1# Copyright (c) 1983 Eric P. Allman 2# Copyright (c) 1988 The Regents of the University of California. 3# All rights reserved. 4# 5# %sccs.include.redist.sh% 6# 7# @(#)READ_ME 8.78 (Berkeley) 02/06/95 8# 9 10This directory contains the source files for sendmail. 11 12For detailed instructions, please read the document ../doc/op.me: 13 14 eqn ../doc/op.me | pic | ditroff -me 15 16********************* 17!! DO NOT USE MAKE !! to compile sendmail -- instead, use the 18********************* "makesendmail" script located in the src 19directory. It will find an appropriate Makefile, and create an 20appropriate obj.* subdirectory so that multiplatform support 21works easily. 22 23The Makefile is for the new (4.4BSD) Berkeley make and uses syntax 24that is not recognized by older makes. It also has assumptions 25about the 4.4 file system layout built in. See below for details 26about other Makefiles. 27 28If you are porting to a new architecture for which there is no existing 29Makefile, you might start with Makefile.dist. This works on the old 30traditional make, but isn't customized for any particular architecture. 31 32 ************************************************** 33 ** Read below for more details of Makefiles. ** 34 ************************************************** 35 36************************************************************************** 37** IMPORTANT: DO NOT USE OPTIMIZATION (``-O'') IF YOU ARE RUNNING ** 38** GCC 2.4.x or 2.5.x. THERE IS A BUG IN THE GCC OPTIMIZER THAT ** 39** CAUSES SENDMAIL COMPILES TO FAIL MISERABLY. ** 40************************************************************************** 41 42Jim Wilson of Cygnus believes he has found the problem -- it will 43probably be fixed in GCC 2.5.6 -- but until this is verified, be 44very suspicious of gcc -O. 45 46************************************************************************** 47** IMPORTANT: Read the appropriate paragraphs in the section on ** 48** ``Operating System and Compile Quirks''. ** 49************************************************************************** 50 51 52+-----------+ 53| MAKEFILES | 54+-----------+ 55 56By far, the easiest way to compile sendmail is to use the "makesendmail" 57script: 58 59 sh makesendmail 60 61This uses the "uname" command to figure out what architecture you are 62on and selects a proper Makefile accordingly. It also creates a 63subdirectory per object format, so that multiarchitecture support is 64easy. In general this should be all you need. However, if for some 65reason this doesn't work (e.g., NeXT systems don't have the "uname" 66command) you may have to set up your compile environment by hand. 67 68The "Makefile"s in these directories are from 4.4 BSD, and hence 69really only work properly if you are on a 4.4 system. In particular, 70they use new syntax that will not be recognized on old make programs, 71and some of them do things like ``.include ../../Makefile.inc'' to 72pick up some system defines. If you are getting sendmail separately, 73these files won't be included in the distribution, as they are 74outside of the sendmail tree. 75 76Instead, you should use one of the other Makefiles, such as 77Makefile.SunOS for a SunOS system, and so forth. These should 78work with the version of make that is appropriate for that 79system. All other Makefiles are in the "src/Makefiles" subdirectory. 80They use the version of make that is native for that system. These 81are the Makefiles that I use, and they have "Berkeley quirks" in them. 82I can't guarantee that they will work unmodified in your environment. 83In particular, Many of them include -I/usr/sww/include/db and 84-L/usr/sww/lib -- these are Berkeley's locations in the ``Software 85Warehouse'' for the new database libraries, described below. You don't 86have to remove these definitions if you don't have these directories, 87but you may have to remove -DNEWDB from the DBMDEF definition. 88 89Please look for an appropriate Makefile before you start trying to 90compile with Makefile or Makefile.dist. 91 92If you want to port the new Berkeley make, you can get it from 93ftp.uu.net in the directory /systems/unix/bsd-sources/usr.bin/make. 94Diffs and instructions for building this version of make under 95SunOS 4.1.x are available on ftp.css.itd.umich.edu in 96/pub/systems/sun/Net2-make.sun4.diff.Z. Diffs and instructions 97for building this version of make under IBM AIX 3.2.4 are available 98on ftp.uni-stuttgart.de in /sw/src/patches/bsd-make-rus-patches. 99For Ultrix, try ftp.vix.com:~ftp/pub/patches/pmake-for-ultrix.Z. 100Paul Southworth <pauls@umich.edu> published a description of porting 101this make in comp.unix.bsd. 102 103The complete text of the Makefile.inc that is in the parent of the 104sendmail directory is: 105 106 # @(#)Makefile.inc 8.1 (Berkeley) 6/6/93 107 108 BINDIR?= /usr/sbin 109 110 111+----------------------+ 112| DATABASE DEFINITIONS | 113+----------------------+ 114 115There are several database formats that can be used for the alias files 116and for general maps. When used for alias files they interact in an 117attempt to be back compatible. 