1# Copyright (c) 1998-2003 Sendmail, Inc. and its suppliers. 2# All rights reserved. 3# Copyright (c) 1983, 1995-1997 Eric P. Allman. All rights reserved. 4# Copyright (c) 1988 5# The Regents of the University of California. All rights reserved. 6# 7# By using this file, you agree to the terms and conditions set 8# forth in the LICENSE file which can be found at the top level of 9# the sendmail distribution. 10# 11# 12# $Id: README,v 8.355.2.14 2003/03/27 21:39:29 ca Exp $ 13# 14 15This directory contains the source files for sendmail(TM). 16 17 ******************************************************************* 18 !! Read sendmail/SECURITY for important installation information !! 19 ******************************************************************* 20 21 ********************************************************** 22 ** Read below for more details on building sendmail. ** 23 ********************************************************** 24 25************************************************************************** 26** IMPORTANT: Read the appropriate paragraphs in the section on ** 27** ``Operating System and Compile Quirks''. ** 28************************************************************************** 29 30For detailed instructions, please read the document ../doc/op/op.me: 31 32 cd ../doc/op ; make op.ps op.txt 33 34Sendmail is a trademark of Sendmail, Inc. 35 36 37+-------------------+ 38| BUILDING SENDMAIL | 39+-------------------+ 40 41By far, the easiest way to compile sendmail is to use the "Build" 42script: 43 44 sh Build 45 46This uses the "uname" command to figure out what architecture you are 47on and creates a proper Makefile accordingly. It also creates a 48subdirectory per object format, so that multiarchitecture support is 49easy. In general this should be all you need. IRIX 6.x users should 50read the note below in the OPERATING SYSTEM AND COMPILE QUIRKS section. 51 52If you need to look at other include or library directories, use the 53-I or -L flags on the command line, e.g., 54 55 sh Build -I/usr/sww/include -L/usr/sww/lib 56 57It's also possible to create local site configuration in the file 58site.config.m4 (or another file settable with the -f flag). This 59file contains M4 definitions for various compilation values; the 60most useful are: 61 62confMAPDEF -D flags to specify database types to be included 63 (see below) 64confENVDEF -D flags to specify other environment information 65confINCDIRS -I flags for finding include files during compilation 66confLIBDIRS -L flags for finding libraries during linking 67confLIBS -l flags for selecting libraries during linking 68confLDOPTS other ld(1) linker options 69 70Others can be found by examining Makefile.m4. Please read 71../devtools/README for more information about the site.config.m4 72file. 73 74You can recompile from scratch using the -c flag with the Build 75command. This removes the existing compilation directory for the 76current platform and builds a new one. The -c flag must also 77be used if any site.*.m4 file in devtools/Site/ is changed. 78 79Porting to a new Unix-based system should be a matter of creating 80an appropriate configuration file in the devtools/OS/ directory. 81 82 83+----------------------+ 84| DATABASE DEFINITIONS | 85+----------------------+ 86 87There are several database formats that can be used for the alias files 88and for general maps. When used for alias files they interact in an 89attempt to be backward compatible. 90 91The options are: 92 93NEWDB The new Berkeley DB package. Some systems (e.g., BSD/OS and 94 Digital UNIX 4.0) have some version of this package 95 pre-installed. If your system does not have Berkeley DB 96 pre-installed, or the version installed is not version 2.0 97 or greater (e.g., is Berkeley DB 1.85 or 1.86), get the 98 current version from http://www.sleepycat.com/. DO NOT 99 use a version from any of the University of California, 100 Berkeley "Net" or other distributions. If you are still 101 running BSD/386 1.x, you will need to upgrade the included 102 Berkeley DB library to a current version. NEWDB is included 103 automatically if the Build script can find a library named 104 libdb.a or libdb.so. 105 See also OPERATING SYSTEM AND COMPILE QUIRKS about Berkeley 106 DB versions, e.g., DB 4.1.x. 107NDBM The older NDBM implementation -- the very old V7 DBM 108 implementation is no longer supported. 109NIS Network Information Services. To use this you must have 110 NIS support on your system. 111NISPLUS NIS+ (the revised NIS released with Solaris 2). You must 112 have NIS+ support on your system to use this flag. 113HESIOD Support for Hesiod (from the DEC/Athena distribution). You 114 must already have Hesiod support on your system for this to 115 work. You may be able to get this to work with the MIT/Athena 116 version of Hesiod, but that's likely to be a lot of work. 117 BIND 8.X also includes Hesiod support. 118LDAPMAP Lightweight Directory Access Protocol support. You will 119 have to install the UMich or OpenLDAP 120 (http://www.openldap.org/) ldap and lber libraries to use 121 this flag. 122MAP_REGEX Regular Expression support. You will need to use an 123 operating system which comes with the POSIX regex() 124 routines or install a regexp library such as libregex from 125 the Free Software Foundation. 126DNSMAP DNS map support. Requires NAMED_BIND. 127PH_MAP PH map support. You will need the libphclient library from 128 the nph package (http://www-dev.cso.uiuc.edu/ph/nph/). 129MAP_NSD nsd map support (IRIX 6.5 and later). 130 131>>> NOTE WELL for NEWDB support: If you want to get ndbm support, for 132>>> Berkeley DB versions under 2.0, it is CRITICAL that you remove 133>>> ndbm.o from libdb.a before you install it and DO NOT install ndbm.h; 134>>> for Berkeley DB versions 2.0 through 2.3.14, remove dbm.o from libdb.a 135>>> before you install it. If you don't delete these, there is absolutely 136>>> no point to including -DNDBM, since it will just get you another 137>>> (inferior) API to the same format database. These files OVERRIDE 138>>> calls to ndbm routines -- in particular, if you leave ndbm.h in, 139>>> you can find yourself using the new db package even if you don't 140>>> define NEWDB. Berkeley DB versions later than 2.3.14 do not need 141>>> to be modified. Please also consult the README in the top level 142>>> directory of the sendmail distribution for other important information. 143>>> 144>>> Further note: DO NOT remove your existing /usr/include/ndbm.h -- 145>>> you need that one. But do not install an updated ndbm.h in 146>>> /usr/include, /usr/local/include, or anywhere else. 147 148If NEWDB and NDBM are defined (but not NIS), then sendmail will read 149NDBM format alias files, but the next time a newaliases is run the 150format will be converted to NEWDB; that format will be used forever 151more. This is intended as a transition feature. 152 153If NEWDB, NDBM, and NIS are all defined and the name of the file includes 154the string "/yp/", sendmail will rebuild BOTH the NEWDB and NDBM format 155alias files. However, it will only read the NEWDB file; the NDBM format 156file is used only by the NIS subsystem. This is needed because the NIS 157maps on an NIS server are built directly from the NDBM files. 158 159If NDBM and NIS are defined (regardless of the definition of NEWDB), 160and the filename includes the string "/yp/", sendmail adds the special 161tokens "YP_LAST_MODIFIED" and "YP_MASTER_NAME", both of which are 162required if the NDBM file is to be used as an NIS map. 163 164All of these flags are normally defined in a confMAPDEF setting in your 165site.config.m4. 