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 8.69 (Berkeley) 07/24/94 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 16The Makefile is for the new (4.4BSD) Berkeley make and uses syntax 17that is not recognized by older makes. It also has assumptions 18about the 4.4 file system layout built in. See below for details 19about other Makefiles. 20 21There is also a Makefile.dist which is much less clever, but works on 22the old traditional make. You can use this using: 23 24 make -f Makefile.dist 25 26************************************************** 27** Read below for more details of Makefiles. ** 28************************************************** 29 30There is also a shell script (makesendmail) that tries to be clever 31about using object subdirectories. It's pretty straightforward, and 32may help if you share a source tree among different architectures. 33 34************************************************************************** 35** IMPORTANT: DO NOT USE OPTIMIZATION (``-O'') IF YOU ARE RUNNING ** 36** GCC 2.4.x or 2.5.x. THERE IS A BUG IN THE GCC OPTIMIZER THAT ** 37** CAUSES SENDMAIL COMPILES TO FAIL MISERABLY. ** 38************************************************************************** 39 40Jim Wilson of Cygnus believes he has found the problem -- it will 41probably be fixed in GCC 2.5.6 -- but until this is verified, be 42very suspicious of gcc -O. 43 44************************************************************************** 45** IMPORTANT: Read the appropriate paragraphs in the section on ** 46** ``Operating System and Compile Quirks''. ** 47************************************************************************** 48 49 50+-----------+ 51| MAKEFILES | 52+-----------+ 53 54The "Makefile"s in these directories are from 4.4 BSD, and hence 55really only work properly if you are on a 4.4 system. In particular, 56they use new syntax that will not be recognized on old make programs, 57and some of them do things like ``.include ../../Makefile.inc'' to 58pick up some system defines. If you are getting sendmail separately, 59these files won't be included in the distribution, as they are 60outside of the sendmail tree. 61 62Instead, you should use one of the other Makefiles, such as 63Makefile.SunOS for a SunOS system, and so forth. These should 64work with the version of make that is appropriate for that 65system. 66 67There are a bunch of other Makefiles for other systems with names 68like Makefile.HPUX for an HP-UX system. They use the version of 69make that is native for that system. These are the Makefiles that 70I use, and they have "Berkeley quirks" in them. I can't guarantee 71that they will work unmodified in your environment. Many of them 72include -I/usr/sww/include/db and -L/usr/sww/lib -- this is Berkeley's 73location (the ``Software Warehouse'') for the new database libraries, 74described below. You don't have to remove these definitions if you 75don't have these directories. 76 77Please look for an appropriate Makefile before you start trying to 78compile with Makefile or Makefile.dist. 79 80If you want to port the new Berkeley make, you can get it from 81ftp.uu.net in the directory /systems/unix/bsd-sources/usr.bin/make. 82Diffs and instructions for building this version of make under 83SunOS 4.1.x are available on ftp.css.itd.umich.edu in 84/pub/systems/sun/Net2-make.sun4.diff.Z. Diffs and instructions 85for building this version of make under IBM AIX 3.2.4 are available 86on ftp.uni-stuttgart.de in /sw/src/patches/bsd-make-rus-patches. 87For Ultrix, try ftp.vix.com:~ftp/pub/patches/pmake-for-ultrix.Z. 88Paul Southworth <pauls@umich.edu> published a description of porting 89this make in comp.unix.bsd. 90 91The complete text of the Makefile.inc that is in the parent of the 92sendmail directory is: 93 94 # @(#)Makefile.inc 8.1 (Berkeley) 6/6/93 95 96 BINDIR?= /usr/sbin 97 98 99+----------------------+ 100| DATABASE DEFINITIONS | 101+----------------------+ 102 103There are several database formats that can be used for the alias files 104and for general maps. When used for alias files they interact in an 105attempt to be back compatible. 106 107The three options are NEWDB (the new Berkeley DB package), NDBM (the 108older DBM implementation -- the very old V7 implementation is no 109longer supported), and NIS (Network Information Services). Used alone 110these just include the support they indicate. [If you are using NEWDB, 111get the latest version from FTP.CS.Berkeley.EDU in /ucb/4bsd. DO NOT 112use the version from the Net2 distribution! However, if you are on 113BSD/386 or 386BSD-based systems, use the one that already exists 114on your system. You may need to #define OLD_NEWDB 1 to do this.] 115 116[NOTE WELL: it is CRITICAL that you remove ndbm.o from libdb.a and 117ndbm.h from the appropriate include directories if you want to get 118ndbm support. These files OVERRIDE calls to ndbm routines -- in 119particular, if you leave ndbm.h in, you can find yourself using 120the new db package even if you don't define NEWDB.] 