135062Sbostic# Copyright (c) 1983 Eric P. Allman 248582Sbostic# Copyright (c) 1988 The Regents of the University of California. 333728Sbostic# All rights reserved. 433728Sbostic# 548582Sbostic# %sccs.include.redist.sh% 633728Sbostic# 7*64364Seric# @(#)READ_ME 8.19 (Berkeley) 08/26/93 833728Sbostic# 948582Sbostic 109881SericThis directory contains the source files for sendmail. 115369Seric 1260565SericFor detailed instructions, please read the document ../doc/op.me: 135369Seric 1460565Seric eqn ../doc/op.me | pic | ditroff -me 155369Seric 1664262SericThe Makefile is for the new (4.4BSD) Berkeley make, available from 1764262Sericftp.uu.net in the directory /systems/unix/bsd-sources/usr.bin/make. 1864262SericIt has assumptions about the 4.4 file system layout built in. There 1964262Sericis also a Makefile.dist which is much less clever, but works on the 2064262Sericold traditional make. You can use this using: 2157418Seric 2257418Seric make -f Makefile.dist 2357418Seric 2464262SericThere are a bunch of other Makefiles for other systems -- these are 2560584Sericthe ones that I use, they have "Berkeley quirks" in them, and I don't 2664262Sericguarantee that they will work unmodified in your environment. However, 2764262Sericthey are all designed for the old make and can be used to help you get 2864262Sericstarted. They have names like "Makefile.HPUX". Many of them include 2964262Seric-I/usr/sww/include/db and -L/usr/sww/lib -- this is Berkeley's 3064262Sericlocation for the new database libraries, described below. 3157943Seric 3264272SericThere is also a shell script (makesendmail) that tries to be clever 3364272Sericabout using object subdirectories. It's pretty straightforward, and 3464272Sericmay help if you share a source tree among different architectures. 3564035Seric 3664272Seric 3764250Seric+----------------------+ 3864250Seric| DATABASE DEFINITIONS | 3964250Seric+----------------------+ 4064250Seric 4164250SericThere are several database formats that can be used for the alias files 4264250Sericand for general maps. When used for alias files they interact in an 4364250Sericattempt to be back compatible. 4464250Seric 4564250SericThe three options are NEWDB (the new Berkeley DB package), NDBM (the 4664250Sericolder DBM implementation -- the very old V7 implementation is no 4764250Sericlonger supported), and NIS (Network Information Services). Used alone 4864250Sericthese just include the support they indicate. 4964250Seric 5064250SericIf NEWDB and NDBM are defined (but not NIS), then sendmail will read 5164250SericNDBM format alias files, but the next time a newaliases is run the 5264250Sericformat will be converted to NEWDB; that format will be used forever 5364250Sericmore. This is intended as a transition feature. [Note however that 5464250Sericthe NEWDB library also catches and maps NDBM calls; you will have to 5564250Sericback out this feature to get this to work. See ``Quirks'' section 5664250Sericbelow for details.] 5764250Seric 5864250SericIf all three are defined, sendmail operates as described above, and also 5964250Sericlooks for the file /var/yp/Makefile. If it exists, newaliases will 6064250Sericbuild BOTH the NEWDB and NDBM format alias files. However, it will 6164250Sericonly use the NEWDB file; the NDBM format file is used only by the 6264250SericNIS subsystem. 6364250Seric 6464250SericIf NDBM and NIS are defined (regardless of the definition of NEWDB 6564250Sericor the existance of /var/yp/Makefile), sendmail adds the special 6664250Serictokens "YP_LAST_MODIFIED" and "YP_MASTER_NAME", both of which are 6764250Sericrequired if the NDBM file is to be used as an NIS map. 6864250Seric 6964250SericAll of -DNEWDB, -DNDBM, and -DNIS are normally defined in the DBMDEF 7064250Sericline in the Makefile. 7164250Seric 7264250Seric 7364035Seric+---------------+ 7464035Seric| COMPILE FLAGS | 7564035Seric+---------------+ 7664035Seric 7760565SericWhereever possible, I try to make sendmail pull in the correct 7860584Sericcompilation options needed to compile on various environments based on 7960584Sericautomatically defined symbols. Some machines don't seem to have useful 8060584Sericsymbols availble, requiring the following compilation flags in the 8160584SericMakefile: 8260565Seric 8360565SericSOLARIS Define this if you are running Solaris 2.0 or higher. 8464077SericNeXT Define this if you are on a NeXT box. (This one may 8564072Seric be pre-defined for you.) There are other hacks you 8664072Seric have to make -- see below. 