1# Copyright (c) 1983, 1995 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 8.86 (Berkeley) 05/18/95 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 46This problem is reported to have been fixed in gcc 2.6. 47 48************************************************************************** 49** IMPORTANT: Read the appropriate paragraphs in the section on ** 50** ``Operating System and Compile Quirks''. ** 51************************************************************************** 52 53 54+-----------+ 55| MAKEFILES | 56+-----------+ 57 58By far, the easiest way to compile sendmail is to use the "makesendmail" 59script: 60 61 sh makesendmail 62 63This uses the "uname" command to figure out what architecture you are 64on and selects a proper Makefile accordingly. It also creates a 65subdirectory per object format, so that multiarchitecture support is 66easy. In general this should be all you need. However, if for some 67reason this doesn't work (e.g., NeXT systems don't have the "uname" 68command) you may have to set up your compile environment by hand. 69 70The "Makefile"s in these directories are from 4.4 BSD, and hence 71really only work properly if you are on a 4.4 system. In particular, 72they use new syntax that will not be recognized on old make programs, 73and some of them do things like ``.include ../../Makefile.inc'' to 74pick up some system defines. If you are getting sendmail separately, 75these files won't be included in the distribution, as they are 76outside of the sendmail tree. 77 78Instead, you should use one of the other Makefiles, such as 79Makefile.SunOS for a SunOS system, and so forth. These should 80work with the version of make that is appropriate for that 81system. All other Makefiles are in the "src/Makefiles" subdirectory. 82They use the version of make that is native for that system. These 83are the Makefiles that I use, and they have "Berkeley quirks" in them. 84I can't guarantee that they will work unmodified in your environment. 85In particular, Many of them include -I/usr/sww/include/db and 86-L/usr/sww/lib -- these are Berkeley's locations in the ``Software 87Warehouse'' for the new database libraries, described below. You don't 88have to remove these definitions if you don't have these directories, 89but you may have to remove -DNEWDB from the DBMDEF definition. 90 91Please look for an appropriate Makefile before you start trying to 92compile with Makefile or Makefile.dist. 93 94If you want to port the new Berkeley make, you can get it from 95ftp.uu.net in the directory /systems/unix/bsd-sources/usr.bin/make. 96Diffs and instructions for building this version of make under 97SunOS 4.1.x are available on ftp.css.itd.umich.edu in 98/pub/systems/sun/Net2-make.sun4.diff.Z. Diffs and instructions 99for building this version of make under IBM AIX 3.2.4 are available 100on ftp.uni-stuttgart.de in /sw/src/patches/bsd-make-rus-patches. 101For Ultrix, try ftp.vix.com:~ftp/pub/patches/pmake-for-ultrix.Z. 102Paul Southworth <pauls@umich.edu> published a description of porting 103this make in comp.unix.bsd. 104 105The complete text of the Makefile.inc that is in the parent of the 106sendmail directory is: 107 108 # @(#)Makefile.inc 8.1 (Berkeley) 6/6/93 109 110 BINDIR?= /usr/sbin 111 112 113+----------------------+ 114| DATABASE DEFINITIONS | 115+----------------------+ 116 117There are several database formats that can be used for the alias files 118and for general maps. When used for alias files they interact in an 119attempt to be back compatible. 120 121The three options are NEWDB (the new Berkeley DB package), NDBM (the 122older DBM implementation -- the very old V7 implementation is no 123longer supported), and NIS (Network Information Services). Used alone 124these just include the support they indicate. [If you are using NEWDB, 125get the latest version from FTP.CS.Berkeley.EDU in /ucb/4bsd/db.tar.Z 126(or db.tar.gz). DO NOT use the version from the Net2 distribution! 127However, if you are on BSD/386 or 386BSD-based systems, use the one 128that already exists on your system. You may need to #define OLD_NEWDB 1291 to do this.] 130 131[NOTE WELL: it is CRITICAL that you remove ndbm.o from libdb.a and 132ndbm.h from the appropriate include directories if you want to get 133ndbm support. These files OVERRIDE calls to ndbm routines -- in 134particular, if you leave ndbm.h in, you can find yourself using 135the new db package even if you don't define NEWDB.] 136 137If NEWDB and NDBM are defined (but not NIS), then sendmail will read 138NDBM format alias files, but the next time a newaliases is run the 139format will be converted to NEWDB; that format will be used forever 140more. This is intended as a transition feature. [Note however that 141the NEWDB library also catches and maps NDBM calls; you will have to 142back out this feature to get this to work. See ``Quirks'' section 143below for details.] 144 145If all three are defined, sendmail operates as described above, and also 146looks for the file /var/yp/Makefile. If it exists, newaliases will 147build BOTH the NEWDB and NDBM format alias files. However, it will 148only use the NEWDB file; the NDBM format file is used only by the 149NIS subsystem. 150 151If NDBM and NIS are defined (regardless of the definition of NEWDB 152or the existance of /var/yp/Makefile), sendmail adds the special 153tokens "YP_LAST_MODIFIED" and "YP_MASTER_NAME", both of which are 154required if the NDBM file is to be used as an NIS map. 155 156There is also preliminary support for NIS+ (-DNISPLUS), Hesiod 157(-DHESIOD), and NetInfo (-DNETINFO). These have not been well 158tested. 