121 122If NEWDB and NDBM are defined (but not NIS), then sendmail will read 123NDBM format alias files, but the next time a newaliases is run the 124format will be converted to NEWDB; that format will be used forever 125more. This is intended as a transition feature. [Note however that 126the NEWDB library also catches and maps NDBM calls; you will have to 127back out this feature to get this to work. See ``Quirks'' section 128below for details.] 129 130If all three are defined, sendmail operates as described above, and also 131looks for the file /var/yp/Makefile. If it exists, newaliases will 132build BOTH the NEWDB and NDBM format alias files. However, it will 133only use the NEWDB file; the NDBM format file is used only by the 134NIS subsystem. 135 136If NDBM and NIS are defined (regardless of the definition of NEWDB 137or the existance of /var/yp/Makefile), sendmail adds the special 138tokens "YP_LAST_MODIFIED" and "YP_MASTER_NAME", both of which are 139required if the NDBM file is to be used as an NIS map. 140 141All of -DNEWDB, -DNDBM, and -DNIS are normally defined in the DBMDEF 142line in the Makefile. 143 144 145+---------------+ 146| COMPILE FLAGS | 147+---------------+ 148 149Whereever possible, I try to make sendmail pull in the correct 150compilation options needed to compile on various environments based on 151automatically defined symbols. Some machines don't seem to have useful 152symbols availble, requiring the following compilation flags in the 153Makefile: 154 155SOLARIS Define this if you are running Solaris 2.0 or higher. 156SOLARIS_2_3 Define this if you are running Solaris 2.3 or higher. 157SUNOS403 Define this if you are running SunOS 4.0.3. 158NeXT Define this if you are on a NeXT box. (This one may 159 be pre-defined for you.) There are other hacks you 160 have to make -- see below. 161_AIX3 Define this if you are IBM AIX 3.x. 162RISCOS Define this if you are running RISC/os from MIPS. 163IRIX Define this if you are running IRIX from SGI. 164_SCO_unix_ Define this if you are on SCO UNIX. 165_SCO_unix_4_2 Define this if you are on SCO Open Server 3.2v4. 166DGUX Define this if you are on DG/UX 5.4.3 or later 167DGUX_5_4_2 Define this if you are on DG/UX systems prior to 5.4.3. 168NonStop_UX_BXX Define this if you are on a Tandem NonStop-UX release 169 Bxx system. 170 171If you are a system that sendmail has already been ported to, you 172probably won't have to touch these. But if you are porting, you may 173have to tweak the following compilation flags in conf.h in order to 174get it to compile and link properly: 175 176SYSTEM5 Adjust for System V (not necessarily Release 4). 177SYS5SIGNALS Use System V signal semantics -- the signal handler 178 is automatically dropped when the signal is caught. 179 If this is not set, use POSIX/BSD semantics, where the 180 signal handler stays in force until an exec or an 181 explicit delete. Implied by SYSTEM5. 182SYS5SETPGRP Use System V setpgrp() semantics. Implied by SYSTEM5. 183HASFLOCK Set this if you prefer to use the flock(2) system call 184 rather than using fcntl-based locking. Fcntl locking 185 has some semantic gotchas, but many vendor systems 186 also interface it to lockd(8) to do NFS-style locking. 187 For this reason, this should not be set unless you 188 don't have an alternative. 189HASUNAME Set if you have the "uname" system call. Implied by 190 SYSTEM5. 191HASUNSETENV Define this if your system library has the "unsetenv" 192 subroutine. 193HASSETSID Define this if you have the setsid(2) system call. This 194 is implied if your system appears to be POSIX compliant. 195HASINITGROUPS Define this if you have the initgroups(3) routine. 196HASSETVBUF Define this if you have the setvbuf(3) library call. 197 If you don't, setlinebuf will be used instead. This 198 defaults on if your compiler defines __STDC__. 199HASSETREUID Define this if you have setreuid(2) ***AND*** root can 200 use setreuid to change to an arbitrary user. This second 201 condition is not satisfied on AIX 3.x. You may find that 202 your system has setresuid(2), (for example, on HP-UX) in 203 which case you will also have to #define setreuid(r, e) 204 to be the appropriate call. Some systems (such as Solaris) 205 have a compatibility routine that doesn't work properly, 206 but may have "saved user ids" properly implemented so you 207 can ``#define setreuid(r, e) seteuid(e)'' and have it work. 208 The important thing is that you have a call that will set 209 the effective uid independently of the real or saved uid 210 and be able to set the effective uid back again when done. 211 There's a test program in ../test/t_setreuid.c that will 212 try things on your system. Setting this improves the 213 security, since sendmail doesn't have to read .forward 214 and :include: files as root. There are certain attacks 215 that may be unpreventable without this call. 216HASLSTAT Define this if you have symbolic links (and thus the 217 lstat(2) system call). This improves security. Unlike 218 most other options, this one is on by default, so you 219 need to #undef it in conf.h if you don't have symbolic 220 links (these days everyone does). 