166 167If you define NEWDB or HESIOD you get the User Database (USERDB) 168automatically. Generally you do want to have NEWDB for it to do 169anything interesting. See above for getting the Berkeley DB 170package (i.e., NEWDB). There is no separate "user database" 171package -- don't bother searching for it on the net. 172 173Hesiod and LDAP require libraries that may not be installed with your 174system. These are outside of my ability to provide support. See the 175"Quirks" section for more information. 176 177The regex map can be used to see if an address matches a certain regular 178expression. For example, all-numerics local parts are common spam 179addresses, so "^[0-9]+$" would match this. By using such a map in a 180check_* rule-set, you can block a certain range of addresses that would 181otherwise be considered valid. 182 183 184+---------------+ 185| COMPILE FLAGS | 186+---------------+ 187 188Wherever possible, I try to make sendmail pull in the correct 189compilation options needed to compile on various environments based on 190automatically defined symbols. Some machines don't seem to have useful 191symbols available, requiring that a compilation flag be defined in 192the Makefile; see the devtools/OS subdirectory for the supported 193architectures. 194 195If you are a system to which sendmail has already been ported you 196should not have to touch the following symbols. But if you are porting, 197you may have to tweak the following compilation flags in conf.h in order 198to get it to compile and link properly: 199 200SYSTEM5 Adjust for System V (not necessarily Release 4). 201SYS5SIGNALS Use System V signal semantics -- the signal handler 202 is automatically dropped when the signal is caught. 203 If this is not set, use POSIX/BSD semantics, where the 204 signal handler stays in force until an exec or an 205 explicit delete. Implied by SYSTEM5. 206SYS5SETPGRP Use System V setpgrp() semantics. Implied by SYSTEM5. 207HASNICE Define this to zero if you lack the nice(2) system call. 208HASRRESVPORT Define this to zero if you lack the rresvport(3) system call. 209HASFCHMOD Define this to one if you have the fchmod(2) system call. 210 This improves security. 211HASFCHOWN Define this to one if you have the fchown(2) system call. 212 This is required for the TrustedUser option if sendmail 213 must rebuild an (alias) map. 214HASFLOCK Set this if you prefer to use the flock(2) system call 215 rather than using fcntl-based locking. Fcntl locking 216 has some semantic gotchas, but many vendor systems 217 also interface it to lockd(8) to do NFS-style locking. 218 Unfortunately, may vendors implementations of fcntl locking 219 is just plain broken (e.g., locks are never released, 220 causing your sendmail to deadlock; when the kernel runs 221 out of locks your system crashes). For this reason, I 222 recommend always defining this unless you are absolutely 223 certain that your fcntl locking implementation really works. 224HASUNAME Set if you have the "uname" system call. Implied by 225 SYSTEM5. 226HASUNSETENV Define this if your system library has the "unsetenv" 227 subroutine. 228HASSETSID Define this if you have the setsid(2) system call. This 229 is implied if your system appears to be POSIX compliant. 230HASINITGROUPS Define this if you have the initgroups(3) routine. 231HASSETVBUF Define this if you have the setvbuf(3) library call. 232 If you don't, setlinebuf will be used instead. This 233 defaults on if your compiler defines __STDC__. 234HASSETREUID Define this if you have setreuid(2) ***AND*** root can 235 use setreuid to change to an arbitrary user. This second 236 condition is not satisfied on AIX 3.x. You may find that 237 your system has setresuid(2), (for example, on HP-UX) in 238 which case you will also have to #define setreuid(r, e) 239 to be the appropriate call. Some systems (such as Solaris) 240 have a compatibility routine that doesn't work properly, 241 but may have "saved user ids" properly implemented so you 242 can ``#define setreuid(r, e) seteuid(e)'' and have it work. 243 The important thing is that you have a call that will set 244 the effective uid independently of the real or saved uid 245 and be able to set the effective uid back again when done. 246 There's a test program in ../test/t_setreuid.c that will 247 try things on your system. Setting this improves the 248 security, since sendmail doesn't have to read .forward 249 and :include: files as root. There are certain attacks 250 that may be unpreventable without this call. 251USESETEUID Define this to 1 if you have a seteuid(2) system call that 252 will allow root to set only the effective user id to an 253 arbitrary value ***AND*** you have saved user ids. This is 254 preferable to HASSETREUID if these conditions are fulfilled. 255 These are the semantics of the to-be-released revision of 256 Posix.1. The test program ../test/t_seteuid.c will try 257 this out on your system. If you define both HASSETREUID 258 and USESETEUID, the former is ignored. 259HASSETEGID Define this if you have setegid(2) and it can be 260 used to set the saved gid. Please run t_dropgid in 261 test/ if you are not sure whether the call works. 262HASSETREGID Define this if you have setregid(2) and it can be 263 used to set the saved gid. Please run t_dropgid in 264 test/ if you are not sure whether the call works. 265HASSETRESGID Define this if you have setresgid(2) and it can be 266 used to set the saved gid. Please run t_dropgid in 267 test/ if you are not sure whether the call works. 268HASLSTAT Define this if you have symbolic links (and thus the 269 lstat(2) system call). This improves security. Unlike 270 most other options, this one is on by default, so you 271 need to #undef it in conf.h if you don't have symbolic 272 links (these days everyone does). 273HASSETRLIMIT Define this to 1 if you have the setrlimit(2) syscall. 274 You can define it to 0 to force it off. It is assumed 275 if you are running a BSD-like system. 276HASULIMIT Define this if you have the ulimit(2) syscall (System V 277 style systems). HASSETRLIMIT overrides, as it is more 278 general. 279HASWAITPID Define this if you have the waitpid(2) syscall. 280HASGETDTABLESIZE 281 Define this if you have the getdtablesize(2) syscall. 282HAS_ST_GEN Define this to 1 if your system has the st_gen field in 283 the stat structure (see stat(2)). 284HASSRANDOMDEV Define this if your system has the srandomdev(3) function 285 call. 286HASURANDOMDEV Define this if your system has /dev/urandom(4). 287HASSTRERROR Define this if you have the libc strerror(3) function (which 288 should be declared in <errno.h>), and it should be used 289 instead of sys_errlist. 290SM_CONF_GETOPT Define this as 0 if you need a reimplementation of getopt(3). 291 On some systems, getopt does very odd things if called 292 to scan the arguments twice. This flag will ask sendmail 293 to compile in a local version of getopt that works 294 properly. You may also need this if you build with 295 another library that introduces a non-standard getopt(3). 296NEEDSTRTOL Define this if your standard C library does not define 297 strtol(3). This will compile in a local version. 