159 160All of -DNEWDB, -DNDBM, -DNIS, -DNISPLUS, -DHESIOD, and -DNETINFO are 161normally defined in the DBMDEF line in the Makefile. 162 163 164+---------------+ 165| COMPILE FLAGS | 166+---------------+ 167 168Whereever possible, I try to make sendmail pull in the correct 169compilation options needed to compile on various environments based on 170automatically defined symbols. Some machines don't seem to have useful 171symbols availble, requiring the following compilation flags in the 172Makefile: 173 174SOLARIS Define this if you are running Solaris 2.0 or higher. 175SOLARIS_2_3 Define this if you are running Solaris 2.3 or higher. 176SUNOS403 Define this if you are running SunOS 4.0.3. 177NeXT Define this if you are on a NeXT box. (This one may 178 be pre-defined for you.) There are other hacks you 179 have to make -- see below. 180_AIX3 Define this if you are IBM AIX 3.x. 181RISCOS Define this if you are running RISC/os from MIPS. 182IRIX Define this if you are running IRIX from SGI. 183_SCO_unix_ Define this if you are on SCO UNIX. 184_SCO_unix_4_2 Define this if you are on SCO Open Server 3.2v4. 185DGUX Define this if you are on DG/UX 5.4.3 or later 186DGUX_5_4_2 Define this if you are on DG/UX systems prior to 5.4.3. 187NonStop_UX_BXX Define this if you are on a Tandem NonStop-UX release 188 Bxx system. 189IRIX64 Define this if you are on an IRIX64 system. 190 191If you are a system that sendmail has already been ported to, you 192probably won't have to touch these. But if you are porting, you may 193have to tweak the following compilation flags in conf.h in order to 194get it to compile and link properly: 195 196SYSTEM5 Adjust for System V (not necessarily Release 4). 197SYS5SIGNALS Use System V signal semantics -- the signal handler 198 is automatically dropped when the signal is caught. 199 If this is not set, use POSIX/BSD semantics, where the 200 signal handler stays in force until an exec or an 201 explicit delete. Implied by SYSTEM5. 202SYS5SETPGRP Use System V setpgrp() semantics. Implied by SYSTEM5. 203HASFLOCK Set this if you prefer to use the flock(2) system call 204 rather than using fcntl-based locking. Fcntl locking 205 has some semantic gotchas, but many vendor systems 206 also interface it to lockd(8) to do NFS-style locking. 207 For this reason, this should not be set unless you 208 don't have an alternative. 209HASUNAME Set if you have the "uname" system call. Implied by 210 SYSTEM5. 211HASUNSETENV Define this if your system library has the "unsetenv" 212 subroutine. 213HASSETSID Define this if you have the setsid(2) system call. This 214 is implied if your system appears to be POSIX compliant. 215HASINITGROUPS Define this if you have the initgroups(3) routine. 216HASSETVBUF Define this if you have the setvbuf(3) library call. 217 If you don't, setlinebuf will be used instead. This 218 defaults on if your compiler defines __STDC__. 219HASSETREUID Define this if you have setreuid(2) ***AND*** root can 220 use setreuid to change to an arbitrary user. This second 221 condition is not satisfied on AIX 3.x. You may find that 222 your system has setresuid(2), (for example, on HP-UX) in 223 which case you will also have to #define setreuid(r, e) 224 to be the appropriate call. Some systems (such as Solaris) 225 have a compatibility routine that doesn't work properly, 226 but may have "saved user ids" properly implemented so you 227 can ``#define setreuid(r, e) seteuid(e)'' and have it work. 228 The important thing is that you have a call that will set 229 the effective uid independently of the real or saved uid 230 and be able to set the effective uid back again when done. 231 There's a test program in ../test/t_setreuid.c that will 232 try things on your system. Setting this improves the 233 security, since sendmail doesn't have to read .forward 234 and :include: files as root. There are certain attacks 235 that may be unpreventable without this call. 236HASLSTAT Define this if you have symbolic links (and thus the 237 lstat(2) system call). This improves security. Unlike 238 most other options, this one is on by default, so you 239 need to #undef it in conf.h if you don't have symbolic 240 links (these days everyone does). 241HASSETRLIMIT Define this to 1 if you have the setrlimit(2) syscall. 242 You can define it to 0 to force it off. It is assumed 243 if you are running a BSD-like system. 244HASULIMIT Define this if you have the ulimit(2) syscall (System V 245 style systems). HASSETRLIMIT overrides, as it is more 246 general. 247NEEDGETOPT Define this if you need a reimplementation of getopt(3). 248 On some systems, getopt does very odd things if called 249 to scan the arguments twice. This flag will ask sendmail 250 to compile in a local version of getopt that works 251 properly. 252NEEDSTRTOL Define this if your standard C library does not define 253 strtol(3). This will compile in a local version. 254NEEDVPRINTF Define this if your standard C library does not define 255 vprintf(3). Note that the resulting fake implementation 256 is not very elegant and may not even work on some 257 architectures. 