118 119The three options are NEWDB (the new Berkeley DB package), NDBM (the 120older DBM implementation -- the very old V7 implementation is no 121longer supported), and NIS (Network Information Services). Used alone 122these just include the support they indicate. [If you are using NEWDB, 123get the latest version from FTP.CS.Berkeley.EDU in /ucb/4bsd/db.tar.Z 124(or db.tar.gz). DO NOT use the version from the Net2 distribution! 125However, if you are on BSD/386 or 386BSD-based systems, use the one 126that already exists on your system. You may need to #define OLD_NEWDB 1271 to do this.] 128 129[NOTE WELL: it is CRITICAL that you remove ndbm.o from libdb.a and 130ndbm.h from the appropriate include directories if you want to get 131ndbm support. These files OVERRIDE calls to ndbm routines -- in 132particular, if you leave ndbm.h in, you can find yourself using 133the new db package even if you don't define NEWDB.] 134 135If NEWDB and NDBM are defined (but not NIS), then sendmail will read 136NDBM format alias files, but the next time a newaliases is run the 137format will be converted to NEWDB; that format will be used forever 138more. This is intended as a transition feature. [Note however that 139the NEWDB library also catches and maps NDBM calls; you will have to 140back out this feature to get this to work. See ``Quirks'' section 141below for details.] 142 143If all three are defined, sendmail operates as described above, and also 144looks for the file /var/yp/Makefile. If it exists, newaliases will 145build BOTH the NEWDB and NDBM format alias files. However, it will 146only use the NEWDB file; the NDBM format file is used only by the 147NIS subsystem. 148 149If NDBM and NIS are defined (regardless of the definition of NEWDB 150or the existance of /var/yp/Makefile), sendmail adds the special 151tokens "YP_LAST_MODIFIED" and "YP_MASTER_NAME", both of which are 152required if the NDBM file is to be used as an NIS map. 153 154There is also preliminary support for NIS+ (-DNISPLUS), Hesiod 155(-DHESIOD), and NetInfo (-DNETINFO). These have not been well 156tested. 157 158All of -DNEWDB, -DNDBM, -DNIS, -DNISPLUS, -DHESIOD, and -DNETINFO are 159normally defined in the DBMDEF line in the Makefile. 160 161 162+---------------+ 163| COMPILE FLAGS | 164+---------------+ 165 166Whereever possible, I try to make sendmail pull in the correct 167compilation options needed to compile on various environments based on 168automatically defined symbols. Some machines don't seem to have useful 169symbols availble, requiring the following compilation flags in the 170Makefile: 171 172SOLARIS Define this if you are running Solaris 2.0 or higher. 173SOLARIS_2_3 Define this if you are running Solaris 2.3 or higher. 174SUNOS403 Define this if you are running SunOS 4.0.3. 175NeXT Define this if you are on a NeXT box. (This one may 176 be pre-defined for you.) There are other hacks you 177 have to make -- see below. 178_AIX3 Define this if you are IBM AIX 3.x. 179RISCOS Define this if you are running RISC/os from MIPS. 180IRIX Define this if you are running IRIX from SGI. 181_SCO_unix_ Define this if you are on SCO UNIX. 182_SCO_unix_4_2 Define this if you are on SCO Open Server 3.2v4. 183DGUX Define this if you are on DG/UX 5.4.3 or later 184DGUX_5_4_2 Define this if you are on DG/UX systems prior to 5.4.3. 185NonStop_UX_BXX Define this if you are on a Tandem NonStop-UX release 186 Bxx system. 187IRIX64 Define this if you are on an IRIX64 system. 188 189If you are a system that sendmail has already been ported to, you 190probably won't have to touch these. But if you are porting, you may 191have to tweak the following compilation flags in conf.h in order to 192get it to compile and link properly: 193 194SYSTEM5 Adjust for System V (not necessarily Release 4). 195SYS5SIGNALS Use System V signal semantics -- the signal handler 196 is automatically dropped when the signal is caught. 197 If this is not set, use POSIX/BSD semantics, where the 198 signal handler stays in force until an exec or an 199 explicit delete. Implied by SYSTEM5. 200SYS5SETPGRP Use System V setpgrp() semantics. Implied by SYSTEM5. 201HASFLOCK Set this if you prefer to use the flock(2) system call 202 rather than using fcntl-based locking. Fcntl locking 203 has some semantic gotchas, but many vendor systems 204 also interface it to lockd(8) to do NFS-style locking. 205 For this reason, this should not be set unless you 206 don't have an alternative. 207HASUNAME Set if you have the "uname" system call. Implied by 208 SYSTEM5. 209HASUNSETENV Define this if your system library has the "unsetenv" 210 subroutine. 211HASSETSID Define this if you have the setsid(2) system call. This 212 is implied if your system appears to be POSIX compliant. 213HASINITGROUPS Define this if you have the initgroups(3) routine. 214HASSETVBUF Define this if you have the setvbuf(3) library call. 215 If you don't, setlinebuf will be used instead. This 216 defaults on if your compiler defines __STDC__. 