298NEEDFSYNC Define this if your standard C library does not define 299 fsync(2). This will try to simulate the operation using 300 fcntl(2); if that is not available it does nothing, which 301 isn't great, but at least it compiles and runs. 302HASGETUSERSHELL Define this to 1 if you have getusershell(3) in your 303 standard C library. If this is not defined, or is defined 304 to be 0, sendmail will scan the /etc/shells file (no 305 NIS-style support, defaults to /bin/sh and /bin/csh if 306 that file does not exist) to get a list of unrestricted 307 user shells. This is used to determine whether users 308 are allowed to forward their mail to a program or a file. 309NEEDPUTENV Define this if your system needs am emulation of the 310 putenv(3) call. Define to 1 to implement it in terms 311 of setenv(3) or to 2 to do it in terms of primitives. 312NOFTRUNCATE Define this if you don't have the ftruncate(2) syscall. 313 If you don't have this system call, there is an unavoidable 314 race condition that occurs when creating alias databases. 315GIDSET_T The type of entries in a gidset passed as the second 316 argument to getgroups(2). Historically this has been an 317 int, so this is the default, but some systems (such as 318 IRIX) pass it as a gid_t, which is an unsigned short. 319 This will make a difference, so it is important to get 320 this right! However, it is only an issue if you have 321 group sets. 322SLEEP_T The type returned by the system sleep() function. 323 Defaults to "unsigned int". Don't worry about this 324 if you don't have compilation problems. 325ARBPTR_T The type of an arbitrary pointer -- defaults to "void *". 326 If you are an very old compiler you may need to define 327 this to be "char *". 328SOCKADDR_LEN_T The type used for the third parameter to accept(2), 329 getsockname(2), and getpeername(2), representing the 330 length of a struct sockaddr. Defaults to int. 331SOCKOPT_LEN_T The type used for the fifth parameter to getsockopt(2) 332 and setsockopt(2), representing the length of the option 333 buffer. Defaults to int. 334LA_TYPE The type of load average your kernel supports. These 335 can be one of: 336 LA_ZERO (1) -- it always returns the load average as 337 "zero" (and does so on all architectures). 338 LA_INT (2) to read /dev/kmem for the symbol avenrun and 339 interpret as a long integer. 340 LA_FLOAT (3) same, but interpret the result as a floating 341 point number. 342 LA_SHORT (6) to interpret as a short integer. 343 LA_SUBR (4) if you have the getloadavg(3) routine in your 344 system library. 345 LA_MACH (5) to use MACH-style load averages (calls 346 processor_set_info()), 347 LA_PROCSTR (7) to read /proc/loadavg and interpret it 348 as a string representing a floating-point 349 number (Linux-style). 350 LA_READKSYM (8) is an implementation suitable for some 351 versions of SVr4 that uses the MIOC_READKSYM ioctl 352 call to read /dev/kmem. 353 LA_DGUX (9) is a special implementation for DG/UX that uses 354 the dg_sys_info system call. 355 LA_HPUX (10) is an HP-UX specific version that uses the 356 pstat_getdynamic system call. 357 LA_IRIX6 (11) is an IRIX 6.x specific version that adapts 358 to 32 or 64 bit kernels; it is otherwise very similar 359 to LA_INT. 360 LA_KSTAT (12) uses the (Solaris-specific) kstat(3k) 361 implementation. 362 LA_DEVSHORT (13) reads a short from a system file (default: 363 /dev/table/avenrun) and scales it in the same manner 364 as LA_SHORT. 365 LA_INT, LA_SHORT, LA_FLOAT, and LA_READKSYM have several 366 other parameters that they try to divine: the name of your 367 kernel, the name of the variable in the kernel to examine, 368 the number of bits of precision in a fixed point load average, 369 and so forth. LA_DEVSHORT uses _PATH_AVENRUN to find the 370 device to be read to find the load average. 371 In desperation, use LA_ZERO. The actual code is in 372 conf.c -- it can be tweaked if you are brave. 373FSHIFT For LA_INT, LA_SHORT, and LA_READKSYM, this is the number 374 of bits of load average after the binary point -- i.e., 375 the number of bits to shift right in order to scale the 376 integer to get the true integer load average. Defaults to 8. 377_PATH_UNIX The path to your kernel. Needed only for LA_INT, LA_SHORT, 378 and LA_FLOAT. Defaults to "/unix" on System V, "/vmunix" 379 everywhere else. 380LA_AVENRUN For LA_INT, LA_SHORT, and LA_FLOAT, the name of the kernel 381 variable that holds the load average. Defaults to "avenrun" 382 on System V, "_avenrun" everywhere else. 383SFS_TYPE Encodes how your kernel can locate the amount of free 384 space on a disk partition. This can be set to SFS_NONE 385 (0) if you have no way of getting this information, 386 SFS_USTAT (1) if you have the ustat(2) system call, 387 SFS_4ARGS (2) if you have a four-argument statfs(2) 388 system call (and the include file is <sys/statfs.h>), 389 SFS_VFS (3), SFS_MOUNT (4), SFS_STATFS (5) if you have 390 the two-argument statfs(2) system call with includes in 391 <sys/vfs.h>, <sys/mount.h>, or <sys/statfs.h> respectively, 392 or SFS_STATVFS (6) if you have the two-argument statvfs(2) 393 call. The default if nothing is defined is SFS_NONE. 394SFS_BAVAIL with SFS_4ARGS you can also set SFS_BAVAIL to the field name 395 in the statfs structure that holds the useful information; 396 this defaults to f_bavail. 397SPT_TYPE Encodes how your system can display what a process is doing 398 on a ps(1) command (SPT stands for Set Process Title). Can 399 be set to: 400 SPT_NONE (0) -- Don't try to set the process title at all. 401 SPT_REUSEARGV (1) -- Pad out your argv with the information; 402 this is the default if none specified. 403 SPT_BUILTIN (2) -- The system library has setproctitle. 404 SPT_PSTAT (3) -- Use the PSTAT_SETCMD option to pstat(2) 405 to set the process title; this is used by HP-UX. 406 SPT_PSSTRINGS (4) -- Use the magic PS_STRINGS pointer (4.4BSD). 407 SPT_SYSMIPS (5) -- Use sysmips() supported by NEWS-OS 6. 408 SPT_SCO (6) -- Write kernel u. area. 409 SPT_CHANGEARGV (7) -- Write pointers to our own strings into 410 the existing argv vector. 411SPT_PADCHAR Character used to pad the process title; if undefined, 412 the space character (0x20) is used. This is ignored if 413 SPT_TYPE != SPT_REUSEARGV 414ERRLIST_PREDEFINED 415 If set, assumes that some header file defines sys_errlist. 416 This may be needed if you get type conflicts on this 417 variable -- otherwise don't worry about it. 418WAITUNION The wait(2) routine takes a "union wait" argument instead 419 of an integer argument. This is for compatibility with 420 old versions of BSD. 421SCANF You can set this to extend the F command to accept a 422 scanf string -- this gives you a primitive parser for 423 class definitions -- BUT it can make you vulnerable to 424 core dumps if the target file is poorly formed. 425SYSLOG_BUFSIZE You can define this to be the size of the buffer that 426 syslog accepts. If it is not defined, it assumes a 427 1024-byte buffer. If the buffer is very small (under 428 256 bytes) the log message format changes -- each 429 e-mail message will log many more messages, since it 430 will log each piece of information as a separate line 431 in syslog. 432BROKEN_RES_SEARCH 433 On Ultrix (and maybe other systems?) if you use the 434 res_search routine with an unknown host name, it returns 435 -1 but sets h_errno to 0 instead of HOST_NOT_FOUND. If 436 you set this, sendmail considers 0 to be the same as 437 HOST_NOT_FOUND. 