221HASSETRLIMIT Define this to 1 if you have the setrlimit(2) syscall. 222 You can define it to 0 to force it off. It is assumed 223 if you are running a BSD-like system. 224HASULIMIT Define this if you have the ulimit(2) syscall (System V 225 style systems). HASSETRLIMIT overrides, as it is more 226 general. 227NEEDGETOPT Define this if you need a reimplementation of getopt(3). 228 On some systems, getopt does very odd things if called 229 to scan the arguments twice. This flag will ask sendmail 230 to compile in a local version of getopt that works 231 properly. 232NEEDSTRTOL Define this if your standard C library does not define 233 strtol(3). This will compile in a local version. 234NEEDVPRINTF Define this if your standard C library does not define 235 vprintf(3). Note that the resulting fake implementation 236 is not very elegant and may not even work on some 237 architectures. 238NEEDFSYNC Define this if your standard C library does not define 239 fsync(2). This will try to simulate the operation using 240 fcntl(2); if that is not available it does nothing, which 241 isn't great, but at least it compiles and runs. 242HASGETUSERSHELL Define this to 1 if you have getusershell(3) in your 243 standard C library. If this is not defined, or is defined 244 to be 0, sendmail will scan the /etc/shells file (no 245 NIS-style support, defaults to /bin/sh and /bin/csh if 246 that file does not exist) to get a list of unrestricted 247 user shells. This is used to determine whether users 248 are allowed to forward their mail to a program or a file. 249GIDSET_T The type of entries in a gidset passed as the second 250 argument to getgroups(2). Historically this has been an 251 int, so this is the default, but some systems (such as 252 IRIX) pass it as a gid_t, which is an unsigned short. 253 This will make a difference, so it is important to get 254 this right! However, it is only an issue if you have 255 group sets. 256SLEEP_T The type returned by the system sleep() function. 257 Defaults to "unsigned int". Don't worry about this 258 if you don't have compilation problems. 259ARBPTR_T The type of an arbitrary pointer -- defaults to "void *". 260 If you are an very old compiler you may need to define 261 this to be "char *". 262LA_TYPE The type of load average your kernel supports. These 263 can be one of: 264 LA_ZERO (1) -- it always returns the load average as 265 "zero" (and does so on all architectures). 266 LA_SUBR (4) if you have the getloadavg(3) routine, 267 LA_MACH (5) to use MACH-style load averages (calls 268 processor_set_info()), 269 LA_PROCSTR (7) to read /proc/loadavg and interpret it 270 as a string representing a floating-point 271 number (Linux-style), 272 LA_FLOAT (3) if you read kmem and interpret the value 273 as a floating point number, 274 LA_INT (2) to interpret as a long integer, 275 LA_SHORT (6) to interpret as a short integer. 276 These last three have several other parameters that they 277 try to divine: the name of your kernel, the name of the 278 variable in the kernel to examine, the number of bits of 279 precision in a fixed point load average, and so forth. 280 In desperation, use LA_ZERO. The actual code is in 281 conf.c -- it can be tweaked if you are brave. 282SFS_TYPE Encodes how your kernel can locate the amount of free 283 space on a disk partition. This can be set to SFS_NONE 284 (0) if you have no way of getting this information, 285 SFS_USTAT (1) if you have the ustat(2) system call, 286 SFS_4ARGS (2) if you have a four-argument statfs(2) 287 system call (and the include file is <sys/statfs.h>), 288 SFS_VFS (3), SFS_MOUNT (4), SFS_STATFS (5) if you have 289 the two-argument statfs(2) system call with includes in 290 <sys/vfs.h>, <sys/mount.h>, or <sys/statfs.h> respectively, 291 or SFS_STATVFS (6) if you have the two-argument statvfs(2) 292 call. The default if nothing is defined is SFS_NONE. 293ERRLIST_PREDEFINED 294 If set, assumes that some header file defines sys_errlist. 295 This may be needed if you get type conflicts on this 296 variable -- otherwise don't worry about it. 297WAITUNION The wait(2) routine takes a "union wait" argument instead 298 of an integer argument. This is for compatibility with 299 old versions of BSD. 300SCANF You can set this to extend the F command to accept a 301 scanf string -- this gives you a primitive parser for 302 class definitions -- BUT it can make you vulnerable to 303 core dumps if the target file is poorly formed. 