217HASSETREUID Define this if you have setreuid(2) ***AND*** root can 218 use setreuid to change to an arbitrary user. This second 219 condition is not satisfied on AIX 3.x. You may find that 220 your system has setresuid(2), (for example, on HP-UX) in 221 which case you will also have to #define setreuid(r, e) 222 to be the appropriate call. Some systems (such as Solaris) 223 have a compatibility routine that doesn't work properly, 224 but may have "saved user ids" properly implemented so you 225 can ``#define setreuid(r, e) seteuid(e)'' and have it work. 226 The important thing is that you have a call that will set 227 the effective uid independently of the real or saved uid 228 and be able to set the effective uid back again when done. 229 There's a test program in ../test/t_setreuid.c that will 230 try things on your system. Setting this improves the 231 security, since sendmail doesn't have to read .forward 232 and :include: files as root. There are certain attacks 233 that may be unpreventable without this call. 234HASLSTAT Define this if you have symbolic links (and thus the 235 lstat(2) system call). This improves security. Unlike 236 most other options, this one is on by default, so you 237 need to #undef it in conf.h if you don't have symbolic 238 links (these days everyone does). 239HASSETRLIMIT Define this to 1 if you have the setrlimit(2) syscall. 240 You can define it to 0 to force it off. It is assumed 241 if you are running a BSD-like system. 242HASULIMIT Define this if you have the ulimit(2) syscall (System V 243 style systems). HASSETRLIMIT overrides, as it is more 244 general. 245NEEDGETOPT Define this if you need a reimplementation of getopt(3). 246 On some systems, getopt does very odd things if called 247 to scan the arguments twice. This flag will ask sendmail 248 to compile in a local version of getopt that works 249 properly. 250NEEDSTRTOL Define this if your standard C library does not define 251 strtol(3). This will compile in a local version. 252NEEDVPRINTF Define this if your standard C library does not define 253 vprintf(3). Note that the resulting fake implementation 254 is not very elegant and may not even work on some 255 architectures. 256NEEDFSYNC Define this if your standard C library does not define 257 fsync(2). This will try to simulate the operation using 258 fcntl(2); if that is not available it does nothing, which 259 isn't great, but at least it compiles and runs. 260HASGETUSERSHELL Define this to 1 if you have getusershell(3) in your 261 standard C library. If this is not defined, or is defined 262 to be 0, sendmail will scan the /etc/shells file (no 263 NIS-style support, defaults to /bin/sh and /bin/csh if 264 that file does not exist) to get a list of unrestricted 265 user shells. This is used to determine whether users 266 are allowed to forward their mail to a program or a file. 267GIDSET_T The type of entries in a gidset passed as the second 268 argument to getgroups(2). Historically this has been an 269 int, so this is the default, but some systems (such as 270 IRIX) pass it as a gid_t, which is an unsigned short. 271 This will make a difference, so it is important to get 272 this right! However, it is only an issue if you have 273 group sets. 274SLEEP_T The type returned by the system sleep() function. 275 Defaults to "unsigned int". Don't worry about this 276 if you don't have compilation problems. 277ARBPTR_T The type of an arbitrary pointer -- defaults to "void *". 278 If you are an very old compiler you may need to define 279 this to be "char *". 280LA_TYPE The type of load average your kernel supports. These 281 can be one of: 282 LA_ZERO (1) -- it always returns the load average as 283 "zero" (and does so on all architectures). 284 LA_SUBR (4) if you have the getloadavg(3) routine, 285 LA_MACH (5) to use MACH-style load averages (calls 286 processor_set_info()), 287 LA_PROCSTR (7) to read /proc/loadavg and interpret it 288 as a string representing a floating-point 289 number (Linux-style), 290 LA_FLOAT (3) if you read kmem and interpret the value 291 as a floating point number, 292 LA_INT (2) to interpret as a long integer, 293 LA_SHORT (6) to interpret as a short integer. 294 These last three have several other parameters that they 295 try to divine: the name of your kernel, the name of the 296 variable in the kernel to examine, the number of bits of 297 precision in a fixed point load average, and so forth. 298 In desperation, use LA_ZERO. The actual code is in 299 conf.c -- it can be tweaked if you are brave. 300SFS_TYPE Encodes how your kernel can locate the amount of free 301 space on a disk partition. This can be set to SFS_NONE 302 (0) if you have no way of getting this information, 303 SFS_USTAT (1) if you have the ustat(2) system call, 304 SFS_4ARGS (2) if you have a four-argument statfs(2) 305 system call (and the include file is <sys/statfs.h>), 306 SFS_VFS (3), SFS_MOUNT (4), SFS_STATFS (5) if you have 307 the two-argument statfs(2) system call with includes in 308 <sys/vfs.h>, <sys/mount.h>, or <sys/statfs.h> respectively, 309 or SFS_STATVFS (6) if you have the two-argument statvfs(2) 310 call. The default if nothing is defined is SFS_NONE. 311SPT_TYPE Encodes how your system can display what a process is doing 312 on a ps(1) command (SPT stands for Set Process Title). Can 313 be set to: 314 SPT_NONE (0) -- Don't try to set the process title at all. 315 SPT_REUSEARGV (1) -- Pad out your argv with the information; 316 this is the default if none specified. 317 SPT_BUILTIN (2) -- The system library has setproctitle. 