8760565Seric_AIX3 Define this if you are IBM AIX 3.x. 8863965SericRISCOS Define this if you are running RISC/os from MIPS. 8960565Seric 9060584SericIf you are a system that sendmail has already been ported to, you 9160584Sericprobably won't have to touch these. But if you are porting, you may 9263962Serichave to tweak the following compilation flags in conf.h in order to 9363962Sericget it to compile and link properly: 9460565Seric 9560565SericSYSTEM5 Adjust for System V. 9664035SericSYS5SIGNALS Use System V signal semantics -- the signal handler 9764035Seric is automatically dropped when the signal is caught. 9864035Seric If this is not set, use POSIX/BSD semantics, where the 9964035Seric signal handler stays in force until an exec or an 10064035Seric explicit delete. Implied by SYSTEM5. 10164035SericHASFLOCK Set this if you prefer to use the flock(2) system call 10264035Seric rather than using fcntl-based locking. Fcntl locking 10364035Seric has some semantic gotchas, but many vendor systems 10464035Seric also interface it to lockd(8) to do NFS-style locking. 10564035Seric For this reason, this should not be set unless you 10664035Seric don't have an alternative. 10760565SericHASUNAME Set if you have the "uname" system call. Implied by 10860565Seric SYSTEM5. 10963962SericHASUNSETENV Define this if your system library has the "unsetenv" 11063962Seric subroutine. 11160584SericHASSTATFS Define this if you have the statfs(2) system call. It's 11260584Seric not a disaster to get this wrong -- but you do lose the 11360584Seric queue free space code. 11460584SericHASUSTAT Define this if you have the ustat(2) system call. It's 11560584Seric not a disaster to get this wrong -- but you do lose the 11660584Seric queue free space code. 11760565SericHASSETSID Define this if you have the setsid(2) system call. This 11860565Seric is implied if your system appears to be POSIX compliant. 11960565SericHASINITGROUPS Define this if you have the initgroups(3) routine. 12063753SericHASSETVBUF Define this if you have the setvbuf(3) library call. 12163753Seric If you don't, setlinebuf will be used instead. This 12263753Seric defaults on if your compiler defines __STDC__. 12363902SericHASSETREUID Define this if you have setreuid(2) ***AND*** root can 12463902Seric use setreuid to change to an arbitrary user. This second 12563902Seric condition is not satisfied on AIX 3.x. You may find that 12663902Seric your system has setresuid(2), (for example, on HP-UX) in 12763902Seric which case you will also have to #define setreuid(r, e) 12863902Seric to be the appropriate call. Some systems (such as Solaris) 12963902Seric have a compatibility routine that doesn't work properly. 13063902Seric The important thing is that you have a call that will set 13163902Seric the effective uid independently of the real or saved uid. 13263902Seric Setting this improves the security somewhat, since 13363902Seric sendmail doesn't have to read .forward and :include: files 13463902Seric as root. 13563937SericGIDSET_T The type of entries in a gidset passed as the second 13663937Seric argument to getgroups(2). Historically this has been an 13763937Seric int, so this is the default, but some systems (such as 13863937Seric IRIX) pass it as a gid_t, which is an unsigned short. 13963937Seric This will make a difference, so it is important to get 14063937Seric this right! However, it is only an issue if you have 14163937Seric group sets. 14263968SericSLEEP_T The type returned by the system sleep() function. 14363968Seric Defaults to "unsigned int". Don't worry about this 14463968Seric if you don't have compilation problems. 14563974SericARBPTR_T The type of an arbitrary pointer -- defaults to "void *". 14663974Seric If you are an very old compiler you may need to define 14763974Seric this to be "char *". 14860584SericLA_TYPE The type of load average your kernel supports. These 14960584Seric can be LA_SUBR (4) if you have the getloadavg(3) routine, 15060584Seric LA_FLOAT (3) if you read kmem and interpret the value 15160584Seric as a floating point number, LA_INT (2) to interpret as 15260584Seric an integer. These last two have several other parameters 15360584Seric that they try to divine: the name of your kernel, the name 15460584Seric of the variable in the kernel to examine, the number of 15560584Seric bits of precision in a fixed point load average, and so 15660584Seric forth. In desparation, use LA_ZERO -- it always returns 15760584Seric the load average as "zero" (and does so on all architectures). 15860584Seric The actual code is in conf.c -- it can be tweaked if you 15960584Seric are brave. 16063962SericERRLIST_PREDEFINED 16163962Seric If set, assumes that some header file defines sys_errlist. 