438NAMELISTMASK If defined, values returned by nlist(3) are masked 439 against this value before use -- a common value is 440 0x7fffffff to strip off the top bit. 441BSD4_4_SOCKADDR If defined, socket addresses have an sa_len field that 442 defines the length of this address. 443SAFENFSPATHCONF Set this to 1 if and only if you have verified that a 444 pathconf(2) call with _PC_CHOWN_RESTRICTED argument on an 445 NFS filesystem where the underlying system allows users to 446 give away files to other users returns <= 0. Be sure you 447 try both on NFS V2 and V3. Some systems assume that their 448 local policy apply to NFS servers -- this is a bad 449 assumption! The test/t_pathconf.c program will try this 450 for you -- you have to run it in a directory that is 451 mounted from a server that allows file giveaway. 452SIOCGIFCONF_IS_BROKEN 453 Set this if your system has an SIOCGIFCONF ioctl defined, 454 but it doesn't behave the same way as "most" systems (BSD, 455 Solaris, SunOS, HP-UX, etc.) 456SIOCGIFNUM_IS_BROKEN 457 Set this if your system has an SIOCGIFNUM ioctl defined, 458 but it doesn't behave the same way as "most" systems 459 (Solaris, HP-UX). 460FAST_PID_RECYCLE 461 Set this if your system can reuse the same PID in the same 462 second. 463SO_REUSEADDR_IS_BROKEN 464 Set this if your system has a setsockopt() SO_REUSEADDR 465 flag but doesn't pay attention to it when trying to bind a 466 socket to a recently closed port. 467NEEDSGETIPNODE Set this if your system supports IPv6 but doesn't include 468 the getipnodeby{name,addr}() functions. Set automatically 469 for Linux's glibc. 470PIPELINING Support SMTP PIPELINING (set by default). 471USING_NETSCAPE_LDAP 472 Deprecated in favor of SM_CONF_LDAP_MEMFREE. See 473 libsm/README. 474NEEDLINK Set this if your system doesn't have a link() call. It 475 will create a copy of the file instead of a hardlink. 476USE_ENVIRON Set this to 1 to access process environment variables from 477 the external variable environ instead of the third 478 parameter of main(). 479USE_DOUBLE_FORK By default this is on (1). Set it to 0 to suppress the 480 extra fork() used to avoid intermediate zombies. 481ALLOW_255 Do not convert (char)0xff to (char)0x7f in headers etc. 482 This can also be done at runtime with the command line 483 option -d82.101. 484 485 486+-----------------------+ 487| COMPILE-TIME FEATURES | 488+-----------------------+ 489 490There are a bunch of features that you can decide to compile in, such 491as selecting various database packages and special protocol support. 492Several are assumed based on other compilation flags -- if you want to 493"un-assume" something, you probably need to edit conf.h. Compilation 494flags that add support for special features include: 495 496NDBM Include support for "new" DBM library for aliases and maps. 497 Normally defined in the Makefile. 498NEWDB Include support for Berkeley DB package (hash & btree) 499 for aliases and maps. Normally defined in the Makefile. 500 If the version of NEWDB you have is the old one that does 501 not include the "fd" call (this call was added in version 502 1.5 of the Berkeley DB code), you must upgrade to the 503 current version of Berkeley DB. 504NIS Define this to get NIS (YP) support for aliases and maps. 505 Normally defined in the Makefile. 506NISPLUS Define this to get NIS+ support for aliases and maps. 507 Normally defined in the Makefile. 508HESIOD Define this to get Hesiod support for aliases and maps. 509 Normally defined in the Makefile. 510NETINFO Define this to get NeXT NetInfo support for aliases and maps. 511 Normally defined in the Makefile. 512LDAPMAP Define this to get LDAP support for maps. 513PH_MAP Define this to get PH support for maps. 514MAP_NSD Define this to get nsd support for maps. 515USERDB Define this to 1 to include support for the User Information 516 Database. Implied by NEWDB or HESIOD. You can use 517 -DUSERDB=0 to explicitly turn it off. 518IDENTPROTO Define this as 1 to get IDENT (RFC 1413) protocol support. 519 This is assumed unless you are running on Ultrix or 520 HP-UX, both of which have a problem in the UDP 521 implementation. You can define it to be 0 to explicitly 522 turn off IDENT protocol support. If defined off, the code 523 is actually still compiled in, but it defaults off; you 524 can turn it on by setting the IDENT timeout in the 525 configuration file. 526IP_SRCROUTE Define this to 1 to get IP source routing information 527 displayed in the Received: header. This is assumed on 528 most systems, but some (e.g., Ultrix) apparently have a 529 broken version of getsockopt that doesn't properly 530 support the IP_OPTIONS call. You probably want this if 531 your OS can cope with it. Symptoms of failure will be that 532 it won't compile properly (that is, no support for fetching 533 IP_OPTIONs), or it compiles but source-routed TCP connections 534 either refuse to open or open and hang for no apparent reason. 535 Ultrix and AIX3 are known to fail this way. 536LOG Set this to get syslog(3) support. Defined by default 537 in conf.h. You want this if at all possible. 538NETINET Set this to get TCP/IP support. Defined by default 539 in conf.h. You probably want this. 540NETINET6 Set this to get IPv6 support. Other configuration may 541 be needed in conf.h for your particular operating system. 542 Also, DaemonPortOptions must be set appropriately for 543 sendmail to accept IPv6 connections. 544NETISO Define this to get ISO networking support. 545NETUNIX Define this to get Unix domain networking support. Defined 546 by default. A few bizarre systems (SCO, ISC, Altos) don't 547 support this networking domain. 548NETNS Define this to get NS networking support. 549NETX25 Define this to get X.25 networking support. 550NAMED_BIND If non-zero, include DNS (name daemon) support, including 551 MX support. The specs say you must use this if you run 552 SMTP. You don't have to be running a name server daemon 553 on your machine to need this -- any use of the DNS resolver, 554 including remote access to another machine, requires this 555 option. Defined by default in conf.h. Define it to zero 556 ONLY on machines that do not use DNS in any way. 557MATCHGECOS Permit fuzzy matching of user names against the full 558 name (GECOS) field in the /etc/passwd file. This should 559 probably be on, since you can disable it from the config 560 file if you want to. Defined by default in conf.h. 561MIME8TO7 If non-zero, include 8 to 7 bit MIME conversions. This 562 also controls advertisement of 8BITMIME in the ESMTP 563 startup dialogue. 564MIME7TO8 If non-zero, include 7 to 8 bit MIME conversions. 565HES_GETMAILHOST Define this to 1 if you are using Hesiod with the 566 hes_getmailhost() routine. This is included with the MIT 567 Hesiod distribution, but not with the DEC Hesiod distribution. 568XDEBUG Do additional internal checking. These don't cost too 569 much; you might as well leave this on. 570TCPWRAPPERS Turns on support for the TCP wrappers library (-lwrap). 571 See below for further information. 572SECUREWARE Enable calls to the SecureWare luid enabling/changing routines. 573 SecureWare is a C2 security package added to several UNIX's 574 (notably ConvexOS) to get a C2 Secure system. This 575 option causes mail delivery to be done with the luid of the 576 recipient. 577SHARE_V1 Support for the fair share scheduler, version 1. Setting to 578 1 causes final delivery to be done using the recipients 579 resource limitations. So far as I know, this is only 580 supported on ConvexOS. 