318 SPT_PSTAT (3) -- Use the PSTAT_SETCMD option to pstat(2) 319 to set the process title; this is used by HP-UX. 320 SPT_PSSTRINGS (4) -- Use the magic PS_STRINGS pointer (4.4BSD). 321SPT_PADCHAR Character used to pad the process title; if undefined, 322 the space character (0x20) is used. This is ignored if 323 SPT_TYPE != SPT_REUSEARGV 324ERRLIST_PREDEFINED 325 If set, assumes that some header file defines sys_errlist. 326 This may be needed if you get type conflicts on this 327 variable -- otherwise don't worry about it. 328WAITUNION The wait(2) routine takes a "union wait" argument instead 329 of an integer argument. This is for compatibility with 330 old versions of BSD. 331SCANF You can set this to extend the F command to accept a 332 scanf string -- this gives you a primitive parser for 333 class definitions -- BUT it can make you vulnerable to 334 core dumps if the target file is poorly formed. 335SYSLOG_BUFSIZE You can define this to be the size of the buffer that 336 syslog accepts. If it is not defined, it assumes a 337 1024-byte buffer. If the buffer is very small (under 338 256 bytes) the log message format changes -- each 339 e-mail message will log many more messages, since it 340 will log each piece of information as a separate line 341 in syslog. 342BROKEN_RES_SEARCH 343 On Ultrix (and maybe other systems?) if you use the 344 res_search routine with an unknown host name, it returns 345 -1 but sets h_errno to 0 instead of HOST_NOT_FOUND. If 346 you set this, sendmail considers 0 to be the same as 347 HOST_NOT_FOUND. 348NAMELISTMASK If defined, values returned by nlist(3) are masked 349 against this value before use -- a common value is 350 0x7fffffff to strip off the top bit. 351 352 353 354+-----------------------+ 355| COMPILE-TIME FEATURES | 356+-----------------------+ 357 358There are a bunch of features that you can decide to compile in, such 359as selecting various database packages and special protocol support. 360Several are assumed based on other compilation flags -- if you want to 361"un-assume" something, you probably need to edit conf.h. Compilation 362flags that add support for special features include: 363 364NDBM Include support for "new" DBM library for aliases and maps. 365 Normally defined in the Makefile. 366NEWDB Include support for Berkeley "db" package (hash & btree) 367 for aliases and maps. Normally defined in the Makefile. 368OLD_NEWDB If non-zero, the version of NEWDB you have is the old 369 one that does not include the "fd" call. This call was 370 added in version 1.5 of the Berkeley DB code. If you 371 use -DOLD_NEWDB=0 it forces you to use the new interface. 372NIS Define this to get NIS (YP) support for aliases and maps. 373 Normally defined in the Makefile. 374USERDB Include support for the User Information Database. Implied 375 by NEWDB in conf.h. 376IDENTPROTO Define this as 1 to get IDENT (RFC 1413) protocol support. 377 This is assumed unless you are running on Ultrix or 378 HP-UX, both of which have a problem in the UDP 379 implementation. You can define it to be 0 to explicitly 380 turn off IDENT protocol support. 381LOG Set this to get syslog(3) support. Defined by default 382 in conf.h. You want this if at all possible. 383NETINET Set this to get TCP/IP support. Defined by default 384 in conf.h. You probably want this. 385NETISO Define this to get ISO networking support. 386SMTP Define this to get the SMTP code. Implied by NETINET 387 or NETISO. 388NAMED_BIND Define this to get DNS (name daemon) support, including 389 MX support. The specs you must use this if you run 390 SMTP. Defined by default in conf.h. 391QUEUE Define this to get queueing code. Implied by NETINET 392 or NETISO; required by SMTP. This gives you other good 393 stuff -- it should be on. 394DAEMON Define this to get general network support. Implied by 395 NETINET or NETISO. Defined by default in conf.h. You 396 almost certainly want it on. 397MATCHGECOS Permit fuzzy matching of user names against the full 398 name (GECOS) field in the /etc/passwd file. This should 399 probably be on, since you can disable it from the config 400 file if you want to. Defined by default in conf.h. 401 402 403+---------------------+ 404| DNS/RESOLVER ISSUES | 405+---------------------+ 406 407Many systems have old versions of the resolver library. At a minimum, 408you should be running BIND 4.8.3; older versions may compile, but they 409have known bugs that should give you pause. 410 411Common problems in old versions include "undefined" errors for 412dn_skipname. 413 414Some people have had a problem with BIND 4.9; it uses some routines 415that it expects to be externally defined such as strerror(). It may 416help to link with "-l44bsd" to solve this problem. 417 418!PLEASE! be sure to link with the same version of the resolver as 419the header files you used -- some people have used the 4.9 headers 420and linked with BIND 4.8 or vice versa, and it doesn't work. 421Unfortunately, it doesn't fail in an obvious way -- things just 422subtly don't work. 423 424 425+-------------------------------------+ 426| OPERATING SYSTEM AND COMPILE QUIRKS | 427+-------------------------------------+ 428 429GCC 2.5.x problems *** IMPORTANT *** 430 Date: Mon, 29 Nov 93 19:08:44 PST 431 From: wilson@cygnus.com (Jim Wilson) 432 Message-Id: <9311300308.AA04608@cygnus.com> 433 To: kenner@vlsi1.ultra.nyu.edu 434 Subject: [cattelan@thebarn.com: gcc 2.5.4-2.5.5 -O bug] 435 Cc: cattelan@thebarn.com, rms@gnu.ai.mit.edu, sendmail@cs.berkeley.edu 436 437 This fixes a problem that occurs when gcc 2.5.5 is used to compile 438 sendmail 8.6.4 with optimization on a sparc. 