16263962Seric This may be needed if you get type conflicts on this 16363962Seric variable -- otherwise don't worry about it. 16460565Seric 16564035Seric 16664035Seric+-----------------------+ 16764035Seric| COMPILE-TIME FEATURES | 16864035Seric+-----------------------+ 16964035Seric 17060584SericThere are a bunch of features that you can decide to compile in, such 17160584Sericas selecting various database packages and special protocol support. 17260584SericSeveral are assumed based on other compilation flags -- if you want to 17360584Seric"un-assume" something, you probably need to edit conf.h. Compilation 17460584Sericflags that add support for special features include: 17560565Seric 17660565SericNDBM Include support for "new" DBM library for aliases and maps. 17764250Seric Normally defined in the Makefile. 17860565SericNEWDB Include support for Berkeley "db" package (hash & btree) 17964250Seric for aliases and maps. Normally defined in the Makefile. 18060565SericNIS Define this to get NIS (YP) support for aliases and maps. 18164250Seric Normally defined in the Makefile. 18260565SericUSERDB Include support for the User Information Database. Implied 18364250Seric by NEWDB in conf.h. 18460565SericIDENTPROTO Define this to get IDENT (RFC 1413) protocol support. 18560565Seric This is assumed unless you are running on Ultrix or 18660565Seric HP-UX, both of which have a problem in the UDP 18760565Seric implementation. 18860565SericMIME Include support for MIME-encapsulated error messages. 18960565SericLOG Set this to get syslog(3) support. Defined by default 19060584Seric in conf.h. You want this if at all possible. 19160565SericNETINET Set this to get TCP/IP support. Defined by default 19260584Seric in conf.h. You probably want this. 19360565SericNETISO Define this to get ISO networking support. 19460565SericSMTP Define this to get the SMTP code. Implied by NETINET 19560565Seric or NETISO. 19660565SericNAMED_BIND Define this to get DNS (name daemon) support, including 19760565Seric MX support. The specs you must use this if you run 19860565Seric SMTP. Defined by default in conf.h. 19960565SericQUEUE Define this to get queueing code. Implied by NETINET 20060584Seric or NETISO; required by SMTP. This gives you other good 20160584Seric stuff -- it should be on. 20260565SericDAEMON Define this to get general network support. Implied by 20360584Seric NETINET or NETISO. Defined by default in conf.h. You 20460584Seric almost certainly want it on. 20560565SericMATCHGECOS Permit fuzzy matching of user names against the full 20660565Seric name (GECOS) field in the /etc/passwd file. This should 20760565Seric probably be on, since you can disable it from the config 20860584Seric file if you want to. Defined by default in conf.h. 20960565SericSETPROCTITLE Try to set the string printed by "ps" to something 21060584Seric informative about what sendmail is doing. Defined by 21160584Seric default in conf.h. 21260565Seric 21364035Seric 21464035Seric+-------------------------------------+ 21564035Seric| OPERATING SYSTEM AND COMPILE QUIRKS | 21664035Seric+-------------------------------------+ 21764035Seric 21864250SericSunOS 21964250Seric You may have to use -lresolv on SunOS. 22064035Seric 221*64364SericSolaris 222*64364Seric From a correspondent: 223*64364Seric 224*64364Seric For solaris 2.2, I have 225*64364Seric 226*64364Seric hosts: files dns 227*64364Seric 228*64364Seric in /etc/nsswitch.conf and /etc/hosts has to have the fully 229*64364Seric qualified host name. I think "files" has to be before "dns" 230*64364Seric in /etc/nsswitch.conf during bootup. 231*64364Seric 23264250SericOSF/1 23364250Seric If you are compiling on OSF/1 (DEC Alpha), you must use -lmld. 23457977Seric 23564250SericNeXT 23664250Seric If you are compiling on NeXT, you will have to create an empty 23764250Seric file "unistd.h" and create a file "dirent.h" containing: 23863753Seric 23964250Seric #include <sys/dir.h> 24064250Seric #define dirent direct 24164035Seric 24264250Seric (The Makefile.NeXT should try to do both of these for you.) 24364077Seric 244*64364Seric Apparently, there is a bug in getservbyname on Nextstep 3.0 245*64364Seric that causes it to fail under some circumstances with the 246*64364Seric message "SYSERR: service "smtp" unknown" logged. You should 247*64364Seric be able to work around this by including the line: 248*64364Seric 249*64364Seric OPort=25 250*64364Seric 251*64364Seric in your .cf file. 252*64364Seric 25364250SericBSDI (BSD/386) 25464250Seric I have reports that the "m4" from BSDI won't handle the config 25564250Seric files properly. I haven't had a chance to test this myself. 