304SYSLOG_BUFSIZE You can define this to be the size of the buffer that 305 syslog accepts. If it is not defined, it assumes a 306 1024-byte buffer. If the buffer is very small (under 307 256 bytes) the log message format changes -- each 308 e-mail message will log many more messages, since it 309 will log each piece of information as a separate line 310 in syslog. 311BROKEN_RES_SEARCH 312 On Ultrix (and maybe other systems?) if you use the 313 res_search routine with an unknown host name, it returns 314 -1 but sets h_errno to 0 instead of HOST_NOT_FOUND. If 315 you set this, sendmail considers 0 to be the same as 316 HOST_NOT_FOUND. 317NAMELISTMASK If defined, values returned by nlist(3) are masked 318 against this value before use -- a common value is 319 0x7fffffff to strip off the top bit. 320 321 322 323+-----------------------+ 324| COMPILE-TIME FEATURES | 325+-----------------------+ 326 327There are a bunch of features that you can decide to compile in, such 328as selecting various database packages and special protocol support. 329Several are assumed based on other compilation flags -- if you want to 330"un-assume" something, you probably need to edit conf.h. Compilation 331flags that add support for special features include: 332 333NDBM Include support for "new" DBM library for aliases and maps. 334 Normally defined in the Makefile. 335NEWDB Include support for Berkeley "db" package (hash & btree) 336 for aliases and maps. Normally defined in the Makefile. 337OLD_NEWDB If non-zero, the version of NEWDB you have is the old 338 one that does not include the "fd" call. This call was 339 added in version 1.5 of the Berkeley DB code. If you 340 use -DOLD_NEWDB=0 it forces you to use the new interface. 341NIS Define this to get NIS (YP) support for aliases and maps. 342 Normally defined in the Makefile. 343USERDB Include support for the User Information Database. Implied 344 by NEWDB in conf.h. 345IDENTPROTO Define this as 1 to get IDENT (RFC 1413) protocol support. 346 This is assumed unless you are running on Ultrix or 347 HP-UX, both of which have a problem in the UDP 348 implementation. You can define it to be 0 to explicitly 349 turn off IDENT protocol support. 350MIME Include support for MIME-encapsulated error messages. 351LOG Set this to get syslog(3) support. Defined by default 352 in conf.h. You want this if at all possible. 353NETINET Set this to get TCP/IP support. Defined by default 354 in conf.h. You probably want this. 355NETISO Define this to get ISO networking support. 356SMTP Define this to get the SMTP code. Implied by NETINET 357 or NETISO. 358NAMED_BIND Define this to get DNS (name daemon) support, including 359 MX support. The specs you must use this if you run 360 SMTP. Defined by default in conf.h. 361QUEUE Define this to get queueing code. Implied by NETINET 362 or NETISO; required by SMTP. This gives you other good 363 stuff -- it should be on. 364DAEMON Define this to get general network support. Implied by 365 NETINET or NETISO. Defined by default in conf.h. You 366 almost certainly want it on. 367MATCHGECOS Permit fuzzy matching of user names against the full 368 name (GECOS) field in the /etc/passwd file. This should 369 probably be on, since you can disable it from the config 370 file if you want to. Defined by default in conf.h. 371SETPROCTITLE Try to set the string printed by "ps" to something 372 informative about what sendmail is doing. Defined by 373 default in conf.h. 374 375 376+---------------------+ 377| DNS/RESOLVER ISSUES | 378+---------------------+ 379 380Many systems have old versions of the resolver library. At a minimum, 381you should be running BIND 4.8.3; older versions may compile, but they 382have known bugs that should give you pause. 383 384Common problems in old versions include "undefined" errors for 385dn_skipname. 386 387Some people have had a problem with BIND 4.9; it uses some routines 388that it expects to be externally defined such as strerror(). It may 389help to link with "-l44bsd" to solve this problem. 390 391!PLEASE! be sure to link with the same version of the resolver as 392the header files you used -- some people have used the 4.9 headers 393and linked with BIND 4.8 or vice versa, and it doesn't work. 394Unfortunately, it doesn't fail in an obvious way -- things just 395subtly don't work. 396 397 398+-------------------------------------+ 399| OPERATING SYSTEM AND COMPILE QUIRKS | 400+-------------------------------------+ 401 402GCC 2.5.x problems *** IMPORTANT *** 403 Date: Mon, 29 Nov 93 19:08:44 PST 404 From: wilson@cygnus.com (Jim Wilson) 405 Message-Id: <9311300308.AA04608@cygnus.com> 406 To: kenner@vlsi1.ultra.nyu.edu 407 Subject: [cattelan@thebarn.com: gcc 2.5.4-2.5.5 -O bug] 408 Cc: cattelan@thebarn.com, rms@gnu.ai.mit.edu, sendmail@cs.berkeley.edu 409 410 This fixes a problem that occurs when gcc 2.5.5 is used to compile 411 sendmail 8.6.4 with optimization on a sparc. 