258NEEDFSYNC Define this if your standard C library does not define 259 fsync(2). This will try to simulate the operation using 260 fcntl(2); if that is not available it does nothing, which 261 isn't great, but at least it compiles and runs. 262HASGETUSERSHELL Define this to 1 if you have getusershell(3) in your 263 standard C library. If this is not defined, or is defined 264 to be 0, sendmail will scan the /etc/shells file (no 265 NIS-style support, defaults to /bin/sh and /bin/csh if 266 that file does not exist) to get a list of unrestricted 267 user shells. This is used to determine whether users 268 are allowed to forward their mail to a program or a file. 269GIDSET_T The type of entries in a gidset passed as the second 270 argument to getgroups(2). Historically this has been an 271 int, so this is the default, but some systems (such as 272 IRIX) pass it as a gid_t, which is an unsigned short. 273 This will make a difference, so it is important to get 274 this right! However, it is only an issue if you have 275 group sets. 276SLEEP_T The type returned by the system sleep() function. 277 Defaults to "unsigned int". Don't worry about this 278 if you don't have compilation problems. 279ARBPTR_T The type of an arbitrary pointer -- defaults to "void *". 280 If you are an very old compiler you may need to define 281 this to be "char *". 282LA_TYPE The type of load average your kernel supports. These 283 can be one of: 284 LA_ZERO (1) -- it always returns the load average as 285 "zero" (and does so on all architectures). 286 LA_INT (2) to read /dev/kmem for the symbol avenrun and 287 interpret as a long integer. 288 LA_FLOAT (3) same, but interpret the result as a floating 289 point number. 290 LA_SHORT (6) to interpret as a short integer. 291 LA_SUBR (4) if you have the getloadavg(3) routine in your 292 system library. 293 LA_MACH (5) to use MACH-style load averages (calls 294 processor_set_info()), 295 LA_PROCSTR (7) to read /proc/loadavg and interpret it 296 as a string representing a floating-point 297 number (Linux-style). 298 LA_READKSYM (8) is an implementation suitable for some 299 versions of SVr4 that uses the MIOC_READKSYM ioctl 300 call to read /dev/kmem. 301 LA_DGUX (9) is a special implementation for DG/UX that uses 302 the dg_sys_info system call. 303 LA_HPUX (10) is an HP-UX specific version that uses the 304 pstat_getdynamic system call. 305 LA_INT, LA_SHORT, LA_FLOAT, and LA_READKSYM have several 306 other parameters that they try to divine: the name of your 307 kernel, the name of the variable in the kernel to examine, 308 the number of bits of precision in a fixed point load average, 309 and so forth. 310 In desperation, use LA_ZERO. The actual code is in 311 conf.c -- it can be tweaked if you are brave. 312FSHIFT For LA_INT, LA_SHORT, and LA_READKSYM, this is the number 313 of bits of load average after the binary point -- i.e., 314 the number of bits to shift right in order to scale the 315 integer to get the true integer load average. Defaults to 8. 316_PATH_UNIX The path to your kernel. Needed only for LA_INT, LA_SHORT, 317 and LA_FLOAT. Defaults to "/unix" on System V, "/vmunix" 318 everywhere else. 319LA_AVENRUN For LA_INT, LA_SHORT, and LA_FLOAT, the name of the kernel 320 variable that holds the load average. Defaults to "avenrun" 321 on System V, "_avenrun" everywhere else. 322SFS_TYPE Encodes how your kernel can locate the amount of free 323 space on a disk partition. This can be set to SFS_NONE 324 (0) if you have no way of getting this information, 325 SFS_USTAT (1) if you have the ustat(2) system call, 326 SFS_4ARGS (2) if you have a four-argument statfs(2) 327 system call (and the include file is <sys/statfs.h>), 328 SFS_VFS (3), SFS_MOUNT (4), SFS_STATFS (5) if you have 329 the two-argument statfs(2) system call with includes in 330 <sys/vfs.h>, <sys/mount.h>, or <sys/statfs.h> respectively, 331 or SFS_STATVFS (6) if you have the two-argument statvfs(2) 332 call. The default if nothing is defined is SFS_NONE. 333SFS_BAVAIL with SFS_4ARGS hou can also set SFS_BAVAIL to the field name 334 in the statfs structure that holds the useful information; 335 this defaults to f_bavail. 336SPT_TYPE Encodes how your system can display what a process is doing 337 on a ps(1) command (SPT stands for Set Process Title). Can 338 be set to: 339 SPT_NONE (0) -- Don't try to set the process title at all. 340 SPT_REUSEARGV (1) -- Pad out your argv with the information; 341 this is the default if none specified. 342 SPT_BUILTIN (2) -- The system library has setproctitle. 343 SPT_PSTAT (3) -- Use the PSTAT_SETCMD option to pstat(2) 344 to set the process title; this is used by HP-UX. 345 SPT_PSSTRINGS (4) -- Use the magic PS_STRINGS pointer (4.4BSD). 346SPT_PADCHAR Character used to pad the process title; if undefined, 347 the space character (0x20) is used. This is ignored if 348 SPT_TYPE != SPT_REUSEARGV 349ERRLIST_PREDEFINED 350 If set, assumes that some header file defines sys_errlist. 351 This may be needed if you get type conflicts on this 352 variable -- otherwise don't worry about it. 353WAITUNION The wait(2) routine takes a "union wait" argument instead 354 of an integer argument. This is for compatibility with 355 old versions of BSD. 356SCANF You can set this to extend the F command to accept a 357 scanf string -- this gives you a primitive parser for 358 class definitions -- BUT it can make you vulnerable to 359 core dumps if the target file is poorly formed. 360SYSLOG_BUFSIZE You can define this to be the size of the buffer that 361 syslog accepts. If it is not defined, it assumes a 362 1024-byte buffer. If the buffer is very small (under 363 256 bytes) the log message format changes -- each 364 e-mail message will log many more messages, since it 365 will log each piece of information as a separate line 366 in syslog. 