439 440 Mon Nov 29 19:00:14 1993 Jim Wilson (wilson@sphagnum.cygnus.com) 441 442 * reload.c (find_reloads_toplev): Replace obsolete reference to 443 BYTE_LOADS_*_EXTEND with LOAD_EXTEND_OP. 444 445 *** clean-ss-931128/reload.c Sun Nov 14 16:20:01 1993 446 --- ss-931128/reload.c Mon Nov 29 18:52:55 1993 447 *************** find_reloads_toplev (x, opnum, type, ind 448 *** 3888,3894 **** 449 force a reload in that case. So we should not do anything here. */ 450 451 else if (regno >= FIRST_PSEUDO_REGISTER 452 ! #if defined(BYTE_LOADS_ZERO_EXTEND) || defined(BYTE_LOADS_SIGN_EXTEND) 453 && (GET_MODE_SIZE (GET_MODE (x)) 454 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 455 #endif 456 --- 3888,3894 ---- 457 force a reload in that case. So we should not do anything here. */ 458 459 else if (regno >= FIRST_PSEUDO_REGISTER 460 ! #ifdef LOAD_EXTEND_OP 461 && (GET_MODE_SIZE (GET_MODE (x)) 462 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 463 #endif 464 465 466SunOS 4.x (Solaris 1.x) 467 You may have to use -lresolv on SunOS. However, beware that 468 this links in a new version of gethostbyname that does not 469 understand NIS, so you must have all of your hosts in DNS. 470 471 Some people have reported problems with the SunOS version of 472 -lresolv and/or in.named, and suggest that you get a newer 473 version. The symptoms are delays when you connect to the 474 SMTP server on a SunOS machine or having your domain added to 475 addresses inappropriately. There is a version of BIND 476 version 4.9 on gatekeeper.DEC.COM in pub/BSD/bind/4.9. 477 478 There is substantial disagreement about whether you can make 479 this work with resolv+, which allows you to specify a search-path 480 of services. Some people report that it works fine, others 481 claim it doesn't work at all (including causing sendmail to 482 drop core when it tries to do multiple resolv+ lookups for a 483 single job). I haven't tried resolv+, as we use DNS exclusively. 484 485 Should you want to try resolv+, it is on ftp.uu.net in 486 /networking/ip/dns. 487 488 Apparently getservbyname() can fail under moderate to high 489 load under some circumstances. This will exhibit itself as 490 the message ``554 makeconnection: service "smtp" unknown''. 491 The problem has been traced to one or more blank lines in 492 /etc/services on the NIS server machine. Delete these 493 and it should work. This info is thanks to Brian Bartholomew 494 <bb@math.ufl.edu> of I-Kinetics, Inc. 495 496Solaris 2.x (SunOS 5.x) 497 To compile for Solaris, be sure you use -DSOLARIS. 498 499 To the best of my knowledge, Solaris does not have the 500 gethostbyname problem described above. However, it does 501 have another one: 502 503 From a correspondent: 504 505 For solaris 2.2, I have 506 507 hosts: files dns 508 509 in /etc/nsswitch.conf and /etc/hosts has to have the fully 510 qualified host name. I think "files" has to be before "dns" 511 in /etc/nsswitch.conf during bootup. 512 513 From another correspondent: 514 515 When running sendmail under Solaris, the gethostbyname() 516 hack in conf.c which should perform proper canonicalization 517 of host names could fail. Result: the host name is not 518 canonicalized despite the hack, and you'll have to define $j 519 and $m in sendmail.cf somewhere. 520 521 The reason could be that /etc/nsswitch.conf is improperly 522 configured (at least from sendmail's point of view). For 523 example, the line 524 525 hosts: files nisplus dns 526 527 will make gethostbyname() look in /etc/hosts first, then ask 528 nisplus, then dns. However, if /etc/hosts does not contain 529 the full canonicalized hostname, then no amount of 530 gethostbyname()s will work. 531 532 Solution (or rather, a workaround): Ask nisplus first, then 533 dns, then local files: 534 535 hosts: nisplus dns [NOTFOUND=return] files 536 537 The Solaris "syslog" function is apparently limited to something 538 about 90 characters because of a kernel limitation. If you have 539 source code, you can probably up this number. You can get patches 540 that fix this problem: the patch ids are: 541 542 Solaris 2.1 100834 543 Solaris 2.2 100999 544 Solaris 2.3 101318 545 546 Be sure you have the appropriate patch installed or you won't 547 see system logging. 548 549OSF/1 550 If you are compiling on OSF/1 (DEC Alpha), you must use 551 -L/usr/shlib (otherwise it core dumps on startup). You may also 552 need -mld to get the nlist() function, although some versions 553 apparently don't need this. 554 555 Also, the enclosed makefile removed /usr/sbin/smtpd; if you need 556 it, just create the link to the sendmail binary. 557 558IRIX 559 The header files on SGI IRIX are completely prototyped, and as 560 a result you can sometimes get some warning messages during 561 compilation. These can be ignored. There are two errors in 562 deliver only if you are using gcc, both of the form ``warning: 563 passing arg N of `execve' from incompatible pointer type''. 