412 413 Mon Nov 29 19:00:14 1993 Jim Wilson (wilson@sphagnum.cygnus.com) 414 415 * reload.c (find_reloads_toplev): Replace obsolete reference to 416 BYTE_LOADS_*_EXTEND with LOAD_EXTEND_OP. 417 418 *** clean-ss-931128/reload.c Sun Nov 14 16:20:01 1993 419 --- ss-931128/reload.c Mon Nov 29 18:52:55 1993 420 *************** find_reloads_toplev (x, opnum, type, ind 421 *** 3888,3894 **** 422 force a reload in that case. So we should not do anything here. */ 423 424 else if (regno >= FIRST_PSEUDO_REGISTER 425 ! #if defined(BYTE_LOADS_ZERO_EXTEND) || defined(BYTE_LOADS_SIGN_EXTEND) 426 && (GET_MODE_SIZE (GET_MODE (x)) 427 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 428 #endif 429 --- 3888,3894 ---- 430 force a reload in that case. So we should not do anything here. */ 431 432 else if (regno >= FIRST_PSEUDO_REGISTER 433 ! #ifdef LOAD_EXTEND_OP 434 && (GET_MODE_SIZE (GET_MODE (x)) 435 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 436 #endif 437 438 439SunOS 4.x (Solaris 1.x) 440 You may have to use -lresolv on SunOS. However, beware that 441 this links in a new version of gethostbyname that does not 442 understand NIS, so you must have all of your hosts in DNS. 443 444 Some people have reported problems with the SunOS version of 445 -lresolv and/or in.named, and suggest that you get a newer 446 version. The symptoms are delays when you connect to the 447 SMTP server on a SunOS machine or having your domain added to 448 addresses inappropriately. There is a version of BIND 449 version 4.9 on gatekeeper.DEC.COM in pub/BSD/bind/4.9. 450 451 There is substantial disagreement about whether you can make 452 this work with resolv+, which allows you to specify a search-path 453 of services. Some people report that it works fine, others 454 claim it doesn't work at all (including causing sendmail to 455 drop core when it tries to do multiple resolv+ lookups for a 456 single job). I haven't tried resolv+, as we use DNS exclusively. 457 458 Should you want to try resolv+, it is on ftp.uu.net in 459 /networking/ip/dns. 460 461 Apparently getservbyname() can fail under moderate to high 462 load under some circumstances. This will exhibit itself as 463 the message ``554 makeconnection: service "smtp" unknown''. 464 The problem has been traced to one or more blank lines in 465 /etc/services on the NIS server machine. Delete these 466 and it should work. This info is thanks to Brian Bartholomew 467 <bb@math.ufl.edu> of I-Kinetics, Inc. 468 469Solaris 2.x (SunOS 5.x) 470 To compile for Solaris, be sure you use -DSOLARIS. 471 472 To the best of my knowledge, Solaris does not have the 473 gethostbyname problem described above. However, it does 474 have another one: 475 476 From a correspondent: 477 478 For solaris 2.2, I have 479 480 hosts: files dns 481 482 in /etc/nsswitch.conf and /etc/hosts has to have the fully 483 qualified host name. I think "files" has to be before "dns" 484 in /etc/nsswitch.conf during bootup. 485 486 From another correspondent: 487 488 When running sendmail under Solaris, the gethostbyname() 489 hack in conf.c which should perform proper canonicalization 490 of host names could fail. Result: the host name is not 491 canonicalized despite the hack, and you'll have to define $j 492 and $m in sendmail.cf somewhere. 493 494 The reason could be that /etc/nsswitch.conf is improperly 495 configured (at least from sendmail's point of view). For 496 example, the line 497 498 hosts: files nisplus dns 499 500 will make gethostbyname() look in /etc/hosts first, then ask 501 nisplus, then dns. However, if /etc/hosts does not contain 502 the full canonicalized hostname, then no amount of 503 gethostbyname()s will work. 504 505 Solution (or rather, a workaround): Ask nisplus first, then 506 dns, then local files: 507 508 hosts: nisplus dns [NOTFOUND=return] files 509 510 The Solaris "syslog" function is apparently limited to something 511 about 90 characters because of a kernel limitation. If you have 512 source code, you can probably up this number. You can get patches 513 that fix this problem: the patch ids are: 514 515 Solaris 2.1 100834 516 Solaris 2.2 100999 517 Solaris 2.3 101318 518 519 Be sure you have the appropriate patch installed or you won't 520 see system logging. 521 522OSF/1 523 If you are compiling on OSF/1 (DEC Alpha), you must use 524 -L/usr/shlib (otherwise it core dumps on startup). You may also 525 need -mld to get the nlist() function, although some versions 526 apparently don't need this. 527 528 Also, the enclosed makefile removed /usr/sbin/smtpd; if you need 529 it, just create the link to the sendmail binary. 530 531IRIX 532 The header files on SGI IRIX are completely prototyped, and as 533 a result you can sometimes get some warning messages during 534 compilation. These can be ignored. There are two errors in 535 deliver only if you are using gcc, both of the form ``warning: 536 passing arg N of `execve' from incompatible pointer type''. 