367BROKEN_RES_SEARCH 368 On Ultrix (and maybe other systems?) if you use the 369 res_search routine with an unknown host name, it returns 370 -1 but sets h_errno to 0 instead of HOST_NOT_FOUND. If 371 you set this, sendmail considers 0 to be the same as 372 HOST_NOT_FOUND. 373NAMELISTMASK If defined, values returned by nlist(3) are masked 374 against this value before use -- a common value is 375 0x7fffffff to strip off the top bit. 376BSD4_4_SOCKADDR If defined, socket addresses have an sa_len field that 377 defines the length of this address. 378 379 380 381+-----------------------+ 382| COMPILE-TIME FEATURES | 383+-----------------------+ 384 385There are a bunch of features that you can decide to compile in, such 386as selecting various database packages and special protocol support. 387Several are assumed based on other compilation flags -- if you want to 388"un-assume" something, you probably need to edit conf.h. Compilation 389flags that add support for special features include: 390 391NDBM Include support for "new" DBM library for aliases and maps. 392 Normally defined in the Makefile. 393NEWDB Include support for Berkeley "db" package (hash & btree) 394 for aliases and maps. Normally defined in the Makefile. 395OLD_NEWDB If non-zero, the version of NEWDB you have is the old 396 one that does not include the "fd" call. This call was 397 added in version 1.5 of the Berkeley DB code. If you 398 use -DOLD_NEWDB=0 it forces you to use the new interface. 399NIS Define this to get NIS (YP) support for aliases and maps. 400 Normally defined in the Makefile. 401NISPLUS Define this to get NIS+ support for aliases and maps. 402 Normally defined in the Makefile. 403HESIOD Define this to get Hesiod support for aliases and maps. 404 Normally defined in the Makefile. 405NETINFO Define this to get NeXT NetInfo support for aliases and maps. 406 Normally defined in the Makefile. 407USERDB Include support for the User Information Database. Implied 408 by NEWDB in conf.h. 409IDENTPROTO Define this as 1 to get IDENT (RFC 1413) protocol support. 410 This is assumed unless you are running on Ultrix or 411 HP-UX, both of which have a problem in the UDP 412 implementation. You can define it to be 0 to explicitly 413 turn off IDENT protocol support. If defined off, the code 414 is actually still compiled in, but it defaults off; you 415 can turn it on by setting the IDENT timeout to 30s in the 416 configuration file. 417LOG Set this to get syslog(3) support. Defined by default 418 in conf.h. You want this if at all possible. 419NETINET Set this to get TCP/IP support. Defined by default 420 in conf.h. You probably want this. 421NETISO Define this to get ISO networking support. 422SMTP Define this to get the SMTP code. Implied by NETINET 423 or NETISO. 424NAMED_BIND If non-zero, include DNS (name daemon) support, including 425 MX support. The specs you must use this if you run 426 SMTP. Defined by default in conf.h. 427QUEUE Define this to get queueing code. Implied by NETINET 428 or NETISO; required by SMTP. This gives you other good 429 stuff -- it should be on. 430DAEMON Define this to get general network support. Implied by 431 NETINET or NETISO. Defined by default in conf.h. You 432 almost certainly want it on. 433MATCHGECOS Permit fuzzy matching of user names against the full 434 name (GECOS) field in the /etc/passwd file. This should 435 probably be on, since you can disable it from the config 436 file if you want to. Defined by default in conf.h. 437MIME8TO7 If non-zero, include 8 to 7 bit MIME conversions. This 438 also controls advertisement of 8BITMIME in the ESMTP 439 startup dialogue. 440MIME7TO8 If non-zero, include 7 to 8 bit MIME conversions. Not yet 441 implemented. 442 443 444+---------------------+ 445| DNS/RESOLVER ISSUES | 446+---------------------+ 447 448Many systems have old versions of the resolver library. At a minimum, 449you should be running BIND 4.8.3; older versions may compile, but they 450have known bugs that should give you pause. 451 452Common problems in old versions include "undefined" errors for 453dn_skipname. 454 455Some people have had a problem with BIND 4.9; it uses some routines 456that it expects to be externally defined such as strerror(). It may 457help to link with "-l44bsd" to solve this problem. 458 459!PLEASE! be sure to link with the same version of the resolver as 460the header files you used -- some people have used the 4.9 headers 461and linked with BIND 4.8 or vice versa, and it doesn't work. 462Unfortunately, it doesn't fail in an obvious way -- things just 463subtly don't work. 464 465 466+-------------------------------------+ 467| OPERATING SYSTEM AND COMPILE QUIRKS | 468+-------------------------------------+ 469 470GCC 2.5.x problems *** IMPORTANT *** 471 Date: Mon, 29 Nov 93 19:08:44 PST 472 From: wilson@cygnus.com (Jim Wilson) 473 Message-Id: <9311300308.AA04608@cygnus.com> 474 To: kenner@vlsi1.ultra.nyu.edu 475 Subject: [cattelan@thebarn.com: gcc 2.5.4-2.5.5 -O bug] 476 Cc: cattelan@thebarn.com, rms@gnu.ai.mit.edu, sendmail@cs.berkeley.edu 477 478 This fixes a problem that occurs when gcc 2.5.5 is used to compile 479 sendmail 8.6.4 with optimization on a sparc. 