supported on systems with older syslog implementations. The severity option can be used to emphasize or to ignore specific events. .SH ACCESS CONTROL .IP "allow" .IP "deny" Grant (deny) service. These options must appear at the end of a rule. .PP The \fIallow\fR and \fIdeny\fR keywords make it possible to keep all access control rules within a single file, for example in the \fIhosts.allow\fR file. .sp To permit access from specific hosts only: .sp .ne 2 .ti +3 ALL: .friendly.domain: ALLOW .ti +3 ALL: ALL: DENY .sp To permit access from all hosts except a few trouble makers: .sp .ne 2 .ti +3 ALL: .bad.domain: DENY .ti +3 ALL: ALL: ALLOW .sp Notice the leading dot on the domain name patterns. .SH RUNNING OTHER COMMANDS .IP "spawn shell_command" Execute, in a child process, the specified shell command, after performing the % expansions described in the hosts_access(5) manual page. The command is executed with stdin, stdout and stderr connected to the null device, so that it won\'t mess up the conversation with the client host. Example: .sp .nf .ti +3 spawn (/some/where/safe_finger -l @%h | /usr/ucb/mail root) & .fi .sp executes, in a background child process, the shell command "safe_finger -l @%h | mail root" after replacing %h by the name or address of the remote host. .sp The example uses the "safe_finger" command instead of the regular "finger" command, to limit possible damage from data sent by the finger server. The "safe_finger" command is part of the daemon wrapper package; it is a wrapper around the regular finger command that filters the data sent by the remote host. .IP "twist shell_command" Replace the current process by an instance of the specified shell command, after performing the % expansions described in the hosts_access(5) manual page. Stdin, stdout and stderr are connected to the client process. This option must appear at the end of a rule. .sp To send a customized bounce message to the client instead of running the real ftp daemon: .sp .nf .ti +3 in.ftpd : ... : twist /bin/echo 421 Some bounce message .fi .sp For an alternative way to talk to client processes, see the \fIbanners\fR option below. .sp To run /some/other/in.telnetd without polluting its command-line array or its process environment: .sp .nf .ti +3 in.telnetd : ... : twist PATH=/some/other; exec in.telnetd .fi .sp Warning: in case of UDP services, do not twist to commands that use the standard I/O or the read(2)/write(2) routines to communicate with the client process; UDP requires other I/O primitives. .SH NETWORK OPTIONS .IP "keepalive" Causes the server to periodically send a message to the client. The connection is considered broken when the client does not respond. The keepalive option can be useful when users turn off their machine while it is still connected to a server. The keepalive option is not useful for datagram (UDP) services. .IP "linger number_of_seconds" Specifies how long the kernel will try to deliver not-yet delivered data after the server process closes a connection. .SH USERNAME LOOKUP .IP "rfc931 [ timeout_in_seconds ]" Look up the client user name with the RFC 931 (TAP, IDENT, RFC 1413) protocol. This option is silently ignored in case of services based on transports other than TCP. It requires that the client system runs an RFC 931 (IDENT, etc.) -compliant daemon, and may cause noticeable delays with connections from non-UNIX clients. The timeout period is optional. If no timeout is specified a compile-time defined default value is taken. .SH MISCELLANEOUS .IP "banners /some/directory" Look for a file in `/some/directory' with the same name as the daemon process (for example in.telnetd for the telnet service), and copy its contents to the client. Newline characters are replaced by carriage-return newline, and % sequences are expanded (see the hosts_access(5) manual page). .sp The tcp wrappers source code distribution provides a sample makefile (Banners.Makefile) for convenient banner maintenance. .sp Warning: banners are supported for connection-oriented (TCP) network services only. .IP "nice [ number ]" Change the nice value of the process (default 10). Specify a positive value to spend more CPU resources on other processes. .IP "setenv name value" Place a (name, value) pair into the process environment. The value is subjected to % expansions and may contain whitespace (but leading and trailing blanks are stripped off). .sp Warning: many network daemons reset their environment before spawning a login or shell process. .IP "umask 022" Like the umask command that is built into the shell. An umask of 022 prevents the creation of files with group and world write permission. The umask argument should be an octal number. .IP "user nobody" .IP "user nobody.kmem" Assume the privileges of the "nobody" userid (or user "nobody", group "kmem"). The first form is useful with inetd implementations that run all services with root privilege. The second form is useful for services that need special group privileges only. .SH DIAGNOSTICS When a syntax error is found in an access control rule, the error is reported to the syslog daemon; further options will be ignored, and service is denied. .SH SEE ALSO hosts_access(5), the default access control language .SH AUTHOR .na .nf Wietse Venema (wietse@wzv.win.tue.nl) Department of Mathematics and Computing Science Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) hosts_options.5 1.10 94/12/28 17:42:28 stcp_wrappers_7.6/CHANGES 444 465 506 45373 6314551410 10266 Request: after building the programs, please run the `tcpdchk' wrapper configuration checker. See the `tcpdchk.8' manual page (`nroff -man' format) for instructions. `tcpdchk' automatically identifies the most common configuration problems, and will save you and me a lot of time. Changes per release 7.6 (Mar 1997) ================================== - Improved the anti source-routing protection. The code in version 7.5 was not as strong as it could be, because I tried to be compatible with Linux. That was a mistake. Sorry for the inconvenience. - The program no longer terminates case of a source-routed connection, making the IP-spoofing code more usable for long-running daemons. - When syslogging DNS hostname problems, always stop after a limited number of characters. Changes per release 7.5 (Feb 1997) ================================== - Optionally refuse source-routed TCP connections requests altogether. Credits to Niels Provos of Universitaet Hamburg. File: fix_options.c. - Support for IRIX 6 (Lael Tucker). - Support for Amdahl UTS 2.1.5 (Richard E. Richmond). - Support for SINIX 5.42 (Klaus Nielsen). - SCO 5 now has vsyslog() (Bill Golden). - Hints and tips for dealing with IRIX inetd (Niko Makila, Aaron M Lee). - Support for BSD/OS (Paul Borman). - Support for Tandem (Emad Qawas). - Support for ISC (Frederick B. Cohen). - Workaround for UNICOS - it would choke on a setjmp() expression (Bruce Kelly). File: hosts_access.c, tcpdchk.c. - Increased the level of buffer overflow paranoia when printing unwanted IP options. File: fix_options.c. Changes per release 7.4 (Mar 1996) ================================== - IRIX 5.3 (and possibly, earlier releases, too) library routines call the non-reentrant strtok() routine. The result is that hosts may slip through allow/deny filters. Workaround is to not rely on the vendor's strtok() routine (#ifdef LIBC_CALLS_STRTOK). Credits to Th. Eifert (Aachen University) for spotting this one. This fix supersedes the earlier workaround for a similar problem in FreeBSD 2.0. Changes per release 7.3 (Feb 1996) ================================== - More tests added to tcpdchk and tcpdmatch: make sure that the REAL_DAEMON_DIR actually is a directory and not a regular file; detect if tcpd recursively calls itself. - Edwin Kremer found an amusing fencepost error in the xgets() routine: lines longer than BUFLEN characters would be garbled. - The access control routines now refuse to execute "dangerous" actions such as `twist' when they are called from within a resident process. This prevents you from shooting yourself into the foot with critical systems programs such as, e.g., portmap or rpcbind. - Support for Unicos 8.x (Bruce Kelly). The program now closes the syslog client socket before running the real daemon: Cray UNICOS refuses to checkpoint processes with open network ports. - Support for MachTen UNIX (Albert M.C Tam). - Support for Interactive UNIX R3.2 V4.0 (Bobby D. Wright). - Support for SCO 3.2v5.0.0 OpenServer 5 (bob@odt.handy.com) - Support for Unixware 1.x and Unixware 2.x. The old Unixware Makefile rule was broken. Sorry about that. - Some FreeBSD 2.0 libc routines call strtok() and severely mess up the allow/deny rule processing. This is very bad. Workaround: call our own strtok() clone (#ifdef USE_STRSEP). - The programs now log a warning when they detect that a non-existent banner directory is specified. - The hosts_access.3 manual page used obsolete names for the RQ_* constants. Changes per release 7.2 (Jan 1995) ================================== - Added a note to the README and manpages on using the IDENT service to detect sequence number spoofing and other host impersonation attacks. - Portability: ConvexOS puts RPC version numbers before the daemon path name (Jukka Ukkonen). - Portability: the AIX compiler disliked the strchr() declaration in socket.c. I should have removed it when I included . - Backwards compatibility: some people relied on the old leading dot or trailing dot magic in daemon process names. - Backwards compatibility: hostname lookup remains enabled when -DPARANOID is turned off. In order to disable hostname lookups you must turn off -DALWAYS_HOSTNAME. - Eliminated false complaints from the tcpdmatch/tcpdchk configuration checking programs about process names not in inetd.conf or about KNOWN username patterns. Changes per release 7.1 (Jan 1995) ================================== - Portability: HP-UX permits you to break inetd.conf entries with backslash-newline. - Portability: EP/IX has no putenv() and some inetd.conf entries are spread out over two lines. - Portability: SCO with NIS support has no *netgrent() routines. Changes per release 7.0 (Jan 1995) ================================== - Added a last-minute workaround for a Solaris 2.4 gethostbyname() foulup with multi-homed hosts in DNS through NIS mode. - Added a last-minute defense against TLI weirdness: address lookups apparently succeed but the result netbuf is empty (ticlts transport). - Dropped several new solutions that were in need of a problem. Beta testers may recognize what new features were kicked out during the last weeks before release 7.0 came out. Such is life. - Got rid of out the environment replacement routines, at least for most architectures. One should not have to replace working system software when all that is needed is a 4.4BSD setenv() emulator. - By popular request I have added an option to send banner messages to clients. There is a Banners.Makefile that gives some aid for sites that are going to use this feature. John C. Wingenbach did some pioneering work here. I used to think that banners are frivolous. Now that I had a personal need for them I know that banners can be useful. - At last: an extensible functional interface to the pattern matching engine. request_init() and request_set() accept a variable-length name-value argument list. The result can be passed to hosts_access(). - When PARANOID mode is disabled (compile time), the wrapper does no hostname lookup or hostname double checks unless required by %letter expansions, or by access control rules that match host names. This is useful for sites that don't care about internet hostnames