25657943Seric 257*64364Seric4.3BSD 258*64364Seric If you are running a "virgin" version of 4.3BSD, you'll have 259*64364Seric a very old resolver and be missing some header files. The 260*64364Seric header files are simple -- create empty versions and everything 261*64364Seric will work fine. For the resolver you should really port a new 262*64364Seric version (4.8.3 or later) of the resolver; 4.9 is available on 263*64364Seric gatekeeper.DEC.COM in pub/BSD/bind/4.9. If you are really 264*64364Seric determined to continue to use your old, buggy version (or as 265*64364Seric a shortcut to get sendmail working -- I'm sure you have the 266*64364Seric best intentions to port a modern version of BIND), you can 267*64364Seric copy ../contrib/oldbind.compat.c into src and add 268*64364Seric oldbind.compat.o to OBJADD in the Makefile. 269*64364Seric 27064250SericBoth NEWDB and NDBM 27164250Seric If you use both -DNDBM and -DNEWDB, you must delete the module 27264250Seric ndbm.o from libdb.a and delete the file "ndbm.h" from the files 27364250Seric that get installed (that is, use the OLD ndbm.h, not the new 27464250Seric ndbm.h). This compatibility module maps ndbm calls into DB 27564250Seric calls, and breaks things rather badly. 27658709Seric 27764250Seric 27864035Seric+-----------------------------+ 27964035Seric| DESCRIPTION OF SOURCE FILES | 28064035Seric+-----------------------------+ 28164035Seric 2829881SericThe following list describes the files in this directory: 2835369Seric 28457418SericMakefile The makefile used here; this version only works with 28557418Seric the new Berkeley make. 28657418SericMakefile.dist A trimmed down version of the makefile that works with 28757418Seric the old make. 2885369SericREAD_ME This file. 28960565SericTRACEFLAGS My own personal list of the trace flags -- not guaranteed 29060565Seric to be particularly up to date. 2915369Sericalias.c Does name aliasing in all forms. 2929881Sericarpadate.c A subroutine which creates ARPANET standard dates. 2939881Sericclock.c Routines to implement real-time oriented functions 2949881Seric in sendmail -- e.g., timeouts. 2955369Sericcollect.c The routine that actually reads the mail into a temp 2965369Seric file. It also does a certain amount of parsing of 2975369Seric the header, etc. 2985369Sericconf.c The configuration file. This contains information 2995369Seric that is presumed to be quite static and non- 3005369Seric controversial, or code compiled in for efficiency 3015369Seric reasons. Most of the configuration is in sendmail.cf. 3029881Sericconf.h Configuration that must be known everywhere. 3035369Sericconvtime.c A routine to sanely process times. 3049881Sericdaemon.c Routines to implement daemon mode. This version is 3059881Seric specifically for Berkeley 4.1 IPC. 3065369Sericdeliver.c Routines to deliver mail. 30760565Sericdomain.c Routines that interface with DNS (the Domain Name 30860565Seric System). 3095369Sericerr.c Routines to print error messages. 3109881Sericenvelope.c Routines to manipulate the envelope structure. 3115369Sericheaders.c Routines to process message headers. 3125369Sericmacro.c The macro expander. This is used internally to 3135369Seric insert information from the configuration file. 3145369Sericmain.c The main routine to sendmail. This file also 3155369Seric contains some miscellaneous routines. 31660565Sericmap.c Support for database maps. 31760565Sericmci.c Routines that handle mail connection information caching. 3189881Sericparseaddr.c The routines which do address parsing. 3195369Sericqueue.c Routines to implement message queueing. 3205369Sericreadcf.c The routine that reads the configuration file and 3215369Seric translates it to internal form. 3229881Sericrecipient.c Routines that manipulate the recipient list. 3235369Sericsavemail.c Routines which save the letter on processing errors. 3245369Sericsendmail.h Main header file for sendmail. 3255369Sericsrvrsmtp.c Routines to implement server SMTP. 3265369Sericstab.c Routines to manage the symbol table. 3275369Sericstats.c Routines to collect and post the statistics. 3285369Sericsysexits.c List of error messages associated with error codes 3295369Seric in sysexits.h. 3309881Serictrace.c The trace package. These routines allow setting and 3319881Seric testing of trace flags with a high granularity. 33260565Sericudb.c The user database interface module. 3335369Sericusersmtp.c Routines to implement user SMTP. 3345369Sericutil.c Some general purpose routines used by sendmail. 33560565Sericversion.c The version number and information about this 33660565Seric version of sendmail. Theoretically, this gets 33760565Seric modified on every change. 3385369Seric 3395369SericEric Allman 3405369Seric 341*64364Seric(Version 8.19, last update 08/26/93 11:22:15) 342