480 481 Mon Nov 29 19:00:14 1993 Jim Wilson (wilson@sphagnum.cygnus.com) 482 483 * reload.c (find_reloads_toplev): Replace obsolete reference to 484 BYTE_LOADS_*_EXTEND with LOAD_EXTEND_OP. 485 486 *** clean-ss-931128/reload.c Sun Nov 14 16:20:01 1993 487 --- ss-931128/reload.c Mon Nov 29 18:52:55 1993 488 *************** find_reloads_toplev (x, opnum, type, ind 489 *** 3888,3894 **** 490 force a reload in that case. So we should not do anything here. */ 491 492 else if (regno >= FIRST_PSEUDO_REGISTER 493 ! #if defined(BYTE_LOADS_ZERO_EXTEND) || defined(BYTE_LOADS_SIGN_EXTEND) 494 && (GET_MODE_SIZE (GET_MODE (x)) 495 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 496 #endif 497 --- 3888,3894 ---- 498 force a reload in that case. So we should not do anything here. */ 499 500 else if (regno >= FIRST_PSEUDO_REGISTER 501 ! #ifdef LOAD_EXTEND_OP 502 && (GET_MODE_SIZE (GET_MODE (x)) 503 <= GET_MODE_SIZE (GET_MODE (SUBREG_REG (x)))) 504 #endif 505 506 507SunOS 4.x (Solaris 1.x) 508 You may have to use -lresolv on SunOS. However, beware that 509 this links in a new version of gethostbyname that does not 510 understand NIS, so you must have all of your hosts in DNS. 511 512 Some people have reported problems with the SunOS version of 513 -lresolv and/or in.named, and suggest that you get a newer 514 version. The symptoms are delays when you connect to the 515 SMTP server on a SunOS machine or having your domain added to 516 addresses inappropriately. There is a version of BIND 517 version 4.9 on gatekeeper.DEC.COM in pub/BSD/bind/4.9. 518 519 There is substantial disagreement about whether you can make 520 this work with resolv+, which allows you to specify a search-path 521 of services. Some people report that it works fine, others 522 claim it doesn't work at all (including causing sendmail to 523 drop core when it tries to do multiple resolv+ lookups for a 524 single job). I haven't tried resolv+, as we use DNS exclusively. 525 526 Should you want to try resolv+, it is on ftp.uu.net in 527 /networking/ip/dns. 528 529 Apparently getservbyname() can fail under moderate to high 530 load under some circumstances. This will exhibit itself as 531 the message ``554 makeconnection: service "smtp" unknown''. 532 The problem has been traced to one or more blank lines in 533 /etc/services on the NIS server machine. Delete these 534 and it should work. This info is thanks to Brian Bartholomew 535 <bb@math.ufl.edu> of I-Kinetics, Inc. 536 537Solaris 2.x (SunOS 5.x) 538 To compile for Solaris, be sure you use -DSOLARIS. 539 540 To the best of my knowledge, Solaris does not have the 541 gethostbyname problem described above. However, it does 542 have another one: 543 544 From a correspondent: 545 546 For solaris 2.2, I have 547 548 hosts: files dns 549 550 in /etc/nsswitch.conf and /etc/hosts has to have the fully 551 qualified host name. I think "files" has to be before "dns" 552 in /etc/nsswitch.conf during bootup. 553 554 From another correspondent: 555 556 When running sendmail under Solaris, the gethostbyname() 557 hack in conf.c which should perform proper canonicalization 558 of host names could fail. Result: the host name is not 559 canonicalized despite the hack, and you'll have to define $j 560 and $m in sendmail.cf somewhere. 561 562 The reason could be that /etc/nsswitch.conf is improperly 563 configured (at least from sendmail's point of view). For 564 example, the line 565 566 hosts: files nisplus dns 567 568 will make gethostbyname() look in /etc/hosts first, then ask 569 nisplus, then dns. However, if /etc/hosts does not contain 570 the full canonicalized hostname, then no amount of 571 gethostbyname()s will work. 572 573 Solution (or rather, a workaround): Ask nisplus first, then 574 dns, then local files: 575 576 hosts: nisplus dns [NOTFOUND=return] files 577 578 The Solaris "syslog" function is apparently limited to something 579 about 90 characters because of a kernel limitation. If you have 580 source code, you can probably up this number. You can get patches 581 that fix this problem: the patch ids are: 582 583 Solaris 2.1 100834 584 Solaris 2.2 100999 585 Solaris 2.3 101318 586 587 Be sure you have the appropriate patch installed or you won't 588 see system logging. 589 590Ultrix 591 By default, the IDENT protocol is turned off on Ultrix. If you 592 are running Ultrix 4.4 or later, or if you have gotten a patch 593 the TCP problem for an earlier version of Ultrix, you can turn 594 IDENT on in the configuration file by setting the "ident" timeout 595 to 30 seconds. 596 597OSF/1 598 If you are compiling on OSF/1 (DEC Alpha), you must use 599 -L/usr/shlib (otherwise it core dumps on startup). You may also 600 need -mld to get the nlist() function, although some versions 601 apparently don't need this. 602 603 Also, the enclosed makefile removed /usr/sbin/smtpd; if you need 604 it, just create the link to the sendmail binary. 605 606IRIX 607 The header files on SGI IRIX are completely prototyped, and as 608 a result you can sometimes get some warning messages during 609 compilation. These can be ignored. There are two errors in 610 deliver only if you are using gcc, both of the form ``warning: 611 passing arg N of `execve' from incompatible pointer type''. 612 Also, if you compile with -DNIS, you will get a complaint 613 about a declaration of struct dom_binding in a prototype 614 when compiling map.c; this is not important because the 615 function being prototyped is not used in that file. 616 617 In order to compile sendmail you will have had to install 618 the developers' option in order to get the necessary include 619 files. 