anyway. Inspired by the authors of the firewalls and internet security book. - When PARANOID mode is disabled (compile time), hosts with a name/name or name/address conflict can be matched with the PARANOID host wildcard pattern, so that you can take some intelligent action instead of just dropping clients. Like showing a banner that explains the problem. - New percent escapes: %A expands to the server address; %H expands to the corresponding hostname (or address if no name is available); %n and %N expand to the client and server hostname (or "unknown"); %s expands to everything we know about the server endpoint (the opposite of the %c sequence for client information). - Symmetry: server and client host information is now treated on equal footing, so that we can reuse a lot of code. - Lazy evaluation of host names, host addresses, usernames, and so on, to avoid doing unnecessary work. - Dropping #ifdefs for some archaic systems made the code simpler. - Dropping the FAIL pattern made the pattern matcher much simpler. Run the "tcpdchk" program to scan your access control files for any uses of this obscure language feature. - Moving host-specific pattern matching from string_match() to the host_match() routine made the code more accurate. Run the "tcpdchk" program to scan your access control files for any dependencies on undocumented or obscure language features that are gone. - daemon@host patterns trigger on clients that connect to a specific internet address. This can be useful for service providers that offer multiple ftp or www archives on different internet addresses, all belonging to one and the same host (www.foo.com, ftp.bar.com, you get the idea). Inspired by a discussion with Rop Gonggrijp, Cor Bosman, and Casper Dik, and earlier discussions with Adrian van Bloois. - The new "tcpdchk" program critcizes all your access control rules and inetd.conf entries. Great for spotting obscure bugs in my own hosts.xxx files. This program also detects hosts with name/address conflicts and with other DNS-related problems. See the "tcpdchk.8" manual page. - The "tcpdmatch" program replaces the poor old "try" command. The new program looks in your inetd.conf file and therefore produces much more accurate predictions. In addition, it detects hosts with name/address conflicts and with other DNS-related problems. See the "tcpdmatch.8" manual page. The inetd.conf lookup was suggested by Everett F Batey. - In the access control tables, the `=' between option name and value is no longer required. - Added 60-second timeout to the safe_finger command, to cover another potential problem. Suggested by Peter Wemm. - Andrew Maffei provided code that works with WIN-TCP on NCR System V.4 UNIX. It reportedly works with versions 02.02.01 and 02.03.00. The code pops off all streams modules above the device driver, pushes the timod module to get at the peer address, and then restores the streams stack to the initial state. Changes per release 6.3 (Mar 1994) ================================== - Keepalives option, to get rid of stuck daemons when people turn off their PC while still connected. Files: options.c, hosts_options.5. - Nice option, to calm down network daemons that take away too much CPU time. Files: options.c, hosts_options.5. - Ultrix perversion: the environ global pointer may be null. The environment replacement routines now check for this. File: environ.c. - Fixed a few places that still assumed the socket is on standard input. Fixed some error messages that did not provide access control file name and line number. File: options.c. - Just when I was going to release 6.2 I received code for Dynix/PTX. That code is specific to PTX 2.x, so I'll keep around my generic PTX code just in case. The difference is in the handling of UDP services. Files: tli_sequent.[hc]. Changes per release 6.2 (Feb 1994) ================================== - Resurrected my year-old code to reduce DNS load by appending a dot to the gethostbyname() argument. This feature is still experimental and it may go away if it causes more problems than it solves. File: socket.c. - Auxiliary code for the Pyramid, BSD universe. Karl Vogel figured out what was missing: yp_get_default_domain() and vfprintf(). Files: workarounds.c, vfprintf.c. - Improved support for Dynix/PTX. The wrapper should now be able to deal with all TLI over IP services. File: ptx.c. - The try command now uses the hostname that gethostbyaddr() would return, instead of the hostname returned by gethostbyname(). This can be significant on systems with NIS that have short host names in the hosts map. For example, gethostbyname("wzv.win.tue.nl") returns "wzv.win.tue.nl"; gethostbyaddr(131.155.210.17) returns "wzv", and that is what we should test with. File: try.c. Changes per release 6.1 (Dec 1993) ================================== - Re-implemented all environment access routines. Most systems have putenv() but no setenv(), some systems have setenv() but no putenv(), and there are even systems that have neither setenv() nor putenv(). The benefit of all this is that more systems can now be treated in the same way. File: environ.c. - Workaround for a weird problem with DG/UX when the wrapper is run as nobody (i.e. fingerd). For some reason the ioctl(fd, I_FIND, "sockmod") call fails even with socket-based applications. The "fix" is to always assume sockets when the ioctl(fd, I_FIND, "timod") call fails. File: fromhost.c. Thanks to Paul de Vries (vries@dutentb.et.tudelft.nl) for helping me to figure out this one. - Implemented a workaround for Dynix/PTX and other systems with TLI that lack some essential support routines. Thanks to Bugs Brouillard (brouill@hsuseq.humboldt.edu) for the hospitality to try things out. The trick is to temporarily switch to the socket API to identify the client, and to switch back to TLI when done. It still does not work right for basic network services such as telnet. File: fromhost.c. - Easy-to-build procedures for SCO UNIX, ConvexOS with UltraNet, EP/IX, Dynix 3.2, Dynix/PTX. File: Makefile. - Variable rfc931 timeout. Files: rfc931.c, options.c, log_tcp.h, try.c. - Further simplification of the rfc931 code. File: rfc931.c. - The fromhost() interface stinks: I cannot change that, but at least the from_sock() and from_tli() functions now accept a file descriptor argument. - Fixed a buglet: fromhost() would pass a garbage file descriptor to the isastream() call. - On some systems the finger client program lives in /usr/bsd. File: safe_finger.c. Changes per release 6.0 (Sept 1993) =================================== - Easy build procedures for common platforms (sun, ultrix, aix, hpux and others). - TLI support, System V.4 style (Solaris, DG/UX). - Username lookup integrated with the access control language. Selective username lookups are now the default (was: no username lookups). - A safer finger command for booby traps. This one solves a host of possible problems with automatic reverse fingers. Thanks, Borja Marcos (borjam@we.lc.ehu.es) for some inspiring discussions. - KNOWN pattern that matches hosts whose name and address are known. - Cleanup of diagnostics. Errors in access-control files are now shown with file name and line number. - With AIX 3.2, hostnames longer than 32 would be truncated. This caused hostname verification failures, so that service would be refused when paranoid mode was enabled. Found by: Adrian van Bloois (A.vanBloois@info.nic.surfnet.nl). - With some IRIX versions, remote username lookups failed because the fgets() library function does not handle partial read()s from sockets. Found by: Daniel O'Callaghan (danny@austin.unimelb.edu.au). - Added a DISCLAIMER document to help you satisfy legal departments. The extension language module has undergone major revisions and extensions. Thanks, John P. Rouillard (rouilj@ra.cs.umb.edu) for discussions, experiments, and for being a good guinea pig. The extensions are documented in hosts_options.5, and are enabled by editing the Makefile STYLE macro definition. - (Extension language) The ":" separator may now occur within options as long as it is protected with a backslash. A warning is issued when a rule ends on ":". - (Extension language) Better verification mode. When the `try' command is run, each option function now explains what it would do. - (Extension language) New "allow" and "deny" keywords so you can now have all rules within a single file. See "nroff -man hosts_options.5" for examples. - (Extension language) "linger" keyword to set the socket linger time (SO_LINGER). From: Marc Boucher . - (Extension language) "severity" keyword to turn the logging noise up or down. Many sites wanted a means to shut up the program; other sites wanted to emphasize specific events. Adapted from code contributed by Dave Mitchell . Changes per release 5.1 (Mar 1993) ================================== - The additional protection against source-routing attacks from hosts that pretend to have someone elses network address has become optional because it causes kernel panics with SunOS <= 4.1.3. Changes per release 5.0 (Mar 1993) ================================== - Additional protection against source-routing attacks from hosts that pretend to have someone elses network address. For example, the address of a trusted host within your own network. - The access control language has been extended with a simple but powerful operator that greatly simplifies the design of rule sets (ALL: .foo.edu EXCEPT dialup.foo.edu). Blank lines are permitted, and long lines can be continued with backslash-newline. - All configurable stuff, including path names, has been moved into the Makefile so that you no longer have to hack source code to just configure the programs. - Ported to Solaris 2. TLI-based applications not yet supported. Several workarounds for System V bugs. - A small loophole in the netgroup lookup code was closed, and the remote username lookup code was made more portable. - Still more documentation. The README file now provides tutorial sections with introductions to client, server, inetd and syslogd. Changes per release 4.3 (Aug 1992) ================================== - Some sites reported that connections would be rejected because localhost != localhost.domain. The host name checking code now special-cases localhost (problem reported by several sites). - The programs now report an error if an existing access control file cannot be opened (e.g. due to lack of privileges). Until now, the programs would just pretend that the access control file does not exist (reported by Darren Reed, avalon@coombs.anu.edu.au). - The timeout period for remote userid lookups was upped to 30 seconds, in order to cope with slow hosts or networks. If this is too long for you, adjust the TIMEOUT definition in file rfc931.c (problem reported by several sites). - On hosts with more than one IP network interface, remote userid lookups could use the IP address of the "wrong" local interface. The problem and its solution were discussed on the rfc931-users mailing list. Scott Schwartz (schwartz@cs.psu.edu) folded the fix into the rfc931.c module. - The result of % expansion (in shell commands) is now checked for stuff that may confuse the shell; it is replaced by underscores (problem reported by Icarus Sparry, I.Sparry@gdr.bath.ac.uk). - A portability problem was fixed that caused compile-time problems on a CRAY (problem reported by Michael Barnett, mikeb@rmit.edu.au). Changes per release 4.0 (Jun 1992) ================================== 1 - network daemons no longer have to live within a common directory 2 - the access control code now uses both the host address and name 3 - an access control pattern that supports netmasks 4 - additional protection against forged host names 5 - a pattern that matches hosts whose name or address lookup fails 6 - an operator that prevents hosts or services from being matched 7 - optional remote username lookup with the RFC 931 protocol 8 - an optional umask to prevent the creation of world-writable files 9 - hooks for access control language extensions 10 - last but not least, thoroughly revised documentation. Changes per release 3.0 (Oct 1991) ================================== Enhancements over the previous release are: support for datagram (UDP and RPC) services, and execution of shell commands when a (remote host, requested service) pair matches a pattern in the access control tables. Changes per release 2.0 (May 1991) ================================== Enhancements over the previous release are: protection against rlogin and rsh attacks through compromised domain name servers, optional netgroup support for systems with NIS (formerly YP), and an extension of the wild card patterns supported by the access control files. Release 1.0 (Jan 1991) Adapted from code contributed by Dave Mitchell . Changes per release 5.1 (Mar 1993) ================================== - The additional protection against source-routing attacks from hosts that pretend to have someone elses networktcp_wrappers_7.6/try-from.c 444 465 506 4012 5700312617 11161 /* * This program can be called via a remote shell command to find out if the * hostname and address are properly recognized, if username lookup works, * and (SysV only) if the TLI on top of IP heuristics work. * * Example: "rsh host /some/where/try-from". * * Diagnostics are reported through syslog(3) and redirected to stderr. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) try-from.c 1.2 94/12/28 17:42:55"; #endif /* System libraries. */ #include #include #include #include #ifdef TLI #include #include #endif #ifndef STDIN_FILENO #define STDIN_FILENO 0 #endif /* Local stuff. */ #include "tcpd.h" int allow_severity = SEVERITY; /* run-time adjustable */ int deny_severity = LOG_WARNING; /* ditto */ main(argc, argv) int argc; char **argv; { struct request_info request; char buf[BUFSIZ]; char *cp; /* * Simplify the process name, just like tcpd would. */ if ((cp = strrchr(argv[0], '/')) != 0) argv[0] = cp + 1; /* * Turn on the "IP-underneath-TLI" detection heuristics. */ #ifdef TLI if (ioctl(0, I_FIND, "timod") == 0) ioctl(0, I_PUSH, "timod"); #endif /* TLI */ /* * Look up the endpoint information. */ request_init(&request, RQ_DAEMON, argv[0], RQ_FILE, STDIN_FILENO, 0); (void) fromhost(&request); /* * Show some results. Name and address information is looked up when we * ask for it. */ #define EXPAND(str) percent_x(buf, sizeof(buf), str, &request) puts(EXPAND("client address (%%a): %a")); puts(EXPAND("client hostname (%%n): %n")); puts(EXPAND("client username (%%u): %u")); puts(EXPAND("client info (%%c): %c")); puts(EXPAND("server address (%%A): %A")); puts(EXPAND("server hostname (%%N): %N")); puts(EXPAND("server process (%%d): %d")); puts(EXPAND("server info (%%s): %s")); return (0); } the access control file does not exist (reported by Darren Reed, avalon@coombs.anu.edu.au). - The timeout period for remote userid lookups was upped to 30 seconds, in order to cope with slow hosts or networks. If this is too long for you, adjust the TIMEOUT definition in file rfc931.c (problem reported by several sites). - On hosts with more than one IP network interface, remote userid lookups could use the IP address of the "wrong" local interface. The problem and its solution were discussedtcp_wrappers_7.6/update.c 444 465 506 5542 5700312620 10667 /* * Routines for controlled update/initialization of request structures. * * request_init() initializes its argument. Pointers and string-valued members * are initialized to zero, to indicate that no lookup has been attempted. * * request_set() adds information to an already initialized request structure. * * Both functions take a variable-length name-value list. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) update.c 1.1 94/12/28 17:42:56"; #endif /* System libraries */ #include #include #include /* Local stuff. */ #include "mystdarg.h" #include "tcpd.h" /* request_fill - request update engine */ static struct request_info *request_fill(request, ap) struct request_info *request; va_list ap; { int key; char *ptr; while ((key = va_arg(ap, int)) > 0) { switch (key) { default: tcpd_warn("request_fill: invalid key: %d", key); return (request); case RQ_FILE: request->fd = va_arg(ap, int); continue; case RQ_CLIENT_SIN: request->client->sin = va_arg(ap, struct sockaddr_in *); continue; case RQ_SERVER_SIN: request->server->sin = va_arg(ap, struct sockaddr_in *); continue; /* * All other fields are strings with the same maximal length. */ case RQ_DAEMON: ptr = request->daemon; break; case RQ_USER: ptr = request->user; break; case RQ_CLIENT_NAME: ptr = request->client->name; break; case RQ_CLIENT_ADDR: ptr = request->client->addr; break; case RQ_SERVER_NAME: ptr = request->server->name; break; case RQ_SERVER_ADDR: ptr = request->server->addr; break; } STRN_CPY(ptr, va_arg(ap, char *), STRING_LENGTH); } return (request); } /* request_init - initialize request structure */ struct request_info *VARARGS(request_init, struct request_info *, request) { static struct request_info default_info; struct request_info *r; va_list ap; /* * Initialize data members. We do not assign default function pointer * members, to avoid pulling in the whole socket module when it is not * really needed. */ VASTART(ap, struct request_info *, request); *request = default_info; request->fd = -1; strcpy(request->daemon, unknown); sprintf(request->pid, "%d", getpid()); request->client->request = request; request->server->request = request; r = request_fill(request, ap); VAEND(ap); return (r); } /* request_set - update request structure */ struct request_info *VARARGS(request_set, struct request_info *, request) { struct request_info *r; va_list ap; VASTART(ap, struct request_info *, request); r = request_fill(request, ap); VAEND(ap); return (r); } tcp_wrappers_7.6/ptx.c 444 465 506 5226 5700312577 10232 /* * The Dynix/PTX TLI implementation is not quite compatible with System V * Release 4. Some important functions are not present so we are limited to * IP-based services. * * Diagnostics are reported through syslog(3). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) ptx.c 1.3 94/12/28 17:42:38"; #endif #ifdef PTX /* System libraries. */ #include #include #include #include #include #include #include #include /* Local stuff. */ #include "tcpd.h" /* Forward declarations. */ static void ptx_sink(); /* tli_host - determine TLI endpoint info, PTX version */ void tli_host(request) struct request_info *request; { static struct sockaddr_in client; static struct sockaddr_in server; /* * getpeerinaddr() was suggested by someone at Sequent. It seems to work * with connection-oriented (TCP) services such as rlogind and telnetd, * but it returns 0.0.0.0 with datagram (UDP) services. No problem: UDP * needs special treatment anyway, in case we must refuse service. */ if (getpeerinaddr(request->fd, &client, sizeof(client)) == 0 && client.sin_addr.s_addr != 0) { request->client->sin = &client; if (getmyinaddr(request->fd, &server, sizeof(server)) == 0) { request->server->sin = &server; } else { tcpd_warn("warning: getmyinaddr: %m"); } sock_methods(request); } else { /* * Another suggestion was to temporarily switch to the socket * interface, identify the endpoint addresses with socket calls, then * to switch back to TLI. This seems to works OK with UDP services, * which is exactly what we should be looking at right now. */ #define SWAP_MODULE(f, old, new) (ioctl(f, I_POP, old), ioctl(f, I_PUSH, new)) if (SWAP_MODULE(request->fd, "timod", "sockmod") != 0) tcpd_warn("replace timod by sockmod: %m"); sock_host(request); if (SWAP_MODULE(request->fd, "sockmod", "timod") != 0) tcpd_warn("replace sockmod by timod: %m"); if (request->sink != 0) request->sink = ptx_sink; } } /* ptx_sink - absorb unreceived IP datagram */ static void ptx_sink(fd) int fd; { char buf[BUFSIZ]; struct sockaddr sa; int size = sizeof(sa); /* * Eat up the not-yet received datagram. Where needed, switch to the * socket programming interface. */ if (ioctl(fd, I_FIND, "timod") != 0) ioctl(fd, I_POP, "timod"); if (ioctl(fd, I_FIND, "sockmod") == 0) ioctl(fd, I_PUSH, "sockmod"); (void) recvfrom(fd, buf, sizeof(buf), 0, &sa, &size); } #endif /* PTX */ * members, to avoid pulling in the whole socket module when it is not * really needed. */ VASTART(ap, struct request_info *, request); *request = default_info; request->fd = -1; strcpy(request->daemon, unknown); sprintf(request->pid, "%d", getpid()); request->client->request = request; request->server->request = request;tcp_wrappers_7.6/vfprintf.c 444 465 506 6176 5544071000 11247 /* * vfprintf() and vprintf() clones. They will produce unexpected results * when excessive dynamic ("*") field widths are specified. To be used for * testing purposes only. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) vfprintf.c 1.2 94/03/23 17:44:46"; #endif #include #include #ifdef __STDC__ #include #else #include #endif /* vfprintf - print variable-length argument list to stream */ int vfprintf(fp, format, ap) FILE *fp; char *format; va_list ap; { char fmt[BUFSIZ]; /* format specifier */ register char *fmtp; register char *cp; int count = 0; /* * Iterate over characters in the format string, picking up arguments * when format specifiers are found. */ for (cp = format; *cp; cp++) { if (*cp != '%') { putc(*cp, fp); /* ordinary character */ count++; } else { /* * Format specifiers are handled one at a time, since we can only * deal with arguments one at a time. Try to determine the end of * the format specifier. We do not attempt to fully parse format * strings, since we are ging to let fprintf() do the hard work. * In regular expression notation, we recognize: * * %-?0?([0-9]+|\*)?\.?([0-9]+|\*)?l?[a-z] * * which includes some combinations that do not make sense. */ fmtp = fmt; *fmtp++ = *cp++; if (*cp == '-') /* left-adjusted field? */ *fmtp++ = *cp++; if (*cp == '0') /* zero-padded field? */ *fmtp++ = *cp++; if (*cp == '*') { /* dynamic field witdh */ sprintf(fmtp, "%d", va_arg(ap, int)); fmtp += strlen(fmtp); cp++; } else { while (isdigit(*cp)) /* hard-coded field width */ *fmtp++ = *cp++; } if (*cp == '.') /* width/precision separator */ *fmtp++ = *cp++; if (*cp == '*') { /* dynamic precision */ sprintf(fmtp, "%d", va_arg(ap, int)); fmtp += strlen(fmtp); cp++; } else { while (isdigit(*cp)) /* hard-coded precision */ *fmtp++ = *cp++; } if (*cp == 'l') /* long whatever */ *fmtp++ = *cp++; if (*cp == 0) /* premature end, punt */ break; *fmtp++ = *cp; /* type (checked below) */ *fmtp = 0; /* Execute the format string - let fprintf() do the hard work. */ switch (fmtp[-1]) { case 's': /* string-valued argument */ count += fprintf(fp, fmt, va_arg(ap, char *)); break; case 'c': /* integral-valued argument */ case 'd': case 'u': case 'o': case 'x': if (fmtp[-2] == 'l') count += fprintf(fp, fmt, va_arg(ap, long)); else count += fprintf(fp, fmt, va_arg(ap, int)); break; case 'e': /* float-valued argument */ case 'f': case 'g': count += fprintf(fp, fmt, va_arg(ap, double)); break; default: /* anything else */ putc(fmtp[-1], fp); count++; break; } } } return (count); } /* vprintf - print variable-length argument list to stdout */ vprintf(format, ap) char *format; va_list ap; { return (vfprintf(stdout, format, ap)); } e); } #endif /* PTX */ * members, to avoid pulling in the whole socket module when it is not * really needed. */ VASTART(ap, struct request_info *, request); *request = default_info; request->fd = -1; strcpy(request->daemon, unknown); sprintf(request->pid, "%d", getpid()); request->client->request = request; request->server->request = request;tcp_wrappers_7.6/tli-sequent.c 444 465 506 11553 5700312613 11700 /* * Warning - this relies heavily on the TLI implementation in PTX 2.X and will * probably not work under PTX 4. * * Author: Tim Wright, Sequent Computer Systems Ltd., UK. * * Modified slightly to conform to the new internal interfaces - Wietse */ #ifndef lint static char sccsid[] = "@(#) tli-sequent.c 1.1 94/12/28 17:42:51"; #endif #ifdef TLI_SEQUENT /* System libraries. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include extern int errno; extern char *sys_errlist[]; extern int sys_nerr; extern int t_errno; extern char *t_errlist[]; extern int t_nerr; /* Local stuff. */ #include "tcpd.h" #include "tli-sequent.h" /* Forward declarations. */ static char *tli_error(); static void tli_sink(); /* tli_host - determine endpoint info */ int tli_host(request) struct request_info *request; { static struct sockaddr_in client; static struct sockaddr_in server; struct _ti_user *tli_state_ptr; union T_primitives *TSI_prim_ptr; struct strpeek peek; int len; /* * Use DNS and socket routines for name and address conversions. */ sock_methods(request); /* * Find out the client address using getpeerinaddr(). This call is the * TLI equivalent to getpeername() under Dynix/ptx. */ len = sizeof(client); t_sync(request->fd); if (getpeerinaddr(request->fd, &client, len) < 0) { tcpd_warn("can't get client address: %s", tli_error()); return; } request->client->sin = &client; /* Call TLI utility routine to get information on endpoint */ if ((tli_state_ptr = _t_checkfd(request->fd)) == NULL) return; if (tli_state_ptr->ti_servtype == T_CLTS) { /* UDP - may need to get address the hard way */ if (client.sin_addr.s_addr == 0) { /* The UDP endpoint is not connected so we didn't get the */ /* remote address - get it the hard way ! */ /* Look at the control part of the top message on the stream */ /* we don't want to remove it from the stream so we use I_PEEK */ peek.ctlbuf.maxlen = tli_state_ptr->ti_ctlsize; peek.ctlbuf.len = 0; peek.ctlbuf.buf = tli_state_ptr->ti_ctlbuf; /* Don't even look at the data */ peek.databuf.maxlen = -1; peek.databuf.len = 0; peek.databuf.buf = 0; peek.flags = 0; switch (ioctl(request->fd, I_PEEK, &peek)) { case -1: tcpd_warn("can't peek at endpoint: %s", tli_error()); return; case 0: /* No control part - we're hosed */ tcpd_warn("can't get UDP info: %s", tli_error()); return; default: /* FALL THROUGH */ ; } /* Can we even check the PRIM_type ? */ if (peek.ctlbuf.len < sizeof(long)) { tcpd_warn("UDP control info garbage"); return; } TSI_prim_ptr = (union T_primitives *) peek.ctlbuf.buf; if (TSI_prim_ptr->type != T_UNITDATA_IND) { tcpd_warn("wrong type for UDP control info"); return; } /* Validate returned unitdata indication packet */ if ((peek.ctlbuf.len < sizeof(struct T_unitdata_ind)) || ((TSI_prim_ptr->unitdata_ind.OPT_length != 0) && (peek.ctlbuf.len < TSI_prim_ptr->unitdata_ind.OPT_length + TSI_prim_ptr->unitdata_ind.OPT_offset))) { tcpd_warn("UDP control info garbaged"); return; } /* Extract the address */ memcpy(&client, peek.ctlbuf.buf + TSI_prim_ptr->unitdata_ind.SRC_offset, TSI_prim_ptr->unitdata_ind.SRC_length); } request->sink = tli_sink; } if (getmyinaddr(request->fd, &server, len) < 0) tcpd_warn("can't get local address: %s", tli_error()); else request->server->sin = &server; } /* tli_error - convert tli error number to text */ static char *tli_error() { static char buf[40]; if (t_errno != TSYSERR) { if (t_errno < 0 || t_errno >= t_nerr) { sprintf(buf, "Unknown TLI error %d", t_errno); return (buf); } else { return (t_errlist[t_errno]); } } else { if (errno < 0 || errno >= sys_nerr) { sprintf(buf, "Unknown UNIX error %d", errno); return (buf); } else { return (sys_errlist[errno]); } } } /* tli_sink - absorb unreceived datagram */ static void tli_sink(fd) int fd; { struct t_unitdata *unit; int flags; /* * Something went wrong. Absorb the datagram to keep inetd from looping. * Allocate storage for address, control and data. If that fails, sleep * for a couple of seconds in an attempt to keep inetd from looping too * fast. */ if ((unit = (struct t_unitdata *) t_alloc(fd, T_UNITDATA, T_ALL)) == 0) { tcpd_warn("t_alloc: %s", tli_error()); sleep(5); } else { (void) t_rcvudata(fd, unit, &flags); t_free((void *) unit, T_UNITDATA); } } #endif /* TLI_SEQUENT */ tcp_wrappers_7.6/tli-sequent.h 444 465 506 633 5700312614 11643 #ifdef __STDC__ #define __P(X) X #else #define __P(X) () #endif extern int t_sync __P((int)); extern char *t_alloc __P((int, int, int)); extern int t_free __P((char *, int)); extern int t_rcvudata __P((int, struct t_unitdata *, int *)); extern int getpeerinaddr __P((int, struct sockaddr_in *, int)); extern int getmyinaddr __P((int, struct sockaddr_in *, int)); extern struct _ti_user *_t_checkfd __P((int)); d@Ï scaffold.hdx@· tcpdmatch.8@» README.NIScgetpeername() under Dynix/ptx. tcp_wrappers_7.6/misc.c 444 465 506 3115 6107411332 10334 /* * Misc routines that are used by tcpd and by tcpdchk. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsic[] = "@(#) misc.c 1.2 96/02/11 17:01:29"; #endif #include #include #include #include #include #include #include "tcpd.h" extern char *fgets(); #ifndef INADDR_NONE #define INADDR_NONE (-1) /* XXX should be 0xffffffff */ #endif /* xgets - fgets() with backslash-newline stripping */ char *xgets(ptr, len, fp) char *ptr; int len; FILE *fp; { int got; char *start = ptr; while (fgets(ptr, len, fp)) { got = strlen(ptr); if (got >= 1 && ptr[got - 1] == '\n') { tcpd_context.line++; if (got >= 2 && ptr[got - 2] == '\\') { got -= 2; } else { return (start); } } ptr += got; len -= got; ptr[0] = 0; } return (ptr > start ? start : 0); } /* split_at - break string at delimiter or return NULL */ char *split_at(string, delimiter) char *string; int delimiter; { char *cp; if ((cp = strchr(string, delimiter)) != 0) *cp++ = 0; return (cp); } /* dot_quad_addr - convert dotted quad to internal form */ unsigned long dot_quad_addr(str) char *str; { int in_run = 0; int runs = 0; char *cp = str; /* Count the number of runs of non-dot characters. */ while (*cp) { if (*cp == '.') { in_run = 0; } else if (in_run == 0) { in_run = 1; runs++; } cp++; } return (runs == 4 ? inet_addr(str) : INADDR_NONE); } can't peek at endpoint: %s", tli_error()); return; case 0: /* No control part - we're hosed */ tcpd_warn("can't get UDP info: %s", tli_error()); return; default: /* FALL THROUGH */ ; } /* Can we even check the PRIM_type ? */ if (peek.ctlbuf.len < sizeof(long)) { tcpd_warn("UDP control info garbage"); return; } TSI_prim_ptr = (union T_primitives *) peek.ctlbuf.buf; if (TSI_primtcp_wrappers_7.6/diag.c 444 465 506 2672 5700312554 10320 /* * Routines to report various classes of problems. Each report is decorated * with the current context (file name and line number), if available. * * tcpd_warn() reports a problem and proceeds. * * tcpd_jump() reports a problem and jumps. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) diag.c 1.1 94/12/28 17:42:20"; #endif /* System libraries */ #include #include #include /* Local stuff */ #include "tcpd.h" #include "mystdarg.h" struct tcpd_context tcpd_context; jmp_buf tcpd_buf; /* tcpd_diag - centralize error reporter */ static void tcpd_diag(severity, tag, format, ap) int severity; char *tag; char *format; va_list ap; { char fmt[BUFSIZ]; if (tcpd_context.file) sprintf(fmt, "%s: %s, line %d: %s", tag, tcpd_context.file, tcpd_context.line, format); else sprintf(fmt, "%s: %s", tag, format); vsyslog(severity, fmt, ap); } /* tcpd_warn - report problem of some sort and proceed */ void VARARGS(tcpd_warn, char *, format) { va_list ap; VASTART(ap, char *, format); tcpd_diag(LOG_ERR, "warning", format, ap); VAEND(ap); } /* tcpd_jump - report serious problem and jump */ void VARARGS(tcpd_jump, char *, format) { va_list ap; VASTART(ap, char *, format); tcpd_diag(LOG_ERR, "error", format, ap); VAEND(ap); longjmp(tcpd_buf, AC_ERROR); } in_run = 0; } else if (in_run == 0) { in_run = 1; runstcp_wrappers_7.6/ncr.c 444 465 506 5023 5700312572 10167 /* * This part for NCR UNIX with is from Andrew Maffei (arm@aqua.whoi.edu). It * assumes TLI throughout. In order to look up endpoint address information * we must talk to the "timod" streams module. For some reason "timod" wants * to sit directly on top of the device driver. Therefore we pop off all * streams modules except the driver, install the "timod" module so that we * can figure out network addresses, and then restore the original state. */ #ifndef lint static char sccsid[] = "@(#) ncr.c 1.1 94/12/28 17:42:34"; #endif #include #include #include #include #include #include #include "tcpd.h" #define MAX_MODULE_COUNT 10 /* XXX */ /* fromhost - tear down the streams stack then rebuild it */ void fromhost(request) struct request_info *request; { int i; int num_mod; struct str_list str_list; struct str_mlist mod_buffer[MAX_MODULE_COUNT]; int fd = request->fd; str_list.sl_nmods = MAX_MODULE_COUNT; str_list.sl_modlist = &mod_buffer[0]; /* * On systems with WIN streams support we have to be careful about what * is on the stream we are passed. This code POPs off all modules above * the pseudo driver, pushes timod, gets the host address information, * pops timod and then pushes all modules back on the stream. * * Some state may be lost in this process. /usr/etc/tlid seems to do special * things to the stream depending on the TCP port being serviced. (not a * very nice thing to do!). It is unclear what to do if this code breaks * - the stream may be left in an unknown condition. */ if ((num_mod = ioctl(fd, I_LIST, NULL)) < 0) tcpd_warn("fromhost: LIST failed: %m"); if (ioctl(fd, I_LIST, &str_list) < 0) tcpd_warn("fromhost: LIST failed: %m"); /* * POP stream modules except for the driver. */ for (i = 0; i < num_mod - 1; i++) if (ioctl(fd, I_POP, 0) < 0) tcpd_warn("fromhost: POP %s: %m", mod_buffer[i].l_name); /* * PUSH timod so that host address ioctls can be executed. */ if (ioctl(fd, I_PUSH, "timod") < 0) tcpd_warn("fromhost: PUSH timod: %m"); tli_host(request); /* * POP timod, we're done with it now. */ if (ioctl(fd, I_POP, 0) < 0) tcpd_warn("fromhost: POP timod: %m"); /* * Restore stream modules. */ for (i = num_mod - 2; i >= 0; i--) if (ioctl(fd, I_PUSH, mod_buffer[i].l_name) < 0) tcpd_warn("fromhost: PUSH %s: %m", mod_buffer[i].l_name); } 7:42:34"; #endif #include #include #include #include #include #include #include "tcpd.h" #define MAX_MODULE_COUNT 10 /* XXX */ /* fromhost - tear down the streams stack then rebuild it */ void fromhost(request) struct request_info *request; { int i; int num_mod; struct str_list str_list; struct str_mlist mod_buffer[MAX_MODULE_COUNT]; int fd = request->fd; str_list.sl_nmotcp_wrappers_7.6/tcpdchk.c 444 465 506 26167 6300214465 11060 /* * tcpdchk - examine all tcpd access control rules and inetd.conf entries * * Usage: tcpdchk [-a] [-d] [-i inet_conf] [-v] * * -a: complain about implicit "allow" at end of rule. * * -d: rules in current directory. * * -i: location of inetd.conf file. * * -v: show all rules. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) tcpdchk.c 1.8 97/02/12 02:13:25"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #include #include extern int errno; extern void exit(); extern int optind; extern char *optarg; #ifndef INADDR_NONE #define INADDR_NONE (-1) /* XXX should be 0xffffffff */ #endif #ifndef S_ISDIR #define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) #endif /* Application-specific. */ #include "tcpd.h" #include "inetcf.h" #include "scaffold.h" /* * Stolen from hosts_access.c... */ static char sep[] = ", \t\n"; #define BUFLEN 2048 int resident = 0; int hosts_access_verbose = 0; char *hosts_allow_table = HOSTS_ALLOW; char *hosts_deny_table = HOSTS_DENY; extern jmp_buf tcpd_buf; /* * Local stuff. */ static void usage(); static void parse_table(); static void print_list(); static void check_daemon_list(); static void check_client_list(); static void check_daemon(); static void check_user(); static int check_host(); static int reserved_name(); #define PERMIT 1 #define DENY 0 #define YES 1 #define NO 0 static int defl_verdict; static char *myname; static int allow_check; static char *inetcf; int main(argc, argv) int argc; char **argv; { struct request_info request; struct stat st; int c; myname = argv[0]; /* * Parse the JCL. */ while ((c = getopt(argc, argv, "adi:v")) != EOF) { switch (c) { case 'a': allow_check = 1; break; case 'd': hosts_allow_table = "hosts.allow"; hosts_deny_table = "hosts.deny"; break; case 'i': inetcf = optarg; break; case 'v': hosts_access_verbose++; break; default: usage(); /* NOTREACHED */ } } if (argc != optind) usage(); /* * When confusion really strikes... */ if (check_path(REAL_DAEMON_DIR, &st) < 0) { tcpd_warn("REAL_DAEMON_DIR %s: %m", REAL_DAEMON_DIR); } else if (!S_ISDIR(st.st_mode)) { tcpd_warn("REAL_DAEMON_DIR %s is not a directory", REAL_DAEMON_DIR); } /* * Process the inet configuration file (or its moral equivalent). This * information is used later to find references in hosts.allow/deny to * unwrapped services, and other possible problems. */ inetcf = inet_cfg(inetcf); if (hosts_access_verbose) printf("Using network configuration file: %s\n", inetcf); /* * These are not run from inetd but may have built-in access control. */ inet_set("portmap", WR_NOT); inet_set("rpcbind", WR_NOT); /* * Check accessibility of access control files. */ (void) check_path(hosts_allow_table, &st); (void) check_path(hosts_deny_table, &st); /* * Fake up an arbitrary service request. */ request_init(&request, RQ_DAEMON, "daemon_name", RQ_SERVER_NAME, "server_hostname", RQ_SERVER_ADDR, "server_addr", RQ_USER, "user_name", RQ_CLIENT_NAME, "client_hostname", RQ_CLIENT_ADDR, "client_addr", RQ_FILE, 1, 0); /* * Examine all access-control rules. */ defl_verdict = PERMIT; parse_table(hosts_allow_table, &request); defl_verdict = DENY; parse_table(hosts_deny_table, &request); return (0); } /* usage - explain */ static void usage() { fprintf(stderr, "usage: %s [-a] [-d] [-i inet_conf] [-v]\n", myname); fprintf(stderr, " -a: report rules with implicit \"ALLOW\" at end\n"); fprintf(stderr, " -d: use allow/deny files in current directory\n"); fprintf(stderr, " -i: location of inetd.conf file\n"); fprintf(stderr, " -v: list all rules\n"); exit(1); } /* parse_table - like table_match(), but examines _all_ entries */ static void parse_table(table, request) char *table; struct request_info *request; { FILE *fp; int real_verdict; char sv_list[BUFLEN]; /* becomes list of daemons */ char *cl_list; /* becomes list of requests */ char *sh_cmd; /* becomes optional shell command */ char buf[BUFSIZ]; int verdict; struct tcpd_context saved_context; saved_context = tcpd_context; /* stupid compilers */ if (fp = fopen(table, "r")) { tcpd_context.file = table; tcpd_context.line = 0; while (xgets(sv_list, sizeof(sv_list), fp)) { if (sv_list[strlen(sv_list) - 1] != '\n') { tcpd_warn("missing newline or line too long"); continue; } if (sv_list[0] == '#' || sv_list[strspn(sv_list, " \t\r\n")] == 0) continue; if ((cl_list = split_at(sv_list, ':')) == 0) { tcpd_warn("missing \":\" separator"); continue; } sh_cmd = split_at(cl_list, ':'); if (hosts_access_verbose) printf("\n>>> Rule %s line %d:\n", tcpd_context.file, tcpd_context.line); if (hosts_access_verbose) print_list("daemons: ", sv_list); check_daemon_list(sv_list); if (hosts_access_verbose) print_list("clients: ", cl_list); check_client_list(cl_list); #ifdef PROCESS_OPTIONS real_verdict = defl_verdict; if (sh_cmd) { verdict = setjmp(tcpd_buf); if (verdict != 0) { real_verdict = (verdict == AC_PERMIT); } else { dry_run = 1; process_options(sh_cmd, request); if (dry_run == 1 && real_verdict && allow_check) tcpd_warn("implicit \"allow\" at end of rule"); } } else if (defl_verdict && allow_check) { tcpd_warn("implicit \"allow\" at end of rule"); } if (hosts_access_verbose) printf("access: %s\n", real_verdict ? "granted" : "denied"); #else if (sh_cmd) shell_cmd(percent_x(buf, sizeof(buf), sh_cmd, request)); if (hosts_access_verbose) printf("access: %s\n", defl_verdict ? "granted" : "denied"); #endif } (void) fclose(fp); } else if (errno != ENOENT) { tcpd_warn("cannot open %s: %m", table); } tcpd_context = saved_context; } /* print_list - pretty-print a list */ static void print_list(title, list) char *title; char *list; { char buf[BUFLEN]; char *cp; char *next; fputs(title, stdout); strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = next) { fputs(cp, stdout); next = strtok((char *) 0, sep); if (next != 0) fputs(" ", stdout); } fputs("\n", stdout); } /* check_daemon_list - criticize daemon list */ static void check_daemon_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int daemons = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((char *) 0, sep)) { if (STR_EQ(cp, "EXCEPT")) { daemons = 0; } else { daemons++; if ((host = split_at(cp + 1, '@')) != 0 && check_host(host) > 1) { tcpd_warn("host %s has more than one address", host); tcpd_warn("(consider using an address instead)"); } check_daemon(cp); } } if (daemons == 0) tcpd_warn("daemon list is empty or ends in EXCEPT"); } /* check_client_list - criticize client list */ static void check_client_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int clients = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((char *) 0, sep)) { if (STR_EQ(cp, "EXCEPT")) { clients = 0; } else { clients++; if (host = split_at(cp + 1, '@')) { /* user@host */ check_user(cp); check_host(host); } else { check_host(cp); } } } if (clients == 0) tcpd_warn("client list is empty or ends in EXCEPT"); } /* check_daemon - criticize daemon pattern */ static void check_daemon(pat) char *pat; { if (pat[0] == '@') { tcpd_warn("%s: daemon name begins with \"@\"", pat); } else if (pat[0] == '.') { tcpd_warn("%s: daemon name begins with dot", pat); } else if (pat[strlen(pat) - 1] == '.') { tcpd_warn("%s: daemon name ends in dot", pat); } else if (STR_EQ(pat, "ALL") || STR_EQ(pat, unknown)) { /* void */ ; } else if (STR_EQ(pat, "FAIL")) { /* obsolete */ tcpd_warn("FAIL is no longer recognized"); tcpd_warn("(use EXCEPT or DENY instead)"); } else if (reserved_name(pat)) { tcpd_warn("%s: daemon name may be reserved word", pat); } else { switch (inet_get(pat)) { case WR_UNKNOWN: tcpd_warn("%s: no such process name in %s", pat, inetcf); inet_set(pat, WR_YES); /* shut up next time */ break; case WR_NOT: tcpd_warn("%s: service possibly not wrapped", pat); inet_set(pat, WR_YES); break; } } } /* check_user - criticize user pattern */ static void check_user(pat) char *pat; { if (pat[0] == '@') { /* @netgroup */ tcpd_warn("%s: user name begins with \"@\"", pat); } else if (pat[0] == '.') { tcpd_warn("%s: user name begins with dot", pat); } else if (pat[strlen(pat) - 1] == '.') { tcpd_warn("%s: user name ends in dot", pat); } else if (STR_EQ(pat, "ALL") || STR_EQ(pat, unknown) || STR_EQ(pat, "KNOWN")) { /* void */ ; } else if (STR_EQ(pat, "FAIL")) { /* obsolete */ tcpd_warn("FAIL is no longer recognized"); tcpd_warn("(use EXCEPT or DENY instead)"); } else if (reserved_name(pat)) { tcpd_warn("%s: user name may be reserved word", pat); } } /* check_host - criticize host pattern */ static int check_host(pat) char *pat; { char *mask; int addr_count = 1; if (pat[0] == '@') { /* @netgroup */ #ifdef NO_NETGRENT /* SCO has no *netgrent() support */ #else #ifdef NETGROUP char *machinep; char *userp; char *domainp; setnetgrent(pat + 1); if (getnetgrent(&machinep, &userp, &domainp) == 0) tcpd_warn("%s: unknown or empty netgroup", pat + 1); endnetgrent(); #else tcpd_warn("netgroup support disabled"); #endif #endif } else if (mask = split_at(pat, '/')) { /* network/netmask */ if (dot_quad_addr(pat) == INADDR_NONE || dot_quad_addr(mask) == INADDR_NONE) tcpd_warn("%s/%s: bad net/mask pattern", pat, mask); } else if (STR_EQ(pat, "FAIL")) { /* obsolete */ tcpd_warn("FAIL is no longer recognized"); tcpd_warn("(use EXCEPT or DENY instead)"); } else if (reserved_name(pat)) { /* other reserved */ /* void */ ; } else if (NOT_INADDR(pat)) { /* internet name */ if (pat[strlen(pat) - 1] == '.') { tcpd_warn("%s: domain or host name ends in dot", pat); } else if (pat[0] != '.') { addr_count = check_dns(pat); } } else { /* numeric form */ if (STR_EQ(pat, "0.0.0.0") || STR_EQ(pat, "255.255.255.255")) { /* void */ ; } else if (pat[0] == '.') { tcpd_warn("%s: network number begins with dot", pat); } else if (pat[strlen(pat) - 1] != '.') { check_dns(pat); } } return (addr_count); } /* reserved_name - determine if name is reserved */ static int reserved_name(pat) char *pat; { return (STR_EQ(pat, unknown) || STR_EQ(pat, "KNOWN") || STR_EQ(pat, paranoid) || STR_EQ(pat, "ALL") || STR_EQ(pat, "LOCAL")); } tead)"); } check_daemon(cp); } } if (daemons == 0) tcpd_warn("daemon list is empty or ends in EXCEPT"); } /* check_client_list - criticize client list */ static void check_client_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int clients = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((chtcp_wrappers_7.6/percent_m.c 444 465 506 1435 5700312575 11367 /* * Replace %m by system error message. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) percent_m.c 1.1 94/12/28 17:42:37"; #endif #include #include #include extern int errno; #ifndef SYS_ERRLIST_DEFINED extern char *sys_errlist[]; extern int sys_nerr; #endif #include "mystdarg.h" char *percent_m(obuf, ibuf) char *obuf; char *ibuf; { char *bp = obuf; char *cp = ibuf; while (*bp = *cp) if (*cp == '%' && cp[1] == 'm') { if (errno < sys_nerr && errno > 0) { strcpy(bp, sys_errlist[errno]); } else { sprintf(bp, "Unknown error %d", errno); } bp += strlen(bp); cp += 2; } else { bp++, cp++; } return (obuf); } list */ static void check_client_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int clients = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((chtcp_wrappers_7.6/myvsyslog.c 444 465 506 1376 5700312572 11470 /* * vsyslog() for sites without. In order to enable this code, build with * -Dvsyslog=myvsyslog. We use a different name so that no accidents will * happen when vsyslog() exists. On systems with vsyslog(), syslog() is * typically implemented in terms of vsyslog(). * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) myvsyslog.c 1.1 94/12/28 17:42:33"; #endif #ifdef vsyslog #include #include "tcpd.h" #include "mystdarg.h" myvsyslog(severity, format, ap) int severity; char *format; va_list ap; { char fbuf[BUFSIZ]; char obuf[3 * STRING_LENGTH]; vsprintf(obuf, percent_m(fbuf, format), ap); syslog(severity, "%s", obuf); } #endif cp++; } return (obuf); } list */ static void check_client_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int clients = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((chtcp_wrappers_7.6/mystdarg.h 444 465 506 1045 5700312571 11243 /* * What follows is an attempt to unify varargs.h and stdarg.h. I'd rather * have this than #ifdefs all over the code. */ #ifdef __STDC__ #include #define VARARGS(func,type,arg) func(type arg, ...) #define VASTART(ap,type,name) va_start(ap,name) #define VAEND(ap) va_end(ap) #else #include #define VARARGS(func,type,arg) func(va_alist) va_dcl #define VASTART(ap,type,name) {type name; va_start(ap); name = va_arg(ap, type) #define VAEND(ap) va_end(ap);} #endif