581SASL Enables SMTP AUTH (RFC 2554). This requires the Cyrus SASL 582 library (ftp://ftp.andrew.cmu.edu/pub/cyrus-mail/). Please 583 install at least version 1.5.13. See below for further 584 information: SASL COMPILATION AND CONFIGURATION. If your 585 SASL library is older than 1.5.10, you have to set this 586 to its version number using a simple conversion: a.b.c 587 -> c + b*100 + a*10000, e.g. for 1.5.9 define SASL=10509. 588 Note: Using an older version than 1.5.5 of Cyrus SASL is 589 not supported. Starting with version 1.5.10, setting SASL=1 590 is sufficient. Any value other than 1 (or 0) will be 591 compared with the actual version found and if there is a 592 mismatch, compilation will fail. 593EGD Define this if your system has EGD installed, see 594 http://egd.sourceforge.net/ . It should be used to 595 seed the PRNG for STARTTLS if HASURANDOMDEV is not defined. 596STARTTLS Enables SMTP STARTTLS (RFC 2487). This requires OpenSSL 597 (http://www.OpenSSL.org/); use OpenSSL 0.9.5a or later 598 (if compatible with this version), do not use 0.9.3. 599 See STARTTLS COMPILATION AND CONFIGURATION for further 600 information. 601TLS_NO_RSA Turn off support for RSA algorithms in STARTTLS. 602MILTER Turn on support for external filters using the Milter API. 603 See libmilter/README for more information. 604REQUIRES_DIR_FSYNC Turn on support for file systems that require to 605 call fsync() for a directory if the meta-data in it has 606 been changed. This should be turned on at least for older 607 versions of ReiserFS; it is enabled by default for Linux. 608 According to some information this flag is not needed 609 anymore for kernel 2.4.16 and newer. We would appreciate 610 feedback about the semantics of the various file systems 611 available for Linux. 612 An alternative to this compile time flag is to mount the 613 queue directory without the -async option, or using 614 chattr +S on Linux. 615DBMMODE The default file permissions to use when creating new 616 database files for maps and aliases. Defaults to 0640. 617 618Generic notice: If you enable a compile time option that needs 619libraries or include files that don't come with sendmail or are 620installed in a location that your C compiler doesn't use by default 621you should set confINCDIRS and confLIBDIRS as explained in the 622first section: BUILDING SENDMAIL. 623 624 625+---------------------+ 626| DNS/RESOLVER ISSUES | 627+---------------------+ 628 629Many systems have old versions of the resolver library. At a minimum, 630you should be running BIND 4.8.3; older versions may compile, but they 631have known bugs that should give you pause. 632 633Common problems in old versions include "undefined" errors for 634dn_skipname. 635 636Some people have had a problem with BIND 4.9; it uses some routines 637that it expects to be externally defined such as strerror(). It may 638help to link with "-l44bsd" to solve this problem. This has apparently 639been fixed in later versions of BIND, starting around 4.9.3. In other 640words, if you use 4.9.0 through 4.9.2, you need -l44bsd; for earlier or 641later versions, you do not. 642 643!PLEASE! be sure to link with the same version of the resolver as 644the header files you used -- some people have used the 4.9 headers 645and linked with BIND 4.8 or vice versa, and it doesn't work. 646Unfortunately, it doesn't fail in an obvious way -- things just 647subtly don't work. 648 649WILDCARD MX RECORDS ARE A BAD IDEA! The only situation in which they 650work reliably is if you have two versions of DNS, one in the real world 651which has a wildcard pointing to your firewall, and a completely 652different version of the database internally that does not include 653wildcard MX records that match your domain. ANYTHING ELSE WILL GIVE 654YOU HEADACHES! 655 656When attempting to canonify a hostname, some broken name servers will 657return SERVFAIL (a temporary failure) on T_AAAA (IPv6) lookups. If you 658want to excuse this behavior, include WorkAroundBrokenAAAA in 659ResolverOptions. However, instead, we recommend catching the problem and 660reporting it to the name server administrator so we can rid the world of 661broken name servers. 662 663 664+----------------------------------------+ 665| STARTTLS COMPILATION AND CONFIGURATION | 666+----------------------------------------+ 667 668Please read the documentation accompanying the OpenSSL library. You 669have to compile and install the OpenSSL libraries before you can compile 670sendmail. See devtools/README how to set the correct compile time 671parameters; you should at least set the following variables: 672 673APPENDDEF(`conf_sendmail_ENVDEF', `-DSTARTTLS') 674APPENDDEF(`conf_sendmail_LIBS', `-lssl -lcrypto') 675 676If you have installed the OpenSSL libraries and include files in 677a location that your C compiler doesn't use by default you should 678set confINCDIRS and confLIBDIRS as explained in the first section: 679BUILDING SENDMAIL. 680 681Configuration information can be found in doc/op/op.me (required 682certificates) and cf/README (how to tell sendmail about certificates). 683 684To perform an initial test, connect to your sendmail daemon 685(telnet localhost 25) and issue a EHLO localhost and see whether 686250-STARTTLS 687is in the response. If it isn't, run the daemon with 688-O LogLevel=14 689and try again. Then take a look at the logfile and see whether 690there are any problems listed about permissions (unsafe files) 691or the validity of X.509 certificates. 692 693From: Garrett Wollman <wollman@lcs.mit.edu> 694 695 If your certificate authority is hierarchical, and you only include 696 the top-level CA certificate in the CACertFile file, some mail clients 697 may be unable to infer the proper certificate chain when selecting a 698 client certificate. Including the bottom-level CA certificate(s) in 699 the CACertFile file will allow these clients to work properly. This 700 is not necessary if you are not using client certificates for 701 authentication, or if all your clients are running Sendmail or other 702 programs using the OpenSSL library (which get it right automatically). 703 In addition, some mail clients are totally incapable of using 704 certificate authentication -- even some of those which already support 705 SSL/TLS for confidentiality. 706 707Further information can be found via: 708http://www.sendmail.org/tips/ 709 710 711+------------------------------------+ 712| SASL COMPILATION AND CONFIGURATION | 713+------------------------------------+ 714 715Please read the documentation accompanying the Cyrus SASL library 716(INSTALL and README). If you use Berkeley DB for Cyrus SASL then 717you must compile sendmail with the same version of Berkeley DB. 718See devtools/README for how to set the correct compile time parameters; 719you should at least set the following variables: 720 721APPENDDEF(`conf_sendmail_ENVDEF', `-DSASL') 722APPENDDEF(`conf_sendmail_LIBS', `-lsasl') 723 724If you have installed the Cyrus SASL library and include files in 725a location that your C compiler doesn't use by default you should 726set confINCDIRS and confLIBDIRS as explained in the first section: 727BUILDING SENDMAIL. 728 729You have to select and install authentication mechanisms and tell 730sendmail where to find the sasl library and the include files (see 731devtools/README for the parameters to set). Set up the required 732users and passwords as explained in the SASL documentation. See 733also cf/README for authentication related options (especially 734DefaultAuthInfo if you want authentication between MTAs). 