620 621NeXT 622 If you are compiling on NeXT, you will have to create an empty 623 file "unistd.h" and create a file "dirent.h" containing: 624 625 #include <sys/dir.h> 626 #define dirent direct 627 628 (The Makefile.NeXT should try to do both of these for you.) 629 630 Apparently, there is a bug in getservbyname on Nextstep 3.0 631 that causes it to fail under some circumstances with the 632 message "SYSERR: service "smtp" unknown" logged. You should 633 be able to work around this by including the line: 634 635 OOPort=25 636 637 in your .cf file. 638 639 You may have to use -DNeXT. 640 641BSDI (BSD/386) 1.0, NetBSD 0.9, FreeBSD 1.0 642 The "m4" from BSDI won't handle the config files properly. 643 I haven't had a chance to test this myself. 644 645 The M4 shipped in FreeBSD and NetBSD 0.9 don't handle the config 646 files properly. One must use either GNU m4 1.1 or the PD-M4 647 recently posted in comp.os.386bsd.bugs (and maybe others). 648 NetBSD-current includes the PD-M4 (as stated in the NetBSD file 649 CHANGES). 650 651 FreeBSD 1.0 RELEASE has uname(2) now. Use -DUSEUNAME in order to 652 use it (look into Makefile.FreeBSD). NetBSD-current may have 653 it too but it has not been verified. 654 655 You cannot port the latest version of the Berkeley db library 656 and use it with sendmail without recompiling the world. This 657 is because C library routines use the older version which have 658 incompatible header files -- the result is that it can't read 659 other system files, such as /etc/passwd, unless you use the 660 new db format throughout your system. You should normally just 661 use the version of db supplied in your release. You may need 662 to use -DOLD_NEWDB=1 to make this work -- this turns off some 663 new interface calls (for file locking) that are not in older 664 versions of db. You'll get compile errors if you need this 665 flag and don't have it set. 666 6674.3BSD 668 If you are running a "virgin" version of 4.3BSD, you'll have 669 a very old resolver and be missing some header files. The 670 header files are simple -- create empty versions and everything 671 will work fine. For the resolver you should really port a new 672 version (4.8.3 or later) of the resolver; 4.9 is available on 673 gatekeeper.DEC.COM in pub/BSD/bind/4.9. If you are really 674 determined to continue to use your old, buggy version (or as 675 a shortcut to get sendmail working -- I'm sure you have the 676 best intentions to port a modern version of BIND), you can 677 copy ../contrib/oldbind.compat.c into src and add 678 oldbind.compat.o to OBJADD in the Makefile. 679 680A/UX 681 Date: Tue, 12 Oct 1993 18:28:28 -0400 (EDT) 682 From: "Eric C. Hagberg" <hagberg@med.cornell.edu> 683 Subject: Fix for A/UX ndbm 684 685 I guess this isn't really a sendmail bug, however, it is something 686 that A/UX users should be aware of when compiling sendmail 8.6. 687 688 Apparently, the calls that sendmail is using to the ndbm routines 689 in A/UX 3.0.x contain calls to "broken" routines, in that the 690 aliases database will break when it gets "just a little big" 691 (sorry I don't have exact numbers here, but it broke somewhere 692 around 20-25 aliases for me.), making all aliases non-functional 693 after exceeding this point. 694 695 What I did was to get the gnu-dbm-1.6 package, compile it, and 696 then re-compile sendmail with "-lgdbm", "-DNDBM", and using the 697 ndbm.h header file that comes with the gnu-package. This makes 698 things behave properly. 699 700 I suppose porting the New Berkeley db package is another route, 701 however, I made a quick attempt at it, and found it difficult 702 (not easy at least); the gnu-dbm package "configured" and 703 compiled easily. 704 705SCO Unix 706 From: Thomas Essebier <tom@stallion.oz.au> 707 Organisation: Stallion Technologies Pty Ltd. 708 709 It will probably help those who are trying to configure sendmail 8.6.9 710 to know that if they are on SCO, they had better set 711 OI-dnsrch 712 or they will core dump as soon as they try to use the resolver. 713 ie. although SCO has _res.dnsrch defined, and is kinda BIND 4.8.3, it 714 does not inititialise it, nor does it understand 'search' in 715 /etc/named.boot. 716 - sigh - 717 718DG/UX 719 Doug Anderson <dlander@afterlife.ncsc.mil> has successfully run 720 V8 on the DG/UX 5.4.2 and 5.4R3.x platforms under heavy usage. 721 Originally, the DG /bin/mail program wasn't compatible with 722 the V8 sendmail, since the DG /bin/mail requires the environment 723 variable "_FORCE_MAIL_LOCAL_=yes" be set. Version 8.7 now includes 724 this in the environment before invoking the local mailer. Some 725 have used procmail to avoid this problem in the past. It works 726 but some have experienced file locking problems with their DG/UX 727 ports of procmail. 728 729Apollo DomainOS 730 If you are compiling on Apollo, you will have to create an empty 731 file "unistd.h" and create a file "dirent.h" containing: 732 733 #include <sys/dir.h> 734 #define dirent direct 735 736 (The Makefile.DomainOS will attempt to do both of these for you.) 737 738HP-UX 8.00 739 Date: Mon, 24 Jan 1994 13:25:45 +0200 740 From: Kimmo Suominen <Kimmo.Suominen@lut.fi> 741 Subject: 8.6.5 w/ HP-UX 8.00 on s300 742 743 Just compiled and fought with sendmail 8.6.5 on a HP9000/360 (ie. a 744 series 300 machine) running HP-UX 8.00. 745 746 I was getting segmentation fault when delivering to a local user. 747 With debugging I saw it was faulting when doing _free@libc... *sigh* 748 It seems the new implementation of malloc on s300 is buggy as of 8.0, 749 so I tried out the one in -lmalloc (malloc(3X)). With that it seems 750 to work just dandy. 