564 Also, if you compile with -DNIS, you will get a complaint 565 about a declaration of struct dom_binding in a prototype 566 when compiling map.c; this is not important because the 567 function being prototyped is not used in that file. 568 569 In order to compile sendmail you will have had to install 570 the developers' option in order to get the necessary include 571 files. 572 573NeXT 574 If you are compiling on NeXT, you will have to create an empty 575 file "unistd.h" and create a file "dirent.h" containing: 576 577 #include <sys/dir.h> 578 #define dirent direct 579 580 (The Makefile.NeXT should try to do both of these for you.) 581 582 Apparently, there is a bug in getservbyname on Nextstep 3.0 583 that causes it to fail under some circumstances with the 584 message "SYSERR: service "smtp" unknown" logged. You should 585 be able to work around this by including the line: 586 587 OOPort=25 588 589 in your .cf file. 590 591 You may have to use -DNeXT. 592 593BSDI (BSD/386) 1.0, NetBSD 0.9, FreeBSD 1.0 594 The "m4" from BSDI won't handle the config files properly. 595 I haven't had a chance to test this myself. 596 597 The M4 shipped in FreeBSD and NetBSD 0.9 don't handle the config 598 files properly. One must use either GNU m4 1.1 or the PD-M4 599 recently posted in comp.os.386bsd.bugs (and maybe others). 600 NetBSD-current includes the PD-M4 (as stated in the NetBSD file 601 CHANGES). 602 603 FreeBSD 1.0 RELEASE has uname(2) now. Use -DUSEUNAME in order to 604 use it (look into Makefile.FreeBSD). NetBSD-current may have 605 it too but it has not been verified. 606 607 You cannot port the latest version of the Berkeley db library 608 and use it with sendmail without recompiling the world. This 609 is because C library routines use the older version which have 610 incompatible header files -- the result is that it can't read 611 other system files, such as /etc/passwd, unless you use the 612 new db format throughout your system. You should normally just 613 use the version of db supplied in your release. You may need 614 to use -DOLD_NEWDB=1 to make this work -- this turns off some 615 new interface calls (for file locking) that are not in older 616 versions of db. You'll get compile errors if you need this 617 flag and don't have it set. 618 6194.3BSD 620 If you are running a "virgin" version of 4.3BSD, you'll have 621 a very old resolver and be missing some header files. The 622 header files are simple -- create empty versions and everything 623 will work fine. For the resolver you should really port a new 624 version (4.8.3 or later) of the resolver; 4.9 is available on 625 gatekeeper.DEC.COM in pub/BSD/bind/4.9. If you are really 626 determined to continue to use your old, buggy version (or as 627 a shortcut to get sendmail working -- I'm sure you have the 628 best intentions to port a modern version of BIND), you can 629 copy ../contrib/oldbind.compat.c into src and add 630 oldbind.compat.o to OBJADD in the Makefile. 631 632A/UX 633 Date: Tue, 12 Oct 1993 18:28:28 -0400 (EDT) 634 From: "Eric C. Hagberg" <hagberg@med.cornell.edu> 635 Subject: Fix for A/UX ndbm 636 637 I guess this isn't really a sendmail bug, however, it is something 638 that A/UX users should be aware of when compiling sendmail 8.6. 639 640 Apparently, the calls that sendmail is using to the ndbm routines 641 in A/UX 3.0.x contain calls to "broken" routines, in that the 642 aliases database will break when it gets "just a little big" 643 (sorry I don't have exact numbers here, but it broke somewhere 644 around 20-25 aliases for me.), making all aliases non-functional 645 after exceeding this point. 646 647 What I did was to get the gnu-dbm-1.6 package, compile it, and 648 then re-compile sendmail with "-lgdbm", "-DNDBM", and using the 649 ndbm.h header file that comes with the gnu-package. This makes 650 things behave properly. 651 652 I suppose porting the New Berkeley db package is another route, 653 however, I made a quick attempt at it, and found it difficult 654 (not easy at least); the gnu-dbm package "configured" and 655 compiled easily. 656 657SCO Unix 658 From: Thomas Essebier <tom@stallion.oz.au> 659 Organisation: Stallion Technologies Pty Ltd. 660 661 It will probably help those who are trying to configure sendmail 8.6.9 662 to know that if they are on SCO, they had better set 663 OI-dnsrch 664 or they will core dump as soon as they try to use the resolver. 665 ie. although SCO has _res.dnsrch defined, and is kinda BIND 4.8.3, it 666 does not inititialise it, nor does it understand 'search' in 667 /etc/named.boot. 668 - sigh - 669 670DG/UX 671 Doug Anderson <dlander@afterlife.ncsc.mil> has successfully run 672 V8 on the DG/UX 5.4.2 and 5.4R3.x platforms under heavy usage. 673 Originally, the DG /bin/mail program wasn't compatible with 674 the V8 sendmail, since the DG /bin/mail requires the environment 675 variable "_FORCE_MAIL_LOCAL_=yes" be set. Version 8.7 now includes 676 this in the environment before invoking the local mailer. Some 677 have used procmail to avoid this problem in the past. It works 678 but some have experienced file locking problems with their DG/UX 679 ports of procmail. 