735 736To perform an initial test, connect to your sendmail daemon 737(telnet localhost 25) and issue a EHLO localhost and see whether 738250-AUTH .... 739is in the response. If it isn't, run the daemon with 740-O LogLevel=14 741and try again. Then take a look at the logfile and see whether 742there are any security related problems listed (unsafe files). 743 744Further information can be found via: 745http://www.sendmail.org/tips/ 746 747 748+-------------------------------------+ 749| OPERATING SYSTEM AND COMPILE QUIRKS | 750+-------------------------------------+ 751 752GCC problems 753 When compiling with "gcc -O -Wall" specify "-DSM_OMIT_BOGUS_WARNINGS" 754 too (see include/sm/cdefs.h for more info). 755 756 ***************************************************************** 757 ** IMPORTANT: DO NOT USE OPTIMIZATION (``-O'') IF YOU ARE ** 758 ** RUNNING GCC 2.4.x or 2.5.x. THERE IS A BUG IN THE GCC ** 759 ** OPTIMIZER THAT CAUSES SENDMAIL COMPILES TO FAIL MISERABLY. ** 760 ***************************************************************** 761 762 Jim Wilson of Cygnus believes he has found the problem -- it will 763 probably be fixed in GCC 2.5.6 -- but until this is verified, be 764 very suspicious of gcc -O. This problem is reported to have been 765 fixed in gcc 2.6. 766 767 A bug in gcc 2.5.5 caused problems compiling sendmail 8.6.5 with 768 optimization on a Sparc. If you are using gcc 2.5.5, youi should 769 upgrade to the latest version of gcc. 770 771 Apparently GCC 2.7.0 on the Pentium processor has optimization 772 problems. I recommend against using -O on that architecture. This 773 has been seen on FreeBSD 2.0.5 RELEASE. 774 775 Solaris 2.X users should use version 2.7.2.3 over 2.7.2. 776 777 We have been told there are problems with gcc 2.8.0. If you are 778 using this version, you should upgrade to 2.8.1 or later. 779 780Berkeley DB 781 Berkeley DB 4.1.x with x <= 24 does not work with sendmail. 782 You need at least 4.1.25. 783 784GDBM GDBM does not work with sendmail because the additional 785 security checks and file locking cause problems. Unfortunately, 786 gdbm does not provide a compile flag in its version of ndbm.h so 787 the code can adapt. Until the GDBM authors can fix these problems, 788 GDBM will not be supported. Please use Berkeley DB instead. 789 790Configuration file location 791 Up to 8.6, sendmail tried to find the sendmail.cf file in the same 792 place as the vendors had put it, even when this was obviously 793 stupid. As of 8.7, sendmail ALWAYS looks for /etc/sendmail.cf. 794 Beginning with 8.10, sendmail uses /etc/mail/sendmail.cf. 795 You can get sendmail to use the stupid vendor .cf location by 796 adding -DUSE_VENDOR_CF_PATH during compilation, but this may break 797 support programs and scripts that need to find sendmail.cf. You 798 are STRONGLY urged to use symbolic links if you want to use the 799 vendor location rather than changing the location in the sendmail 800 binary. 801 802 NETINFO systems use NETINFO to determine the location of 803 sendmail.cf. The full path to sendmail.cf is stored as the value of 804 the "sendmail.cf" property in the "/locations/sendmail" 805 subdirectory of NETINFO. Set the value of this property to 806 "/etc/mail/sendmail.cf" (without the quotes) to use this new 807 default location for Sendmail 8.10.0 and higher. 808 809ControlSocket permissions 810 Paraphrased from BIND 8.2.1's README: 811 812 Solaris and other pre-4.4BSD kernels do not respect ownership or 813 protections on UNIX-domain sockets. The short term fix for this is to 814 override the default path and put such control sockets into root- 815 owned directories which do not permit non-root to r/w/x through them. 816 The long term fix is for all kernels to upgrade to 4.4BSD semantics. 817 818HP MPE/iX 819 The MPE-specific code within sendmail emulates a set-user-id root 820 environment for the sendmail binary. But there is no root uid 0 on 821 MPE, nor is there any support for set-user-id programs. Even when 822 sendmail thinks it is running as uid 0, it will still have the file 823 access rights of the underlying non-zero uid, but because sendmail is 824 an MPE priv-mode program it will still be able to call setuid() to 825 successfully switch to a new uid. 826 827 MPE setgid() semantics don't quite work the way sendmail expects, so 828 special emulation is done here also. 829 830 This uid/gid emulation is enabled via the setuid/setgid file mode bits 831 which are not currently used by MPE. Code in libsm/mpeix.c examines 832 these bits and enables emulation if they have been set, i.e., 833 chmod u+s,g+s /SENDMAIL/CURRENT/SENDMAIL. 834 835SunOS 4.x (Solaris 1.x) 836 You may have to use -lresolv on SunOS. However, beware that 837 this links in a new version of gethostbyname that does not 838 understand NIS, so you must have all of your hosts in DNS. 839 840 Some people have reported problems with the SunOS version of 841 -lresolv and/or in.named, and suggest that you get a newer 842 version. The symptoms are delays when you connect to the 843 SMTP server on a SunOS machine or having your domain added to 844 addresses inappropriately. There is a version of BIND 845 version 4.9 on gatekeeper.DEC.COM in pub/BSD/bind/4.9. 846 847 There is substantial disagreement about whether you can make 848 this work with resolv+, which allows you to specify a search-path 849 of services. Some people report that it works fine, others 850 claim it doesn't work at all (including causing sendmail to 851 drop core when it tries to do multiple resolv+ lookups for a 852 single job). I haven't tried resolv+, as we use DNS exclusively. 853 854 Should you want to try resolv+, it is on ftp.uu.net in 855 /networking/ip/dns. 856 857 Apparently getservbyname() can fail under moderate to high 858 load under some circumstances. This will exhibit itself as 859 the message ``554 makeconnection: service "smtp" unknown''. 860 The problem has been traced to one or more blank lines in 861 /etc/services on the NIS server machine. Delete these 862 and it should work. This info is thanks to Brian Bartholomew 863 <bb@math.ufl.edu> of I-Kinetics, Inc. 864 865 NOTE: The SunOS 4.X linker uses library paths specified during 866 compilation using -L for run-time shared library searches. 867 Therefore, it is vital that relative and unsafe directory paths not 868 be used when compiling sendmail. 869 870SunOS 4.0.2 (Sun 386i) 871 Date: Fri, 25 Aug 1995 11:13:58 +0200 (MET DST) 872 From: teus@oce.nl 873 874 Sendmail 8.7.Beta.12 compiles and runs nearly out of the box with the 875 following changes: 876 * Don't use /usr/5bin in your PATH, but make /usr/5bin/uname 877 available as "uname" command. 878 * Use the defines "-DBSD4_3 -DNAMED_BIND=0" in 879 devtools/OS/SunOS.4.0, which is selected via the "uname" command. 880 I recommend to make available the db-library on the system first 881 (and change the Makefile to use this library). 882 Note that the sendmail.cf and aliases files are found in /etc. 883 884SunOS 4.1.3, 4.1.3_U1 885 Sendmail causes crashes on SunOS 4.1.3 and 4.1.3_U1. According 886 to Sun bug number 1077939: 887 888 If an application does a getsockopt() on a SOCK_STREAM (TCP) socket 889 after the other side of the connection has sent a TCP RESET for 890 the stream, the kernel gets a Bus Trap in the tcp_ctloutput() or 891 ip_ctloutput() routine. 