751 752 When linking, you will get the following error: 753 754 ld: multiply defined symbol _freespace in file /usr/lib/libmalloc.a 755 756 but you can just ignore it. You might want to add this info to the 757 README file for the future... 758 759Linux 760 Something broke between versions 0.99.13 and 0.99.14 of Linux: 761 the flock() system call gives errors. If you are running .14, 762 you must not use flock. You can do this with -DHASFLOCK=0. 763 764 Around the inclusion of bind-4.9.3 & linux libc-4.6.20, the 765 initialization of the _res structure changed. If /etc/hosts.conf 766 was configured as "hosts, bind" the resolver code could return 767 "Name server failure" errors. This is supposedly fixed in 768 later versions of libc (>= 4.6.29?), and later versions of 769 sendmail (> 8.6.10) try to work around the problem. 770 771 Some older versions (< 4.6.20?) of the libc/include files conflict 772 with sendmail's version of cdefs.h. Deleting sendmail's version 773 on those systems should be non-harmful, and new versions don't care. 774 775AIX 776 This version of sendmail does not support MB, MG, and MR resource 777 records, which are supported by AIX sendmail. 778 779RISC/os 780 RISC/os from MIPS is a merged AT&T/Berkeley system. When you 781 compile on that platform you will get duplicate definitions 782 on many files. You can ignore these. 783 784System V Release 4 Based Systems 785 There is a single Makefile that is intended for all SVR4-based 786 systems (called Makefile.SVR4). It defines __svr4__, which is 787 predefined by some compilers. If your compiler already defines 788 this compile variable, you can delete the definition from the 789 Makefile. 790 791 It's been tested on Dell Issue 2.2. 792 793DELL SVR4 794 Date: Mon, 06 Dec 1993 10:42:29 EST 795 From: "Kimmo Suominen" <kim@grendel.lut.fi> 796 Message-ID: <2d0352f9.lento29@lento29.UUCP> 797 To: eric@cs.berkeley.edu 798 Cc: sendmail@cs.berkeley.edu 799 Subject: Notes for DELL SVR4 800 801 Eric, 802 803 Here are some notes for compiling Sendmail 8.6.4 on DELL SVR4. I ran 804 across these things when helping out some people who contacted me by 805 e-mail. 806 807 1) Use gcc 2.4.5 (or later?). Dell distributes gcc 2.1 with their 808 Issue 2.2 Unix. It is too old, and gives you problems with 809 clock.c, because sigset_t won't get defined in <sys/signal.h>. 810 This is due to a problematic protection rule in there, and is 811 fixed with gcc 2.4.5. 812 813 2) If you don't use the new Berkeley DB (-DNEWDB), then you need 814 to add "-lc -lucb" to the libraries to link with. This is because 815 the -ldbm distributed by Dell needs the bcopy, bcmp and bzero 816 functions. It is important that you specify both libraries in 817 the given order to be sure you only get the BSTRING functions 818 from the UCB library (and not the signal routines etc.). 819 820 3) Don't leave out "-lelf" even if compiling with "-lc -lucb". 821 The UCB library also has another copy of the nlist routines, 822 but we do want the ones from "-lelf". 823 824 If anyone needs a compiled gcc 2.4.5 and/or a ported DB library, they 825 can use anonymous ftp to fetch them from lut.fi in the /kim directory. 826 They are copies of what I use on grendel.lut.fi, and offering them 827 does not imply that I would also support them. I have sent the DB 828 port for SVR4 back to Keith Bostic for inclusion in the official 829 distribution, but I haven't heard anything from him as of today. 830 831 - gcc-2.4.5-svr4.tar.gz (gcc 2.4.5 and the corresponding libg++) 832 - db-1.72.tar.gz (with source, objects and a installed copy) 833 834 Cheers 835 + Kim 836 -- 837 * Kimmo.Suominen@lut.fi * SysVr4 enthusiast at GRENDEL.LUT.FI * 838 * KIM@FINFILES.BITNET * Postmaster and Hostmaster at LUT.FI * 839 * + 358 200 865 718 * Unix area moderator at NIC.FUNET.FI * 840 841ConvexOS 10.1 and below 842 In order to use the name server, you must create the file 843 /etc/use_nameserver. If this file does not exist, the call 844 to res_init() will fail and you will have absolutely no 845 access to DNS, including MX records. 846 847Amdahl UTS 2.1.5 848 In order to get UTS to work, you will have to port BIND 4.9. 849 The vendor's BIND is reported to be ``totally inadequate.'' 850 See sendmail/contrib/AmdahlUTS.patch for the patches necessary 851 to get BIND 4.9 compiled for UTS. 852 853UnixWare 2.0 854 According to Alexander Kolbasov <sasha@unitech.gamma.ru>, 855 the m4 on UnixWare 2.0 (still in Beta) will core dump on the 856 config files. GNU m4 and the m4 from UnixWare 1.x both work. 857 858Non-DNS based sites 859 This version of sendmail always tries to connect to the Domain 860 Name System (DNS) to resolve names, regardless of the setting 861 of the `I' option. On most systems that are not running DNS, 862 this will fail quickly and sendmail will continue, but on some 863 systems it has a long timeout. If you have this problem, you 864 will have to recompile without NAMED_BIND. Some people have 865 claimed that they have successfully used "OI+USEVC" to force 866 sendmail to use a virtual circuit -- this will always time out 867 quickly, but also tells sendmail that a failed connection 868 should requeue the message (probably not what you intended). 869 A future release of sendmail will correct this problem. 870 871Both NEWDB and NDBM 872 If you use both -DNDBM and -DNEWDB, you must delete the module 873 ndbm.o from libdb.a and delete the file "ndbm.h" from the files 874 that get installed (that is, use the OLD ndbm.h, not the new 875 ndbm.h). This compatibility module maps ndbm calls into DB 876 calls, and breaks things rather badly. 