537 Also, if you compile with -DNIS, you will get a complaint 538 about a declaration of struct dom_binding in a prototype 539 when compiling map.c; this is not important because the 540 function being prototyped is not used in that file. 541 542NeXT 543 If you are compiling on NeXT, you will have to create an empty 544 file "unistd.h" and create a file "dirent.h" containing: 545 546 #include <sys/dir.h> 547 #define dirent direct 548 549 (The Makefile.NeXT should try to do both of these for you.) 550 551 Apparently, there is a bug in getservbyname on Nextstep 3.0 552 that causes it to fail under some circumstances with the 553 message "SYSERR: service "smtp" unknown" logged. You should 554 be able to work around this by including the line: 555 556 OOPort=25 557 558 in your .cf file. 559 560 You may have to use -DNeXT. 561 562BSDI (BSD/386) 1.0, NetBSD 0.9, FreeBSD 1.0 563 The "m4" from BSDI won't handle the config files properly. 564 I haven't had a chance to test this myself. 565 566 The M4 shipped in FreeBSD and NetBSD 0.9 don't handle the config 567 files properly. One must use either GNU m4 1.1 or the PD-M4 568 recently posted in comp.os.386bsd.bugs (and maybe others). 569 NetBSD-current includes the PD-M4 (as stated in the NetBSD file 570 CHANGES). 571 572 FreeBSD 1.0 RELEASE has uname(2) now. Use -DUSEUNAME in order to 573 use it (look into Makefile.FreeBSD). NetBSD-current may have 574 it too but it has not been verified. 575 576 You cannot port the latest version of the Berkeley db library 577 and use it with sendmail without recompiling the world. This 578 is because C library routines use the older version which have 579 incompatible header files -- the result is that it can't read 580 other system files, such as /etc/passwd, unless you use the 581 new db format throughout your system. You should normally just 582 use the version of db supplied in your release. You may need 583 to use -DOLD_NEWDB=1 to make this work -- this turns off some 584 new interface calls (for file locking) that are not in older 585 versions of db. You'll get compile errors if you need this 586 flag and don't have it set. 587 5884.3BSD 589 If you are running a "virgin" version of 4.3BSD, you'll have 590 a very old resolver and be missing some header files. The 591 header files are simple -- create empty versions and everything 592 will work fine. For the resolver you should really port a new 593 version (4.8.3 or later) of the resolver; 4.9 is available on 594 gatekeeper.DEC.COM in pub/BSD/bind/4.9. If you are really 595 determined to continue to use your old, buggy version (or as 596 a shortcut to get sendmail working -- I'm sure you have the 597 best intentions to port a modern version of BIND), you can 598 copy ../contrib/oldbind.compat.c into src and add 599 oldbind.compat.o to OBJADD in the Makefile. 600 601A/UX 602 Date: Tue, 12 Oct 1993 18:28:28 -0400 (EDT) 603 From: "Eric C. Hagberg" <hagberg@med.cornell.edu> 604 Subject: Fix for A/UX ndbm 605 606 I guess this isn't really a sendmail bug, however, it is something 607 that A/UX users should be aware of when compiling sendmail 8.6. 608 609 Apparently, the calls that sendmail is using to the ndbm routines 610 in A/UX 3.0.x contain calls to "broken" routines, in that the 611 aliases database will break when it gets "just a little big" 612 (sorry I don't have exact numbers here, but it broke somewhere 613 around 20-25 aliases for me.), making all aliases non-functional 614 after exceeding this point. 615 616 What I did was to get the gnu-dbm-1.6 package, compile it, and 617 then re-compile sendmail with "-lgdbm", "-DNDBM", and using the 618 ndbm.h header file that comes with the gnu-package. This makes 619 things behave properly. 620 621 I suppose porting the New Berkeley db package is another route, 622 however, I made a quick attempt at it, and found it difficult 623 (not easy at least); the gnu-dbm package "configured" and 624 compiled easily. 625 626DG/UX 627 Apparently, /bin/mail doesn't work properly for delivery on 628 DG/UX -- the person who has this working, Douglas Anderson 629 <dlander@afterlife.ncsc.mil>, used procmail instead. The 630 problem is that DG/UX /bin/mail requires that an environment 631 variable be set (_FORCE_MAIL_LOCAL_=yes); sendmail has no 632 mechanism for this. Several people report that procmail works 633 beautifully. 634 635Apollo DomainOS 636 If you are compiling on Apollo, you will have to create an empty 637 file "unistd.h" and create a file "dirent.h" containing: 638 639 #include <sys/dir.h> 640 #define dirent direct 641 642 (The Makefile.DomainOS will attempt to do both of these for you.) 643 644HP-UX 8.00 645 Date: Mon, 24 Jan 1994 13:25:45 +0200 646 From: Kimmo Suominen <Kimmo.Suominen@lut.fi> 647 Subject: 8.6.5 w/ HP-UX 8.00 on s300 648 649 Just compiled and fought with sendmail 8.6.5 on a HP9000/360 (ie. a 650 series 300 machine) running HP-UX 8.00. 