extern char *percent_m(); format, ap) int severity; char *format; va_list ap; { char fbuf[BUFSIZ]; char obuf[3 * STRING_LENGTH]; vsprintf(obuf, percent_m(fbuf, format), ap); syslog(severity, "%s", obuf); } #endif cp++; } return (obuf); } list */ static void check_client_list(list) char *list; { char buf[BUFLEN]; char *cp; char *host; int clients = 0; strcpy(buf, list); for (cp = strtok(buf, sep); cp != 0; cp = strtok((chtcp_wrappers_7.6/printf.ck 444 465 506 47 5702340353 11023 syslog 1 0 tcpd_warn 0 0 tcpd_jump 0 0 DME.IRIXÌ€„ Banners.Makefile€„à@ßrefuse.c€„@Ä tcpdchk.8€„@âsetenv.c€„(€‡inetcf.c<@Õinetcf.hàP‚0 scaffold.cd@Ï scaffold.hx@· tcpdmatch.8@» README.NIS va_end(ap) #else #include #define VARARGS(func,type,arg) func(va_alist) va_dcl #define VASTART(ap,type,name) {type name; va_start(ap); name = va_arg(ap, type) #define VAEND(ap) va_end(ap)tcp_wrappers_7.6/README.IRIX 444 465 506 5112 5700322127 10667 @(#) README.IRIX 1.2 94/12/28 18:45:58 In the past few months I received several messages with questions from people that tried to use my tcp wrapper on IRIX 5.x. Some mysteries could be solved via email, and then some remained. Today I finally had a chance to do some tests on someones IRIX 5.2 system. Here is my first-hand experience with wrapper release 6.3. (1) Inetd is broken. Normally one edits inetd.conf, sends a HUP signal to inetd and that's it. With IRIX evil things happen: inetd is too stupid to remember that it is already listening on a port. In order to modify an entry in inetd.conf, first comment it out with a # at the beginning of the line, kill -HUP the inetd, then uncomment the inetd.conf entry and kill -HUP again. Even with this amount of care I have seen inetd messing up, like calling rusersd when I make a talk connection. Even killing and restarting inetd does not solve all problems. I find it hard to believe, it but the best thing to do with IRIX is to reboot after changing inetd.conf. (2) When tcpd is built according to the irix4 Makefile rules, it appears to work as expected with TCP-based services such as fingerd, and with UDP-based services such as ntalk and tftp. (3) It does NOT work with RPC over UDP services such as rusersd and rstatd: the wrapper hangs in the recvfrom() system call, and I have spent several hours looking for ways to work around it. No way. After finding that none of the applicable socket primitives can be made to work (recvfrom recvmsg) I give up. So, the IRIX RPC services cannot be wrapped until SGI fixes their system so that it works like everyone elses code (HP Sun Dec AIX and so on). (4) I didn't even bother to try the RPC over TCP services. (5) When an IRIX 5.2 system is a NIS client, it can have problems with hosts that have more than one address: the wrapper will see only one address, and may complain when PARANOID mode is on. The fix is to change the name service lookup order in /etc/resolv.conf so that your system tries DNS before NIS (hostresorder bind nis local). (6) IRIX 5.2 is not System V.4, and it shows. Do not link with the -lsocket and -lnsl libraries. They are completely broken, and the wrapper will be unable to figure out the client internet address. So, TLI services cannot be wrapped until SGI fixes their system so that it works the way it is supposed to. I am not impressed by the quality of the IRIX system software. There are many things that work on almost every other system except with IRIX. Wietse strlen(pat) - 1] == '.') { tcpd_warn("%s: daemon name ends in dot", pat); } else if (STR_EQ(pat, "ALL") || STR_EQ(pat, unknown)) { /* void */ ; } else if (STR_EQ(pat, "FAIL")) { /* obsolete */ tcpd_warn("FAIL is no longer recognized"); tcpd_warn("(use EXCEPT or DENY instead)"); } else if (reserved_name(pat)) { tcpd_warn("%s: daemon name may be reserved word", pat); } else { switch (inet_get(pat)) { case WR_UNKtcp_wrappers_7.6/Banners.Makefile 444 465 506 4261 6300214456 12272 # @(#) Banners.Makefile 1.3 97/02/12 02:13:18 # # Install this file as the Makefile in your directory with banner files. # It will convert a prototype banner text to a form that is suitable for # the ftp, telnet, rlogin, and other services. # # You'll have to comment out the IN definition below if your daemon # names don't start with `in.'. # # The prototype text should live in the banners directory, as a file with # the name "prototype". In the prototype text you can use % # sequences as described in the hosts_access.5 manual page (`nroff -man' # format). The sequences will be expanded while the banner message is # sent to the client. For example: # # Hello %u@%h, what brings you here? # # Expands to: Hello username@hostname, what brings you here? Note: the # use of %u forces a client username lookup. # # In order to use banners, build the tcp wrapper with -DPROCESS_OPTIONS # and use hosts.allow rules like this: # # daemons ... : clients ... : banners /some/directory ... # # Of course, nothing prevents you from using multiple banner directories. # For example, one banner directory for clients that are granted service, # one banner directory for rejected clients, and one banner directory for # clients with a hostname problem. # SHELL = /bin/sh IN = in. BANNERS = $(IN)telnetd $(IN)ftpd $(IN)rlogind # $(IN)fingerd $(IN)rshd all: $(BANNERS) $(IN)telnetd: prototype cp prototype $@ chmod 644 $@ $(IN)ftpd: prototype sed 's/^/220-/' prototype > $@ chmod 644 $@ $(IN)rlogind: prototype nul ( ./nul ; cat prototype ) > $@ chmod 644 $@ # Other services: banners may interfere with normal operation # so they should probably be used only when refusing service. # In particular, banners don't work with standard rsh daemons. # You would have to use an rshd that has built-in tcp wrapper # support, for example the rshd that is part of the logdaemon # utilities. $(IN)fingerd: prototype cp prototype $@ chmod 644 $@ $(IN)rshd: prototype nul ( ./nul ; cat prototype ) > $@ chmod 644 $@ # In case no /dev/zero available, let's hope they have at least # a C compiler of some sort. nul: echo 'main() { write(1,"",1); return(0); }' >nul.c $(CC) $(CFLAGS) -s -o nul nul.c rm -f nul.c hat brings you here? # # Expands to: Hello username@hostname, what brings you here? Note: the # use of %u forces a client username lookup. # # In order to use banners, build the tcp wrapper with -DPROCESS_OPTIONS # and use hosts.allow rules like this: # # daemons ... : clients ... : banners /some/directory ... # # Of course, nothing tcp_wrappers_7.6/refuse.c 444 465 506 1356 5700312600 10673 /* * refuse() reports a refused connection, and takes the consequences: in * case of a datagram-oriented service, the unread datagram is taken from * the input queue (or inetd would see the same datagram again and again); * the program is terminated. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) refuse.c 1.5 94/12/28 17:42:39"; #endif /* System libraries. */ #include #include /* Local stuff. */ #include "tcpd.h" /* refuse - refuse request */ void refuse(request) struct request_info *request; { syslog(deny_severity, "refused connect from %s", eval_client(request)); clean_exit(request); /* NOTREACHED */ } what brings you here? Note: the # use of %u forces a client username lookup. # # In order to use banners, build the tcp wrapper with -DPROCESS_OPTIONS # and use hosts.allow rules like this: # # daemons ... : clients ... : banners /some/directory ... # # Of course, nothing tcp_wrappers_7.6/tcpdchk.8 444 465 506 5012 5704006037 10750 .TH TCPDCHK 8 .SH NAME tcpdchk \- tcp wrapper configuration checker .SH SYNOPSYS tcpdchk [-a] [-d] [-i inet_conf] [-v] .SH DESCRIPTION .PP \fItcpdchk\fR examines your tcp wrapper configuration and reports all potential and real problems it can find. The program examines the \fItcpd\fR access control files (by default, these are \fI/etc/hosts.allow\fR and \fI/etc/hosts.deny\fR), and compares the entries in these files against entries in the \fIinetd\fR or \fItlid\fR network configuration files. .PP \fItcpdchk\fR reports problems such as non-existent pathnames; services that appear in \fItcpd\fR access control rules, but are not controlled by \fItcpd\fR; services that should not be wrapped; non-existent host names or non-internet address forms; occurrences of host aliases instead of official host names; hosts with a name/address conflict; inappropriate use of wildcard patterns; inappropriate use of NIS netgroups or references to non-existent NIS netgroups; references to non-existent options; invalid arguments to options; and so on. .PP Where possible, \fItcpdchk\fR provides a helpful suggestion to fix the problem. .SH OPTIONS .IP -a Report access control rules that permit access without an explicit ALLOW keyword. This applies only when the extended access control language is enabled (build with -DPROCESS_OPTIONS). .IP -d Examine \fIhosts.allow\fR and \fIhosts.deny\fR files in the current directory instead of the default ones. .IP "-i inet_conf" Specify this option when \fItcpdchk\fR is unable to find your \fIinetd.conf\fR or \fItlid.conf\fR network configuration file, or when you suspect that the program uses the wrong one. .IP -v Display the contents of each access control rule. Daemon lists, client lists, shell commands and options are shown in a pretty-printed format; this makes it easier for you to spot any discrepancies between what you want and what the program understands. .SH FILES .PP The default locations of the \fItcpd\fR access control tables are: .PP /etc/hosts.allow .br /etc/hosts.deny .SH SEE ALSO .na .nf tcpdmatch(8), explain what tcpd would do in specific cases. hosts_access(5), format of the tcpd access control tables. hosts_options(5), format of the language extensions. inetd.conf(5), format of the inetd control file. tlid.conf(5), format of the tlid control file. .SH AUTHORS .na .nf Wietse Venema (wietse@wzv.win.tue.nl), Department of Mathematics and Computing Science, Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) tcpdchk.8 1.3 95/01/08 17:00:30 at work on almost every other system except with IRIX. Wietse strlen(pat) - 1] == '.') { tcpd_warn("%s: daemon name ends in dot", pat); } else if (STR_EQ(pat, "ALL") || STR_EQ(pat, unknown)) { /* void */ ; } else if (STR_EQ(pat, "FAIL")) { /* obsolete */ tcpd_warn("FAIL is no longer recognized"); tcpd_warn("(use EXCEPT or DENY instead)"); } else if (reserved_name(pat)) { tcpd_warn("%s: daemon name may be reserved word", pat); } else { switch (inet_get(pat)) { case WR_UNKtcp_wrappers_7.6/setenv.c 444 465 506 1643 5700312603 10710 /* * Some systems do not have setenv(). This one is modeled after 4.4 BSD, but * is implemented in terms of portable primitives only: getenv(), putenv() * and malloc(). It should therefore be safe to use on every UNIX system. * * If clobber == 0, do not overwrite an existing variable. * * Returns nonzero if memory allocation fails. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) setenv.c 1.1 93/03/07 22:47:58"; #endif /* setenv - update or insert environment (name,value) pair */ int setenv(name, value, clobber) char *name; char *value; int clobber; { char *malloc(); char *getenv(); char *cp; if (clobber == 0 && getenv(name) != 0) return (0); if ((cp = malloc(strlen(name) + strlen(value) + 2)) == 0) return (1); sprintf(cp, "%s=%s", name, value); return (putenv(cp)); } \fItcpd\fR access control tables are: .PP /etc/hosts.allow .br /etc/hosts.deny .SH SEE ALSO .tcp_wrappers_7.6/inetcf.c 444 465 506 17575 6300214464 10712 /* * Routines to parse an inetd.conf or tlid.conf file. This would be a great * job for a PERL script. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccsid[] = "@(#) inetcf.c 1.7 97/02/12 02:13:23"; #endif #include #include #include #include #include extern int errno; extern void exit(); #include "tcpd.h" #include "inetcf.h" /* * Network configuration files may live in unusual places. Here are some * guesses. Shorter names follow longer ones. */ char *inet_files[] = { "/private/etc/inetd.conf", /* NEXT */ "/etc/inet/inetd.conf", /* SYSV4 */ "/usr/etc/inetd.conf", /* IRIX?? */ "/etc/inetd.conf", /* BSD */ "/etc/net/tlid.conf", /* SYSV4?? */ "/etc/saf/tlid.conf", /* SYSV4?? */ "/etc/tlid.conf", /* SYSV4?? */ 0, }; static void inet_chk(); static char *base_name(); /* * Structure with everything we know about a service. */ struct inet_ent { struct inet_ent *next; int type; char name[1]; }; static struct inet_ent *inet_list = 0; static char whitespace[] = " \t\r\n"; /* inet_conf - read in and examine inetd.conf (or tlid.conf) entries */ char *inet_cfg(conf) char *conf; { char buf[BUFSIZ]; FILE *fp; char *service; char *protocol; char *user; char *path; char *arg0; char *arg1; struct tcpd_context saved_context; char *percent_m(); int i; struct stat st; saved_context = tcpd_context; /* * The inetd.conf (or tlid.conf) information is so useful that we insist * on its availability. When no file is given run a series of educated * guesses. */ if (conf != 0) { if ((fp = fopen(conf, "r")) == 0) { fprintf(stderr, percent_m(buf, "open %s: %m\n"), conf); exit(1); } } else { for (i = 0; inet_files[i] && (fp = fopen(inet_files[i], "r")) == 0; i++) /* void */ ; if (fp == 0) { fprintf(stderr, "Cannot find your inetd.conf or tlid.conf file.\n"); fprintf(stderr, "Please specify its location.\n"); exit(1); } conf = inet_files[i]; check_path(conf, &st); } /* * Process the file. After the 7.0 wrapper release it became clear that * there are many more inetd.conf formats than the 8 systems that I had * studied. EP/IX uses a two-line specification for rpc services; HP-UX * permits long lines to be broken with backslash-newline. */ tcpd_context.file = conf; tcpd_context.line = 0; while (xgets(buf, sizeof(buf), fp)) { service = strtok(buf, whitespace); /* service */ if (service == 0 || *service == '#') continue; if (STR_NE(service, "stream") && STR_NE(service, "dgram")) strtok((char *) 0, whitespace); /* endpoint */ protocol = strtok((char *) 0, whitespace); (void) strtok((char *) 0, whitespace); /* wait */ if ((user = strtok((char *) 0, whitespace)) == 0) continue; if (user[0] == '/') { /* user */ path = user; } else { /* path */ if ((path = strtok((char *) 0, whitespace)) == 0) continue; } if (path[0] == '?') /* IRIX optional service */ path++; if (STR_EQ(path, "internal")) continue; if (path[strspn(path, "-0123456789")] == 0) { /* * ConvexOS puts RPC version numbers before path names. Jukka * Ukkonen . */ if ((path = strtok((char *) 0, whitespace)) == 0) continue; } if ((arg0 = strtok((char *) 0, whitespace)) == 0) { tcpd_warn("incomplete line"); continue; } if (arg0[strspn(arg0, "0123456789")] == 0) { /* * We're reading a tlid.conf file, the format is: * * ...stuff... path arg_count arguments mod_count modules */ if ((arg0 = strtok((char *) 0, whitespace)) == 0) { tcpd_warn("incomplete line"); continue; } } if ((arg1 = strtok((char *) 0, whitespace)) == 0) arg1 = ""; inet_chk(protocol, path, arg0, arg1); } fclose(fp); tcpd_context = saved_context; return (conf); } /* inet_chk - examine one inetd.conf (tlid.conf?) entry */ static void inet_chk(protocol, path, arg0, arg1) char *protocol; char *path; char *arg0; char *arg1; { char daemon[BUFSIZ]; struct stat st; int wrap_status = WR_MAYBE; char *base_name_path = base_name(path); char *tcpd_proc_name = (arg0[0] == '/' ? base_name(arg0) : arg0); /* * Always warn when the executable does not exist or when it is not * executable. */ if (check_path(path, &st) < 0) { tcpd_warn("%s: not found: %m", path); } else if ((st.st_mode & 0100) == 0) { tcpd_warn("%s: not executable", path); } /* * Cheat on the miscd tests, nobody uses it anymore. */ if (STR_EQ(base_name_path, "miscd")) { inet_set(arg0, WR_YES); return; } /* * While we are here... */ if (STR_EQ(tcpd_proc_name, "rexd") || STR_EQ(tcpd_proc_name, "rpc.rexd")) tcpd_warn("%s may be an insecure service", tcpd_proc_name); /* * The tcpd program gets most of the attention. */ if (STR_EQ(base_name_path, "tcpd")) { if (STR_EQ(tcpd_proc_name, "tcpd")) tcpd_warn("%s is recursively calling itself", tcpd_proc_name); wrap_status = WR_YES; /* * Check: some sites install the wrapper set-uid. */ if ((st.st_mode & 06000) != 0) tcpd_warn("%s: file is set-uid or set-gid", path); /* * Check: some sites insert tcpd in inetd.conf, instead of replacing * the daemon pathname. */ if (arg0[0] == '/' && STR_EQ(tcpd_proc_name, base_name(arg1))) tcpd_warn("%s inserted before %s", path, arg0); /* * Check: make sure files exist and are executable. On some systems * the network daemons are set-uid so we cannot complain. Note that * tcpd takes the basename only in case of absolute pathnames. */ if (arg0[0] == '/') { /* absolute path */ if (check_path(arg0, &st) < 0) { tcpd_warn("%s: not found: %m", arg0); } else if ((st.st_mode & 0100) == 0) { tcpd_warn("%s: not executable", arg0); } } else { /* look in REAL_DAEMON_DIR */ sprintf(daemon, "%s/%s", REAL_DAEMON_DIR, arg0); if (check_path(daemon, &st) < 0) { tcpd_warn("%s: not found in %s: %m", arg0, REAL_DAEMON_DIR); } else if ((st.st_mode & 0100) == 0) { tcpd_warn("%s: not executable", daemon); } } } else { /* * No tcpd program found. Perhaps they used the "simple installation" * recipe. Look for a file with the same basename in REAL_DAEMON_DIR. * Draw some conservative conclusions when a distinct file is found. */ sprintf(daemon, "%s/%s", REAL_DAEMON_DIR, arg0); if (STR_EQ(path, daemon)) { wrap_status = WR_NOT; } else if (check_path(daemon, &st) >= 0) { wrap_status = WR_MAYBE; } else if (errno == ENOENT) { wrap_status = WR_NOT; } else { tcpd_warn("%s: file lookup: %m", daemon); wrap_status = WR_MAYBE; } } /* * Alas, we cannot wrap rpc/tcp services. */ if (wrap_status == WR_YES && STR_EQ(protocol, "rpc/tcp")) tcpd_warn("%s: cannot wrap rpc/tcp services", tcpd_proc_name); inet_set(tcpd_proc_name, wrap_status); } /* inet_set - remember service status */ void inet_set(name, type) char *name; int type; { struct inet_ent *ip = (struct inet_ent *) malloc(sizeof(struct inet_ent) + strlen(name)); if (ip == 0) { fprintf(stderr, "out of memory\n"); exit(1); } ip->next = inet_list; strcpy(ip->name, name); ip->type = type; inet_list = ip; } /* inet_get - look up service status */ int inet_get(name) char *name; { struct inet_ent *ip; if (inet_list == 0) return (WR_MAYBE); for (ip = inet_list; ip; ip = ip->next) if (STR_EQ(ip->name, name)) return (ip->type); return (-1); } /* base_name - compute last pathname component */ static char *base_name(path) char *path; { char *cp; if ((cp = strrchr(path, '/')) != 0) path = cp + 1; return (path); } == 0) arg1 = ""; inet_chk(protocol, path, arg0, arg1); } fclose(fp); tcpd_context = saved_context; return (ctcp_wrappers_7.6/inetcf.h 444 465 506 737 5700312566 10654 /* * @(#) inetcf.h 1.1 94/12/28 17:42:30 * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ extern char *inet_cfg(); /* read inetd.conf file */ extern void inet_set(); /* remember internet service */ extern int inet_get(); /* look up internet service */ #define WR_UNKNOWN (-1) /* service unknown */ #define WR_NOT 1 /* may not be wrapped */ #define WR_MAYBE 2 /* may be wrapped */ #define WR_YES 3 /* service is wrapped */ . */ if (check_path(pathtcp_wrappers_7.6/scaffold.c 444 465 506 12254 6314551414 11214 /* * Routines for testing only. Not really industrial strength. * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ #ifndef lint static char sccs_id[] = "@(#) scaffold.c 1.6 97/03/21 19:27:24"; #endif /* System libraries. */ #include #include #include #include #include #include #include #include #include #include #ifndef INADDR_NONE #define INADDR_NONE (-1) /* XXX should be 0xffffffff */ #endif extern char *malloc(); /* Application-specific. */ #include "tcpd.h" #include "scaffold.h" /* * These are referenced by the options module and by rfc931.c. */ int allow_severity = SEVERITY; int deny_severity = LOG_WARNING; int rfc931_timeout = RFC931_TIMEOUT; /* dup_hostent - create hostent in one memory block */ static struct hostent *dup_hostent(hp) struct hostent *hp; { struct hostent_block { struct hostent host; char *addr_list[1]; }; struct hostent_block *hb; int count; char *data; char *addr; for (count = 0; hp->h_addr_list[count] != 0; count++) /* void */ ; if ((hb = (struct hostent_block *) malloc(sizeof(struct hostent_block) + (hp->h_length + sizeof(char *)) * count)) == 0) { fprintf(stderr, "Sorry, out of memory\n"); exit(1); } memset((char *) &hb->host, 0, sizeof(hb->host)); hb->host.h_length = hp->h_length; hb->host.h_addr_list = hb->addr_list; hb->host.h_addr_list[count] = 0; data = (char *) (hb->host.h_addr_list + count + 1); for (count = 0; (addr = hp->h_addr_list[count]) != 0; count++) { hb->host.h_addr_list[count] = data + hp->h_length * count; memcpy(hb->host.h_addr_list[count], addr, hp->h_length); } return (&hb->host); } /* find_inet_addr - find all addresses for this host, result to free() */ struct hostent *find_inet_addr(host) char *host; { struct in_addr addr; struct hostent *hp; static struct hostent h; static char *addr_list[2]; /* * Host address: translate it to internal form. */ if ((addr.s_addr = dot_quad_addr(host)) != INADDR_NONE) { h.h_addr_list = addr_list; h.h_addr_list[0] = (char *) &addr; h.h_length = sizeof(addr); return (dup_hostent(&h)); } /* * Map host name to a series of addresses. Watch out for non-internet * forms or aliases. The NOT_INADDR() is here in case gethostbyname() has * been "enhanced" to accept numeric addresses. Make a copy of the * address list so that later gethostbyXXX() calls will not clobber it. */ if (NOT_INADDR(host) == 0) { tcpd_warn("%s: not an internet address", host); return (0); } if ((hp = gethostbyname(host)) == 0) { tcpd_warn("%s: host not found", host); return (0); } if (hp->h_addrtype != AF_INET) { tcpd_warn("%d: not an internet host", hp->h_addrtype); return (0); } if (STR_NE(host, hp->h_name)) { tcpd_warn("%s: hostname alias", host); tcpd_warn("(official name: %.