877 878GNU getopt 879 I'm told that GNU getopt has a problem in that it gets confused 880 by the double call. Use the version in conf.c instead. 881 882BIND 4.9.2 and Ultrix 883 If you are running on Ultrix, be sure you read conf/Info.Ultrix 884 in the BIND distribution very carefully -- there is information 885 in there that you need to know in order to avoid errors of the 886 form: 887 888 /lib/libc.a(gethostent.o): sethostent: multiply defined 889 /lib/libc.a(gethostent.o): endhostent: multiply defined 890 /lib/libc.a(gethostent.o): gethostbyname: multiply defined 891 /lib/libc.a(gethostent.o): gethostbyaddr: multiply defined 892 893 during the link stage. 894 895strtoul 896 Some compilers (notably gcc) claim to be ANSI C but do not 897 include the ANSI-required routine "strtoul". If your compiler 898 has this problem, you will get an error in srvrsmtp.c on the 899 code: 900 901 # ifdef defined(__STDC__) && !defined(BROKEN_ANSI_LIBRARY) 902 e->e_msgsize = strtoul(vp, (char **) NULL, 10); 903 # else 904 e->e_msgsize = strtol(vp, (char **) NULL, 10); 905 # endif 906 907 You can use -DBROKEN_ANSI_LIBRARY to get around this problem. 908 909 910+--------------+ 911| MANUAL PAGES | 912+--------------+ 913 914The manual pages have been written against the -mandoc macros 915instead of the -man macros. The latest version of groff has them 916included. You can also get a copy from FTP.UU.NET in directory 917/systems/unix/bsd-sources/share/tmac. 918 919 920+-----------------+ 921| DEBUGGING HOOKS | 922+-----------------+ 923 924As of 8.6.5, sendmail daemons will catch a SIGUSR1 signal and log 925some debugging output (logged at LOG_DEBUG severity). The 926information dumped is: 927 928 * The value of the $j macro. 929 * A warning if $j is not in the set $=w. 930 * A list of the open file descriptors. 931 * The contents of the connection cache. 932 * If ruleset 89 is defined, it is evaluated and the results printed. 933 934This allows you to get information regarding the runtime state of the 935daemon on the fly. This should not be done too frequently, since 936the process of rewriting may lose memory which will not be recovered. 937Also, ruleset 89 may call non-reentrant routines, so there is a small 938non-zero probability that this will cause other problems. It is 939really only for debugging serious problems. 940 941A typical formulation of ruleset 89 would be: 942 943 R$* $@ $>0 some test address 944 945 946+-----------------------------+ 947| DESCRIPTION OF SOURCE FILES | 948+-----------------------------+ 949 950The following list describes the files in this directory: 951 952Makefile The makefile used here; this version only works with 953 the new Berkeley make. 954Makefile.dist A trimmed down version of the makefile that works with 955 the old make. 956READ_ME This file. 957TRACEFLAGS My own personal list of the trace flags -- not guaranteed 958 to be particularly up to date. 959alias.c Does name aliasing in all forms. 960arpadate.c A subroutine which creates ARPANET standard dates. 961clock.c Routines to implement real-time oriented functions 962 in sendmail -- e.g., timeouts. 963collect.c The routine that actually reads the mail into a temp 964 file. It also does a certain amount of parsing of 965 the header, etc. 966conf.c The configuration file. This contains information 967 that is presumed to be quite static and non- 968 controversial, or code compiled in for efficiency 969 reasons. Most of the configuration is in sendmail.cf. 970conf.h Configuration that must be known everywhere. 971convtime.c A routine to sanely process times. 972daemon.c Routines to implement daemon mode. This version is 973 specifically for Berkeley 4.1 IPC. 974deliver.c Routines to deliver mail. 975domain.c Routines that interface with DNS (the Domain Name 976 System). 977err.c Routines to print error messages. 978envelope.c Routines to manipulate the envelope structure. 979headers.c Routines to process message headers. 980macro.c The macro expander. This is used internally to 981 insert information from the configuration file. 982main.c The main routine to sendmail. This file also 983 contains some miscellaneous routines. 984map.c Support for database maps. 985mci.c Routines that handle mail connection information caching. 986parseaddr.c The routines which do address parsing. 987queue.c Routines to implement message queueing. 988readcf.c The routine that reads the configuration file and 989 translates it to internal form. 990recipient.c Routines that manipulate the recipient list. 991savemail.c Routines which save the letter on processing errors. 992sendmail.h Main header file for sendmail. 993srvrsmtp.c Routines to implement server SMTP. 994stab.c Routines to manage the symbol table. 995stats.c Routines to collect and post the statistics. 996sysexits.c List of error messages associated with error codes 997 in sysexits.h. 998trace.c The trace package. These routines allow setting and 999 testing of trace flags with a high granularity. 1000udb.c The user database interface module. 1001usersmtp.c Routines to implement user SMTP. 1002util.c Some general purpose routines used by sendmail. 1003version.c The version number and information about this 1004 version of sendmail. Theoretically, this gets 1005 modified on every change. 1006 1007Eric Allman 1008 1009(Version 8.86, last update 05/18/95 08:30:29) 1010