651 652 I was getting segmentation fault when delivering to a local user. 653 With debugging I saw it was faulting when doing _free@libc... *sigh* 654 It seems the new implementation of malloc on s300 is buggy as of 8.0, 655 so I tried out the one in -lmalloc (malloc(3X)). With that it seems 656 to work just dandy. 657 658 When linking, you will get the following error: 659 660 ld: multiply defined symbol _freespace in file /usr/lib/libmalloc.a 661 662 but you can just ignore it. You might want to add this info to the 663 README file for the future... 664 665Linux 666 Something broke between versions 0.99.13 and 0.99.14 of Linux: 667 the flock() system call gives errors. If you are running .14, 668 you must not use flock. You can do this with -DHASFLOCK=0. 669 670AIX 671 This version of sendmail does not support MB, MG, and MR resource 672 records, which are supported by AIX sendmail. 673 674RISC/os 675 RISC/os from MIPS is a merged AT&T/Berkeley system. When you 676 compile on that platform you will get duplicate definitions 677 on many files. You can ignore these. 678 679System V Release 4 Based Systems 680 There is a single Makefile that is intended for all SVR4-based 681 systems (called Makefile.SVR4). It defines __svr4__, which is 682 predefined by some compilers. If your compiler already defines 683 this compile variable, you can delete the definition from the 684 Makefile. 685 686 It's been tested on Dell Issue 2.2. 687 688DELL SVR4 689 Date: Mon, 06 Dec 1993 10:42:29 EST 690 From: "Kimmo Suominen" <kim@grendel.lut.fi> 691 Message-ID: <2d0352f9.lento29@lento29.UUCP> 692 To: eric@cs.berkeley.edu 693 Cc: sendmail@cs.berkeley.edu 694 Subject: Notes for DELL SVR4 695 696 Eric, 697 698 Here are some notes for compiling Sendmail 8.6.4 on DELL SVR4. I ran 699 across these things when helping out some people who contacted me by 700 e-mail. 701 702 1) Use gcc 2.4.5 (or later?). Dell distributes gcc 2.1 with their 703 Issue 2.2 Unix. It is too old, and gives you problems with 704 clock.c, because sigset_t won't get defined in <sys/signal.h>. 705 This is due to a problematic protection rule in there, and is 706 fixed with gcc 2.4.5. 707 708 2) If you don't use the new Berkeley DB (-DNEWDB), then you need 709 to add "-lc -lucb" to the libraries to link with. This is because 710 the -ldbm distributed by Dell needs the bcopy, bcmp and bzero 711 functions. It is important that you specify both libraries in 712 the given order to be sure you only get the BSTRING functions 713 from the UCB library (and not the signal routines etc.). 714 715 3) Don't leave out "-lelf" even if compiling with "-lc -lucb". 716 The UCB library also has another copy of the nlist routines, 717 but we do want the ones from "-lelf". 718 719 If anyone needs a compiled gcc 2.4.5 and/or a ported DB library, they 720 can use anonymous ftp to fetch them from lut.fi in the /kim directory. 721 They are copies of what I use on grendel.lut.fi, and offering them 722 does not imply that I would also support them. I have sent the DB 723 port for SVR4 back to Keith Bostic for inclusion in the official 724 distribution, but I haven't heard anything from him as of today. 725 726 - gcc-2.4.5-svr4.tar.gz (gcc 2.4.5 and the corresponding libg++) 727 - db-1.72.tar.gz (with source, objects and a installed copy) 728 729 Cheers 730 + Kim 731 -- 732 * Kimmo.Suominen@lut.fi * SysVr4 enthusiast at GRENDEL.LUT.FI * 733 * KIM@FINFILES.BITNET * Postmaster and Hostmaster at LUT.FI * 734 * + 358 200 865 718 * Unix area moderator at NIC.FUNET.FI * 735 736ConvexOS 10.1 and below 737 In order to use the name server, you must create the file 738 /etc/use_nameserver. If this file does not exist, the call 739 to res_init() will fail and you will have absolutely no 740 access to DNS, including MX records. 741 742Non-DNS based sites 743 This version of sendmail always tries to connect to the Domain 744 Name System (DNS) to resolve names, regardless of the setting 745 of the `I' option. On most systems that are not running DNS, 746 this will fail quickly and sendmail will continue, but on some 747 systems it has a long timeout. If you have this problem, you 748 will have to recompile without NAMED_BIND. Some people have 749 claimed that they have successfully used "OI+USEVC" to force 750 sendmail to use a virtual circuit -- this will always time out 751 quickly, but also tells sendmail that a failed connection 752 should requeue the message (probably not what you intended). 753 A future release of sendmail will correct this problem. 754 755Both NEWDB and NDBM 756 If you use both -DNDBM and -DNEWDB, you must delete the module 757 ndbm.o from libdb.a and delete the file "ndbm.h" from the files 758 that get installed (that is, use the OLD ndbm.h, not the new 759 ndbm.h). This compatibility module maps ndbm calls into DB 760 calls, and breaks things rather badly. 