892 893 For 4.1.3, this is fixed in patch 100584-08, available on the 894 Sunsolve 2.7.1 or later CDs. For 4.1.3_U1, this was fixed in patch 895 101790-01 (SunOS 4.1.3_U1: TCP socket and reset problems), later 896 obsoleted by patch 102010-05. 897 898 Sun patch 100584-08 is not currently publicly available on their 899 ftp site but a user has reported it can be found at other sites 900 using a web search engine. 901 902Solaris 2.x (SunOS 5.x) 903 To compile for Solaris, the Makefile built by Build must 904 include a SOLARIS definition which reflects the Solaris version 905 (i.e. -DSOLARIS=20400 for 2.4 or -DSOLARIS=20501 for 2.5.1). 906 If you are using gcc, make sure -I/usr/include is not used (or 907 it might complain about TopFrame). If you are using Sun's cc, 908 make sure /opt/SUNWspro/bin/cc is used instead of /usr/ucb/cc 909 (or it might complain about tm_zone). 910 911 The Solaris 2.x (x <= 3) "syslog" function is apparently limited 912 to something about 90 characters because of a kernel limitation. 913 If you have source code, you can probably up this number. You 914 can get patches that fix this problem: the patch ids are: 915 916 Solaris 2.1 100834 917 Solaris 2.2 100999 918 Solaris 2.3 101318 919 920 Be sure you have the appropriate patch installed or you won't 921 see system logging. 922 923Solaris 2.4 (SunOS 5.4) 924 If you include /usr/lib at the end of your LD_LIBRARY_PATH you run 925 the risk of getting the wrong libraries under some circumstances. 926 This is because of a new feature in Solaris 2.4, described by 927 Rod.Evans@Eng.Sun.COM: 928 929 >> Prior to SunOS 5.4, any LD_LIBRARY_PATH setting was ignored by the 930 >> runtime linker if the application was setxid (secure), thus your 931 >> applications search path would be: 932 >> 933 >> /usr/local/lib LD_LIBRARY_PATH component - IGNORED 934 >> /usr/lib LD_LIBRARY_PATH component - IGNORED 935 >> /usr/local/lib RPATH - honored 936 >> /usr/lib RPATH - honored 937 >> 938 >> the effect is that path 3 would be the first used, and this would 939 >> satisfy your resolv.so lookup. 940 >> 941 >> In SunOS 5.4 we made the LD_LIBRARY_PATH a little more flexible. 942 >> People who developed setxid applications wanted to be able to alter 943 >> the library search path to some degree to allow for their own 944 >> testing and debugging mechanisms. It was decided that the only 945 >> secure way to do this was to allow a `trusted' path to be used in 946 >> LD_LIBRARY_PATH. The only trusted directory we presently define 947 >> is /usr/lib. Thus a set-user-ID root developer could play with some 948 >> alternative shared object implementations and place them in 949 >> /usr/lib (being root we assume they'ed have access to write in this 950 >> directory). This change was made as part of 1155380 - after a 951 >> *huge* amount of discussion regarding the security aspect of things. 952 >> 953 >> So, in SunOS 5.4 your applications search path would be: 954 >> 955 >> /usr/local/lib from LD_LIBRARY_PATH - IGNORED (untrustworthy) 956 >> /usr/lib from LD_LIBRARY_PATH - honored (trustworthy) 957 >> /usr/local/lib from RPATH - honored 958 >> /usr/lib from RPATH - honored 959 >> 960 >> here, path 2 would be the first used. 961 962Solaris 2.5.1 (SunOS 5.5.1) and 2.6 (SunOS 5.6) 963 Apparently Solaris 2.5.1 patch 103663-01 installs a new 964 /usr/include/resolv.h file that defines the __P macro without 965 checking to see if it is already defined. This new resolv.h is also 966 included in the Solaris 2.6 distribution. This causes compile 967 warnings such as: 968 969 In file included from daemon.c:51: 970 /usr/include/resolv.h:208: warning: `__P' redefined 971 cdefs.h:58: warning: this is the location of the previous definition 972 973 These warnings can be safely ignored or you can create a resolv.h 974 file in the obj.SunOS.5.5.1.* or obj.SunOS.5.6.* directory that reads: 975 976 #undef __P 977 #include "/usr/include/resolv.h" 978 979 This problem was fixed in Solaris 7 (Sun bug ID 4081053). 980 981Solaris 7 (SunOS 5.7) 982 Solaris 7 includes LDAP libraries but the implementation was 983 lacking a few things. The following settings can be placed in 984 devtools/Site/site.SunOS.5.7.m4 if you plan on using those 985 libraries. 986 987 APPENDDEF(`confMAPDEF', `-DLDAPMAP') 988 APPENDDEF(`confENVDEF', `-DLDAP_VERSION_MAX=3') 989 APPENDDEF(`confLIBS', `-lldap') 990 991 Also, Sun's patch 107555 is needed to prevent a crash in the call 992 to ldap_set_option for LDAP_OPT_REFERRALS in ldapmap_setopts if 993 LDAP support is compiled in sendmail. 994 995Solaris 8 and later (SunOS 5.8 and later) 996 Solaris 8 and later can optionally install LDAP support. If you 997 have installed the Entire Distribution meta-cluster, you can use 998 the following in devtools/Site/site.SunOS.5.8.m4 (or other 999 appropriately versioned file) to enable LDAP: 1000 1001 APPENDDEF(`confMAPDEF', `-DLDAPMAP') 1002 APPENDDEF(`confLIBS', `-lldap') 1003 1004Solaris 9 and later (SunOS 5.9 and later) 1005 Solaris 9 and later have a revised LDAP library, libldap.so.5, 1006 which is derived from a Netscape implementation, thus requiring 1007 that SM_CONF_LDAP_MEMFREE be defined in conjunction with LDAPMAP: 1008 1009 APPENDDEF(`confMAPDEF', `-DLDAPMAP') 1010 APPENDDEF(`confENVDEF', `-DSM_CONF_LDAP_MEMFREE') 1011 APPENDDEF(`confLIBS', `-lldap') 1012 1013Solaris 1014 If you are using dns for hostname resolution on Solaris, make sure 1015 that the 'dns' entry is last on the hosts line in 1016 '/etc/nsswitch.conf'. For example, use: 1017 1018 hosts: nisplus files dns 1019 1020 Do not use: 1021 1022 hosts: nisplus dns [NOTFOUND=return] files 1023 1024 Note that 'nisplus' above is an illustration. The same comment 1025 applies no matter what naming services you are using. If you have 1026 anything other than dns last, even after "[NOTFOUND=return]", 1027 sendmail may not be able to determine whether an error was 1028 temporary or permanent. The error returned by the solaris 1029 gethostbyname() is the error for the last lookup used, and other 1030 naming services do not have the same concept of temporary failure. 1031 1032Ultrix 1033 By default, the IDENT protocol is turned off on Ultrix. If you 1034 are running Ultrix 4.4 or later, or if you have included patch 1035 CXO-8919 for Ultrix 4.2 or 4.3 to fix the TCP problem, you can turn 1036 IDENT on in the configuration file by setting the "ident" timeout. 1037 1038 The Ultrix 4.5 Y2K patch (ULTV45-022-1) has changed the resolver 1039 included in libc.a. Unfortunately, the __RES symbol hasn't changed 1040 and therefore, sendmail can no longer automatically detect the 1041 newer version. If you get a compiler error: 1042 1043 /lib/libc.a(gethostent.o): local_hostname_length: multiply defined 1044 1045 Then rebuild with this in devtools/Site/site.ULTRIX.m4: 1046 1047 APPENDDEF(`conf_sendmail_ENVDEF', `-DNEEDLOCAL_HOSTNAME_LENGTH=0') 1048 1049Digital UNIX (formerly DEC OSF/1) 1050 If you are compiling on OSF/1 (DEC Alpha), you must use 1051 -L/usr/shlib (otherwise it core dumps on startup). You may also 1052 need -mld to get the nlist() function, although some versions 1053 apparently don't need this. 1054 1055 Also, the enclosed makefile removed /usr/sbin/smtpd; if you need 1056 it, just create the link to the sendmail binary. 