680 681Apollo DomainOS 682 If you are compiling on Apollo, you will have to create an empty 683 file "unistd.h" and create a file "dirent.h" containing: 684 685 #include <sys/dir.h> 686 #define dirent direct 687 688 (The Makefile.DomainOS will attempt to do both of these for you.) 689 690HP-UX 8.00 691 Date: Mon, 24 Jan 1994 13:25:45 +0200 692 From: Kimmo Suominen <Kimmo.Suominen@lut.fi> 693 Subject: 8.6.5 w/ HP-UX 8.00 on s300 694 695 Just compiled and fought with sendmail 8.6.5 on a HP9000/360 (ie. a 696 series 300 machine) running HP-UX 8.00. 697 698 I was getting segmentation fault when delivering to a local user. 699 With debugging I saw it was faulting when doing _free@libc... *sigh* 700 It seems the new implementation of malloc on s300 is buggy as of 8.0, 701 so I tried out the one in -lmalloc (malloc(3X)). With that it seems 702 to work just dandy. 703 704 When linking, you will get the following error: 705 706 ld: multiply defined symbol _freespace in file /usr/lib/libmalloc.a 707 708 but you can just ignore it. You might want to add this info to the 709 README file for the future... 710 711Linux 712 Something broke between versions 0.99.13 and 0.99.14 of Linux: 713 the flock() system call gives errors. If you are running .14, 714 you must not use flock. You can do this with -DHASFLOCK=0. 715 716AIX 717 This version of sendmail does not support MB, MG, and MR resource 718 records, which are supported by AIX sendmail. 719 720RISC/os 721 RISC/os from MIPS is a merged AT&T/Berkeley system. When you 722 compile on that platform you will get duplicate definitions 723 on many files. You can ignore these. 724 725System V Release 4 Based Systems 726 There is a single Makefile that is intended for all SVR4-based 727 systems (called Makefile.SVR4). It defines __svr4__, which is 728 predefined by some compilers. If your compiler already defines 729 this compile variable, you can delete the definition from the 730 Makefile. 731 732 It's been tested on Dell Issue 2.2. 733 734DELL SVR4 735 Date: Mon, 06 Dec 1993 10:42:29 EST 736 From: "Kimmo Suominen" <kim@grendel.lut.fi> 737 Message-ID: <2d0352f9.lento29@lento29.UUCP> 738 To: eric@cs.berkeley.edu 739 Cc: sendmail@cs.berkeley.edu 740 Subject: Notes for DELL SVR4 741 742 Eric, 743 744 Here are some notes for compiling Sendmail 8.6.4 on DELL SVR4. I ran 745 across these things when helping out some people who contacted me by 746 e-mail. 747 748 1) Use gcc 2.4.5 (or later?). Dell distributes gcc 2.1 with their 749 Issue 2.2 Unix. It is too old, and gives you problems with 750 clock.c, because sigset_t won't get defined in <sys/signal.h>. 751 This is due to a problematic protection rule in there, and is 752 fixed with gcc 2.4.5. 753 754 2) If you don't use the new Berkeley DB (-DNEWDB), then you need 755 to add "-lc -lucb" to the libraries to link with. This is because 756 the -ldbm distributed by Dell needs the bcopy, bcmp and bzero 757 functions. It is important that you specify both libraries in 758 the given order to be sure you only get the BSTRING functions 759 from the UCB library (and not the signal routines etc.). 760 761 3) Don't leave out "-lelf" even if compiling with "-lc -lucb". 762 The UCB library also has another copy of the nlist routines, 763 but we do want the ones from "-lelf". 764 765 If anyone needs a compiled gcc 2.4.5 and/or a ported DB library, they 766 can use anonymous ftp to fetch them from lut.fi in the /kim directory. 767 They are copies of what I use on grendel.lut.fi, and offering them 768 does not imply that I would also support them. I have sent the DB 769 port for SVR4 back to Keith Bostic for inclusion in the official 770 distribution, but I haven't heard anything from him as of today. 771 772 - gcc-2.4.5-svr4.tar.gz (gcc 2.4.5 and the corresponding libg++) 773 - db-1.72.tar.gz (with source, objects and a installed copy) 774 775 Cheers 776 + Kim 777 -- 778 * Kimmo.Suominen@lut.fi * SysVr4 enthusiast at GRENDEL.LUT.FI * 779 * KIM@FINFILES.BITNET * Postmaster and Hostmaster at LUT.FI * 780 * + 358 200 865 718 * Unix area moderator at NIC.FUNET.FI * 781 782ConvexOS 10.1 and below 783 In order to use the name server, you must create the file 784 /etc/use_nameserver. If this file does not exist, the call 785 to res_init() will fail and you will have absolutely no 786 access to DNS, including MX records. 787 788Amdahl UTS 2.1.5 789 In order to get UTS to work, you will have to port BIND 4.9. 790 The vendor's BIND is reported to be ``totally inadequate.'' 791 See sendmail/contrib/AmdahlUTS.patch for the patches necessary 792 to get BIND 4.9 compiled for UTS. 793 794Non-DNS based sites 795 This version of sendmail always tries to connect to the Domain 796 Name System (DNS) to resolve names, regardless of the setting 797 of the `I' option. On most systems that are not running DNS, 798 this will fail quickly and sendmail will continue, but on some 799 systems it has a long timeout. If you have this problem, you 800 will have to recompile without NAMED_BIND. Some people have 801 claimed that they have successfully used "OI+USEVC" to force 802 sendmail to use a virtual circuit -- this will always time out 803 quickly, but also tells sendmail that a failed connection 804 should requeue the message (probably not what you intended). 