761 762GNU getopt 763 I'm told that GNU getopt has a problem in that it gets confused 764 by the double call. Use the version in conf.c instead. 765 766BIND 4.9.2 and Ultrix 767 If you are running on Ultrix, be sure you read conf/Info.Ultrix 768 in the BIND distribution very carefully -- there is information 769 in there that you need to know in order to avoid errors of the 770 form: 771 772 /lib/libc.a(gethostent.o): sethostent: multiply defined 773 /lib/libc.a(gethostent.o): endhostent: multiply defined 774 /lib/libc.a(gethostent.o): gethostbyname: multiply defined 775 /lib/libc.a(gethostent.o): gethostbyaddr: multiply defined 776 777 during the link stage. 778 779 780+--------------+ 781| MANUAL PAGES | 782+--------------+ 783 784The manual pages have been written against the -mandoc macros 785instead of the -man macros. The latest version of groff has them 786included. You can also get a copy from FTP.UU.NET in directory 787/systems/unix/bsd-sources/share/tmac. 788 789 790+-----------------+ 791| DEBUGGING HOOKS | 792+-----------------+ 793 794As of 8.6.5, sendmail daemons will catch a SIGUSR1 signal and log 795some debugging output (logged at LOG_DEBUG severity). The 796information dumped is: 797 798 * The value of the $j macro. 799 * A warning if $j is not in the set $=w. 800 * A list of the open file descriptors. 801 * The contents of the connection cache. 802 * If ruleset 89 is defined, it is evaluated and the results printed. 803 804This allows you to get information regarding the runtime state of the 805daemon on the fly. This should not be done too frequently, since 806the process of rewriting may lose memory which will not be recovered. 807Also, ruleset 89 may call non-reentrant routines, so there is a small 808non-zero probability that this will cause other problems. It is 809really only for debugging serious problems. 810 811A typical formulation of ruleset 89 would be: 812 813 R$* $@ $>0 some test address 814 815 816+-----------------------------+ 817| DESCRIPTION OF SOURCE FILES | 818+-----------------------------+ 819 820The following list describes the files in this directory: 821 822Makefile The makefile used here; this version only works with 823 the new Berkeley make. 824Makefile.dist A trimmed down version of the makefile that works with 825 the old make. 826READ_ME This file. 827TRACEFLAGS My own personal list of the trace flags -- not guaranteed 828 to be particularly up to date. 829alias.c Does name aliasing in all forms. 830arpadate.c A subroutine which creates ARPANET standard dates. 831clock.c Routines to implement real-time oriented functions 832 in sendmail -- e.g., timeouts. 833collect.c The routine that actually reads the mail into a temp 834 file. It also does a certain amount of parsing of 835 the header, etc. 836conf.c The configuration file. This contains information 837 that is presumed to be quite static and non- 838 controversial, or code compiled in for efficiency 839 reasons. Most of the configuration is in sendmail.cf. 840conf.h Configuration that must be known everywhere. 841convtime.c A routine to sanely process times. 842daemon.c Routines to implement daemon mode. This version is 843 specifically for Berkeley 4.1 IPC. 844deliver.c Routines to deliver mail. 845domain.c Routines that interface with DNS (the Domain Name 846 System). 847err.c Routines to print error messages. 848envelope.c Routines to manipulate the envelope structure. 849headers.c Routines to process message headers. 850macro.c The macro expander. This is used internally to 851 insert information from the configuration file. 852main.c The main routine to sendmail. This file also 853 contains some miscellaneous routines. 854map.c Support for database maps. 855mci.c Routines that handle mail connection information caching. 856parseaddr.c The routines which do address parsing. 857queue.c Routines to implement message queueing. 858readcf.c The routine that reads the configuration file and 859 translates it to internal form. 860recipient.c Routines that manipulate the recipient list. 861savemail.c Routines which save the letter on processing errors. 862sendmail.h Main header file for sendmail. 863srvrsmtp.c Routines to implement server SMTP. 864stab.c Routines to manage the symbol table. 865stats.c Routines to collect and post the statistics. 866sysexits.c List of error messages associated with error codes 867 in sysexits.h. 868trace.c The trace package. These routines allow setting and 869 testing of trace flags with a high granularity. 870udb.c The user database interface module. 871usersmtp.c Routines to implement user SMTP. 872util.c Some general purpose routines used by sendmail. 873version.c The version number and information about this 874 version of sendmail. Theoretically, this gets 875 modified on every change. 876 877Eric Allman 878 879(Version 8.69, last update 07/24/94 06:41:53) 880