*s)", STRING_LENGTH, hp->h_name); } return (dup_hostent(hp)); } /* check_dns - give each address thorough workout, return address count */ int check_dns(host) char *host; { struct request_info request; struct sockaddr_in sin; struct hostent *hp; int count; char *addr; if ((hp = find_inet_addr(host)) == 0) return (0); request_init(&request, RQ_CLIENT_SIN, &sin, 0); sock_methods(&request); memset((char *) &sin, 0, sizeof(sin)); sin.sin_family = AF_INET; for (count = 0; (addr = hp->h_addr_list[count]) != 0; count++) { memcpy((char *) &sin.sin_addr, addr, sizeof(sin.sin_addr)); /* * Force host name and address conversions. Use the request structure * as a cache. Detect hostname lookup problems. Any name/name or * name/address conflicts will be reported while eval_hostname() does * its job. */ request_set(&request, RQ_CLIENT_ADDR, "", RQ_CLIENT_NAME, "", 0); if (STR_EQ(eval_hostname(request.client), unknown)) tcpd_warn("host address %s->name lookup failed", eval_hostaddr(request.client)); } free((char *) hp); return (count); } /* dummy function to intercept the real shell_cmd() */ /* ARGSUSED */ void shell_cmd(command) char *command; { if (hosts_access_verbose) printf("command: %s", command); } /* dummy function to intercept the real clean_exit() */ /* ARGSUSED */ void clean_exit(request) struct request_info *request; { exit(0); } /* dummy function to intercept the real rfc931() */ /* ARGSUSED */ void rfc931(request) struct request_info *request; { strcpy(request->user, unknown); } /* check_path - examine accessibility */ int check_path(path, st) char *path; struct stat *st; { struct stat stbuf; char buf[BUFSIZ]; if (stat(path, st) < 0) return (-1); #ifdef notdef if (st->st_uid != 0) tcpd_warn("%s: not owned by root", path); if (st->st_mode & 020) tcpd_warn("%s: group writable", path); #endif if (st->st_mode & 002) tcpd_warn("%s: world writable", path); if (path[0] == '/' && path[1] != 0) { strrchr(strcpy(buf, path), '/')[0] = 0; (void) check_path(buf[0] ? buf : "/", &stbuf); } return (0); } tus = WR_MAYBE; } else if (errno == ENOENT) { wrap_status = WR_NOT; } else { tcpd_warn("%s: file lookup: %m", daemon); wrap_status = WR_MAYBE; } } /* * Alas, we cannot wrap rpc/tcp services. */ if (wrap_status == WR_YES && STR_EQ(protocol, "rpc/tcp")) tcpd_warn("%s: cannot wrap rpc/tcp services", tcp_wrappers_7.6/scaffold.h 444 465 506 343 5701311230 11142 /* * @(#) scaffold.h 1.3 94/12/31 18:19:19 * * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. */ extern struct hostent *find_inet_addr(); extern int check_dns(); extern int check_path(); == 0) { tcpd_warn("%s: host not found", host); return (0); } if (hp->h_addrtype != AF_INET) { tcpd_warn("%d: not an internet host", hp->h_addrtype); return (0); } if (STR_NE(host, hp->h_name)) { tcpd_warn("%s: hostname alias", host); tcpd_warn("(official name: %.tcp_wrappers_7.6/tcpdmatch.8 444 465 506 6264 6107411340 11304 .TH TCPDMATCH 8 .SH NAME tcpdmatch \- tcp wrapper oracle .SH SYNOPSYS tcpdmatch [-d] [-i inet_conf] daemon client .sp tcpdmatch [-d] [-i inet_conf] daemon[@server] [user@]client .SH DESCRIPTION .PP \fItcpdmatch\fR predicts how the tcp wrapper would handle a specific request for service. Examples are given below. .PP The program examines the \fItcpd\fR access control tables (default \fI/etc/hosts.allow\fR and \fI/etc/hosts.deny\fR) and prints its conclusion. For maximal accuracy, it extracts additional information from your \fIinetd\fR or \fItlid\fR network configuration file. .PP When \fItcpdmatch\fR finds a match in the access control tables, it identifies the matched rule. In addition, it displays the optional shell commands or options in a pretty-printed format; this makes it easier for you to spot any discrepancies between what you want and what the program understands. .SH ARGUMENTS The following two arguments are always required: .IP daemon A daemon process name. Typically, the last component of a daemon executable pathname. .IP client A host name or network address, or one of the `unknown' or `paranoid' wildcard patterns. .sp When a client host name is specified, \fItcpdmatch\fR gives a prediction for each address listed for that client. .sp When a client address is specified, \fItcpdmatch\fR predicts what \fItcpd\fR would do when client name lookup fails. .PP Optional information specified with the \fIdaemon@server\fR form: .IP server A host name or network address, or one of the `unknown' or `paranoid' wildcard patterns. The default server name is `unknown'. .PP Optional information specified with the \fIuser@client\fR form: .IP user A client user identifier. Typically, a login name or a numeric userid. The default user name is `unknown'. .SH OPTIONS .IP -d Examine \fIhosts.allow\fR and \fIhosts.deny\fR files in the current directory instead of the default ones. .IP "-i inet_conf" Specify this option when \fItcpdmatch\fR is unable to find your \fIinetd.conf\fR or \fItlid.conf\fR network configuration file, or when you suspect that the program uses the wrong one. .SH EXAMPLES To predict how \fItcpd\fR would handle a telnet request from the local system: .sp .ti +5 tcpdmatch in.telnetd localhost .PP The same request, pretending that hostname lookup failed: .sp .ti +5 tcpdmatch in.telnetd 127.0.0.1 .PP To predict what tcpd would do when the client name does not match the client address: .sp .ti +5 tcpdmatch in.telnetd paranoid .PP On some systems, daemon names have no `in.' prefix, or \fItcpdmatch\fR may need some help to locate the inetd configuration file. .SH FILES .PP The default locations of the \fItcpd\fR access control tables are: .PP /etc/hosts.allow .br /etc/hosts.deny .SH SEE ALSO .na .nf tcpdchk(8), tcpd configuration checker hosts_access(5), format of the tcpd access control tables. hosts_options(5), format of the language extensions. inetd.conf(5), format of the inetd control file. tlid.conf(5), format of the tlid control file. .SH AUTHORS .na .nf Wietse Venema (wietse@wzv.win.tue.nl), Department of Mathematics and Computing Science, Eindhoven University of Technology Den Dolech 2, P.O. Box 513, 5600 MB Eindhoven, The Netherlands \" @(#) tcpdmatch.8 1.5 96/02/11 17:01:35 _ent *ip = (struct inet_ent *) malloc(sizeof(struct inet_ent) + strlen(name)); if (ip == 0) { fprintf(stderr, "out of memory\n"); exit(1); } ip->next = inet_list; strcpy(ip->name, name); ip->type = type; inet_list = ip; } /* inet_get - look up service status */ int inet_get(name) char *name; tcp_wrappers_7.6/README.NIS 444 465 506 15030 6107414125 10567 @(#) README.NIS 1.2 96/02/11 17:24:52 > Problem: I have several [machines] with multiple IP addresses, and > when they try to connect to a daemon with tcp wrapper, they are often > rejected. I assume this is due to the -DPARANOID option, and depends > on which IP address is returned first from the nameserver for a given > name. This behavior seems to be random, may depend on ordering in > the YP host map? [Note: the situation described below no longer exists. Presently, my internet gateway uses the same IP address on all interfaces. To avoid confusion I have removed the old name wzv-gw.win.tue.nl from the DNS. I have kept the discussion below for educational reasons]. NIS was not designed to handle multi-homed hosts. With NIS, each address should have its own hostname. For example, wzv-gw is my gateway. It has two interfaces: one connected to the local ethernet, the other to a serial link. In the NIS it is registered as: 131.155.210.23 wzv-gw-ether 131.155.12.78 wzv-gw-slip In principle, wzv-gw could be the official name of one of these interfaces, or it could be an alias for both. The DNS was designed to handle multi-homed hosts. In the DNS my gateway is registered in zone win.tue.nl, with one name that has two A records: wzv-gw IN A 131.155.210.23 IN A 131.155.12.78 And of course there are PTR records in zones 210.155.131.in-addr.arpa and 12.155.131.in-addr.arpa that point to wzv-gw.win.tue.nl. This setup does not cause any problems. You can test your name service with the two programs below. This is what they say on a local NIS client (both client and server running SunOS 4.1.3_U1): % gethostbyname wzv-gw Hostname: wzv-gw.win.tue.nl Aliases: Addresses: 131.155.210.23 131.155.12.78 % gethostbyaddr 131.155.210.23 Hostname: wzv-gw-ether Aliases: Addresses: 131.155.210.23 % gethostbyaddr 131.155.12.78 Hostname: wzv-gw-slip Aliases: Addresses: 131.155.12.78 Things seem less confusing when seen by a NIS client in a different domain (both client and server running SunOS 4.1.3_U1): % gethostbyname wzv-gw.win.tue.nl Hostname: wzv-gw.win.tue.nl Aliases: Addresses: 131.155.210.23 131.155.12.78 % gethostbyaddr 131.155.210.23 Hostname: wzv-gw.win.tue.nl Aliases: Addresses: 131.155.12.78 131.155.210.23 % gethostbyaddr 131.155.12.78 Hostname: wzv-gw.win.tue.nl Aliases: Addresses: 131.155.210.23 131.155.12.78 Alas, Solaris 2.4 still has problems. This is what I get on a Solaris 2.4 NIS client, with a SunOS 4.1.3_U1 NIS server: % gethostbyname wzv-gw.win.tue.nl Hostname: wzv-gw.win.tue.nl Aliases: 131.155.210.23 wzv-gw.win.tue.nl Addresses: 131.155.12.78 The tcpd source comes with a workaround for this problem. The workaround is ugly and is not part of the programs attached below. #! /bin/sh # This is a shell archive. Remove anything before this line, then unpack # it by saving it into a file and typing "sh file". To overwrite existing # files, type "sh file -c". You can also feed this as standard input via # unshar, or by typing "sh gethostbyaddr.c <<'END_OF_gethostbyaddr.c' X /* X * gethostbyaddr tester. compile with: X * X * cc -o gethostbyaddr gethostbyaddr.c (SunOS 4.x) X * X * cc -o gethostbyaddr gethostbyaddr.c -lnsl (SunOS 5.x) X * X * run as: gethostbyaddr address X * X * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. X */ X X#include X#include X#include X#include X#include X#include X Xmain(argc, argv) Xint argc; Xchar **argv; X{ X struct hostent *hp; X long addr; X X if (argc != 2) { X fprintf(stderr, "usage: %s i.p.addres\n", argv[0]); X exit(1); X } X addr = inet_addr(argv[1]); X if (hp = gethostbyaddr((char *) &addr, sizeof(addr), AF_INET)) { X printf("Hostname:\t%s\n", hp->h_name); X printf("Aliases:\t"); X while (hp->h_aliases[0]) X printf("%s ", *hp->h_aliases++); X printf("\n"); X printf("Addresses:\t"); X while (hp->h_addr_list[0]) X printf("%s ", inet_ntoa(*(struct in_addr *) * hp->h_addr_list++)); X printf("\n"); X exit(0); X } X fprintf(stderr, "host %s not found\n", argv[1]); X exit(1); X} END_OF_gethostbyaddr.c if test 1073 -ne `wc -c gethostbyname.c <<'END_OF_gethostbyname.c' X /* X * gethostbyname tester. compile with: X * X * cc -o gethostbyname gethostbyname.c (SunOS 4.x) X * X * cc -o gethostbyname gethostbyname.c -lnsl (SunOS 5.x) X * X * run as: gethostbyname hostname X * X * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. X */ X#include X#include X#include X#include X#include X#include X Xmain(argc, argv) Xint argc; Xchar **argv; X{ X struct hostent *hp; X X if (argc != 2) { X fprintf(stderr, "usage: %s hostname\n", argv[0]); X exit(1); X } X if (hp = gethostbyname(argv[1])) { X printf("Hostname:\t%s\n", hp->h_name); X printf("Aliases:\t"); X while (hp->h_aliases[0]) X printf("%s ", *hp->h_aliases++); X printf("\n"); X printf("Addresses:\t"); X while (hp->h_addr_list[0]) X printf("%s ", inet_ntoa(*(struct in_addr *) * hp->h_addr_list++)); X printf("\n"); X exit(0); X } else { X fprintf(stderr, "host %s not found\n", argv[1]); X exit(1); X } X} END_OF_gethostbyname.c if test 999 -ne `wc -c X#include X#include X#include X#include X#include X Xmain(argc, argv) Xint argc; Xchar **argv; X{ X struct hostent *hp; X long addr; X X if (argc != 2) { X fprintf(stderr, "usage: %s i.p.addres\n", argv[0]); X exit(1); X } X addr = inet_addr(argv[1]); X if (hp = gethostbyaddr((char *) &addr, sizeof(addr), AF_INET)) { X printf("Hostname:\t%s\n", hp->h_name); X printf("Aliases:\t"); X while (hp->h_aliases[0]) X printf("%s ", *hp->h_aliases++); X printf("\n"); X printf("Addresses:\t"); X while (hp->h_addr_list[0]) X printf("%s ", inet_ntoa(*(struct in_addr *) * hp->h_addr_list++)); X printf("\n"); X exit(0); X } X fprintf(stderr, "host %s not found\n", argv[1]); X exit(1); X} END_OF_gethostbyaddr.c if test 1073 -ne `wc -c gethostbyname.c <<'END_OF_gethostbyname.c' X /* X * gethostbyname tester. compile with: X * X * cc -o gethostbyname gethostbyname.c (SunOS 4.x) X * X * cc -o gethostbyname gethostbyname.c -lnsl (SunOS 5.x) X * X * run as: gethostbyname hostname X * X * Author: Wietse Venema, Eindhoven University of Technology, The Netherlands. X */ X#include X#include X#include X#include X#include X#include X Xmain(argc, argv) Xint argc; Xchar **argv; X{ X struct hostent *hp; X X if (argc != 2) { X fprintf(stderr, "usage: %s hostname\n", argv[0]); X exit(1); X } X if (hp = gethostbyname(argv[1])) { X printf("Hostname:\t%s\n", hp->h_name); X printf("Aliases:\t"); X while (hp->h_aliases[0]) X printf("%s ", *hp->h_aliases++); X printf("\n"); X printf("Addresses:\t"); X while (hp->h_addr_list[0]) X printf("%s ", inet_ntoa(*(struct in_addr *) * hp->h_addr_list++)); X printf("\n"); X exit(0); X } else { X fprintf(stderr, "host %s not found\n", argv[1]); X exit(1); X } X} END_OF_gethostbyname.c if test 999 -ne `wc -c