805 A future release of sendmail will correct this problem. 806 807Both NEWDB and NDBM 808 If you use both -DNDBM and -DNEWDB, you must delete the module 809 ndbm.o from libdb.a and delete the file "ndbm.h" from the files 810 that get installed (that is, use the OLD ndbm.h, not the new 811 ndbm.h). This compatibility module maps ndbm calls into DB 812 calls, and breaks things rather badly. 813 814GNU getopt 815 I'm told that GNU getopt has a problem in that it gets confused 816 by the double call. Use the version in conf.c instead. 817 818BIND 4.9.2 and Ultrix 819 If you are running on Ultrix, be sure you read conf/Info.Ultrix 820 in the BIND distribution very carefully -- there is information 821 in there that you need to know in order to avoid errors of the 822 form: 823 824 /lib/libc.a(gethostent.o): sethostent: multiply defined 825 /lib/libc.a(gethostent.o): endhostent: multiply defined 826 /lib/libc.a(gethostent.o): gethostbyname: multiply defined 827 /lib/libc.a(gethostent.o): gethostbyaddr: multiply defined 828 829 during the link stage. 830 831 832+--------------+ 833| MANUAL PAGES | 834+--------------+ 835 836The manual pages have been written against the -mandoc macros 837instead of the -man macros. The latest version of groff has them 838included. You can also get a copy from FTP.UU.NET in directory 839/systems/unix/bsd-sources/share/tmac. 840 841 842+-----------------+ 843| DEBUGGING HOOKS | 844+-----------------+ 845 846As of 8.6.5, sendmail daemons will catch a SIGUSR1 signal and log 847some debugging output (logged at LOG_DEBUG severity). The 848information dumped is: 849 850 * The value of the $j macro. 851 * A warning if $j is not in the set $=w. 852 * A list of the open file descriptors. 853 * The contents of the connection cache. 854 * If ruleset 89 is defined, it is evaluated and the results printed. 855 856This allows you to get information regarding the runtime state of the 857daemon on the fly. This should not be done too frequently, since 858the process of rewriting may lose memory which will not be recovered. 859Also, ruleset 89 may call non-reentrant routines, so there is a small 860non-zero probability that this will cause other problems. It is 861really only for debugging serious problems. 862 863A typical formulation of ruleset 89 would be: 864 865 R$* $@ $>0 some test address 866 867 868+-----------------------------+ 869| DESCRIPTION OF SOURCE FILES | 870+-----------------------------+ 871 872The following list describes the files in this directory: 873 874Makefile The makefile used here; this version only works with 875 the new Berkeley make. 876Makefile.dist A trimmed down version of the makefile that works with 877 the old make. 878READ_ME This file. 879TRACEFLAGS My own personal list of the trace flags -- not guaranteed 880 to be particularly up to date. 881alias.c Does name aliasing in all forms. 882arpadate.c A subroutine which creates ARPANET standard dates. 883clock.c Routines to implement real-time oriented functions 884 in sendmail -- e.g., timeouts. 885collect.c The routine that actually reads the mail into a temp 886 file. It also does a certain amount of parsing of 887 the header, etc. 888conf.c The configuration file. This contains information 889 that is presumed to be quite static and non- 890 controversial, or code compiled in for efficiency 891 reasons. Most of the configuration is in sendmail.cf. 892conf.h Configuration that must be known everywhere. 893convtime.c A routine to sanely process times. 894daemon.c Routines to implement daemon mode. This version is 895 specifically for Berkeley 4.1 IPC. 896deliver.c Routines to deliver mail. 897domain.c Routines that interface with DNS (the Domain Name 898 System). 899err.c Routines to print error messages. 900envelope.c Routines to manipulate the envelope structure. 901headers.c Routines to process message headers. 902macro.c The macro expander. This is used internally to 903 insert information from the configuration file. 904main.c The main routine to sendmail. This file also 905 contains some miscellaneous routines. 906map.c Support for database maps. 907mci.c Routines that handle mail connection information caching. 908parseaddr.c The routines which do address parsing. 909queue.c Routines to implement message queueing. 910readcf.c The routine that reads the configuration file and 911 translates it to internal form. 912recipient.c Routines that manipulate the recipient list. 913savemail.c Routines which save the letter on processing errors. 914sendmail.h Main header file for sendmail. 915srvrsmtp.c Routines to implement server SMTP. 916stab.c Routines to manage the symbol table. 917stats.c Routines to collect and post the statistics. 918sysexits.c List of error messages associated with error codes 919 in sysexits.h. 920trace.c The trace package. These routines allow setting and 921 testing of trace flags with a high granularity. 922udb.c The user database interface module. 923usersmtp.c Routines to implement user SMTP. 924util.c Some general purpose routines used by sendmail. 925version.c The version number and information about this 926 version of sendmail. Theoretically, this gets 927 modified on every change. 928 929Eric Allman 930 931(Version 8.78, last update 02/06/95 07:21:13) 932