1057 1058 On DEC OSF/1 3.2 or earlier, the MatchGECOS option doesn't work 1059 properly due to a bug in the getpw* routines. If you want to use 1060 this, use -DDEC_OSF_BROKEN_GETPWENT=1. The problem is fixed in 3.2C. 1061 1062 Digital's mail delivery agent, /bin/mail (aka /bin/binmail), will 1063 only preserve the envelope sender in the "From " header if 1064 DefaultUserID is set to daemon. Setting this to mailnull will 1065 cause all mail to have the header "From mailnull ...". To use 1066 a different DefaultUserID, you will need to use a different mail 1067 delivery agent (such as mail.local found in the sendmail 1068 distribution). 1069 1070 On Digital UNIX 4.0 and later, Berkeley DB 1.85 is included with the 1071 operating system and already has the ndbm.o module removed. However, 1072 Digital has modified the original Berkeley DB db.h include file. 1073 This results in the following warning while compiling map.c and udb.c: 1074 1075 cc: Warning: /usr/include/db.h, line 74: The redefinition of the macro 1076 "__signed" conflicts with a current definition because the replacement 1077 lists differ. The redefinition is now in effect. 1078 #define __signed signed 1079 ------------------------^ 1080 1081 This warning can be ignored. 1082 1083 Digital UNIX's linker checks /usr/ccs/lib/ before /usr/lib/. 1084 If you have installed a new version of BIND in /usr/include 1085 and /usr/lib, you will experience difficulties as Digital ships 1086 libresolv.a in /usr/ccs/lib/ as well. Be sure to replace both 1087 copies of libresolv.a. 1088 1089IRIX 1090 The header files on SGI IRIX are completely prototyped, and as 1091 a result you can sometimes get some warning messages during 1092 compilation. These can be ignored. There are two errors in 1093 deliver only if you are using gcc, both of the form ``warning: 1094 passing arg N of `execve' from incompatible pointer type''. 1095 Also, if you compile with -DNIS, you will get a complaint 1096 about a declaration of struct dom_binding in a prototype 1097 when compiling map.c; this is not important because the 1098 function being prototyped is not used in that file. 1099 1100 In order to compile sendmail you will have had to install 1101 the developers' option in order to get the necessary include 1102 files. 1103 1104 If you compile with -lmalloc (the fast memory allocator), you may 1105 get warning messages such as the following: 1106 1107 ld32: WARNING 85: definition of _calloc in /usr/lib32/libmalloc.so 1108 preempts that definition in /usr/lib32/mips3/libc.so. 1109 ld32: WARNING 85: definition of _malloc in /usr/lib32/libmalloc.so 1110 preempts that definition in /usr/lib32/mips3/libc.so. 1111 ld32: WARNING 85: definition of _realloc in /usr/lib32/libmalloc.so 1112 preempts that definition in /usr/lib32/mips3/libc.so. 1113 ld32: WARNING 85: definition of _free in /usr/lib32/libmalloc.so 1114 preempts that definition in /usr/lib32/mips3/libc.so. 1115 ld32: WARNING 85: definition of _cfree in /usr/lib32/libmalloc.so 1116 preempts that definition in /usr/lib32/mips3/libc.so. 1117 1118 These are unavoidable and innocuous -- just ignore them. 1119 1120 According to Dave Sill <de5@ornl.gov>, there is a version of the 1121 Berkeley DB library patched to run on Irix 6.2 available from 1122 http://reality.sgi.com/ariel/freeware/#db . 1123 1124IRIX 6.x 1125 If you are using XFS filesystem, avoid using the -32 ABI switch to 1126 the cc compiler if possible. 1127 1128 Broken inet_aton and inet_ntoa on IRIX using gcc: There's 1129 a problem with gcc on IRIX, i.e., gcc can't pass structs 1130 less than 16 bits long unless they are 8 bits; IRIX 6.2 has 1131 some other sized structs. See 1132 http://www.bitmechanic.com/mail-archives/mysql/current/0418.html 1133 This problem seems to be fixed by gcc v2.95.2, gcc v2.8.1 1134 is reported as broken. Check your gcc version for this bug 1135 before installing sendmail. 1136 1137IRIX 6.4 1138 The IRIX 6.5.4 version of /bin/m4 does not work properly with 1139 sendmail. Either install fw_m4.sw.m4 off the Freeware_May99 CD and 1140 use /usr/freeware/bin/m4 or install and use GNU m4. 1141 1142NeXT or NEXTSTEP 1143 NEXTSTEP 3.3 and earlier ship with the old DBM library. Also, 1144 Berkeley DB does not currently run on NEXTSTEP. 1145 1146 If you are compiling on NEXTSTEP, you will have to create an 1147 empty file "unistd.h" and create a file "dirent.h" containing: 1148 1149 #include <sys/dir.h> 1150 #define dirent direct 1151 1152 (devtools/OS/NeXT should try to do both of these for you.) 1153 1154 Apparently, there is a bug in getservbyname on Nextstep 3.0 1155 that causes it to fail under some circumstances with the 1156 message "SYSERR: service "smtp" unknown" logged. You should 1157 be able to work around this by including the line: 1158 1159 OOPort=25 1160 1161 in your .cf file. 1162 1163BSDI (BSD/386) 1.0, NetBSD 0.9, FreeBSD 1.0 1164 The "m4" from BSDI won't handle the config files properly. 1165 I haven't had a chance to test this myself. 1166 1167 The M4 shipped in FreeBSD and NetBSD 0.9 don't handle the config 1168 files properly. One must use either GNU m4 1.1 or the PD-M4 1169 recently posted in comp.os.386bsd.bugs (and maybe others). 1170 NetBSD-current includes the PD-M4 (as stated in the NetBSD file 1171 CHANGES). 1172 1173 FreeBSD 1.0 RELEASE has uname(2) now. Use -DUSEUNAME in order to 1174 use it (look into devtools/OS/FreeBSD). NetBSD-current may have 1175 it too but it has not been verified. 1176 1177 The latest version of Berkeley DB uses a different naming 1178 scheme than the version that is supplied with your release. This 1179 means you will be able to use the current version of Berkeley DB 1180 with sendmail as long you use the new db.h when compiling 1181 sendmail and link it against the new libdb.a or libdb.so. You 1182 should probably keep the original db.h in /usr/include and the 1183 new db.h in /usr/local/include. 1184 11854.3BSD 1186 If you are running a "virgin" version of 4.3BSD, you'll have 1187 a very old resolver and be missing some header files. The 1188 header files are simple -- create empty versions and everything 1189 will work fine. For the resolver you should really port a new 1190 version (4.8.3 or later) of the resolver; 4.9 is available on 1191 gatekeeper.DEC.COM in pub/BSD/bind/4.9. If you are really 1192 determined to continue to use your old, buggy version (or as 1193 a shortcut to get sendmail working -- I'm sure you have the 1194 best intentions to port a modern version of BIND), you can 1195 copy ../contrib/oldbind.compat.c into sendmail and add the 1196 following to devtools/Site/site.config.m4: 1197 1198 APPENDDEF(`confOBJADD', `oldbind.compat.o') 1199 1200OpenBSD (up to 2.9 Release), NetBSD, FreeBSD (up to 4.3-RELEASE) 1201 m4 from *BSD won't handle libsm/Makefile.m4 properly, since the 1202 maximum length for strings is too short. You need to use GNU m4 1203 or patch m4, see for example: 1204 http://FreeBSD.org/cgi/cvsweb.cgi/src/usr.bin/m4/eval.c.diff?r1=1.11&r2=1.12 1205 1206A/UX 1207 Date: Tue, 12 Oct 1993 18:28:28 -0400 (EDT) 1208 From: "Eric C. Hagberg" <hagberg@med.cornell.edu> 1209 Subject: Fix for A/UX ndbm 1210 1211 I guess this isn't really a sendmail bug, however, it is something 1212 that A/UX users should be aware of when compiling sendmail 8.6. 1213 1214 Apparently, the calls that sendmail is using to the ndbm routines 1215 in A/UX 3.0.x contain calls to "broken" routines, in that the 1216 aliases database will break when it gets "just a little big" 1217 (sorry I don't have exact numbers here, but it broke somewhere 1218 around 20-25 aliases for me.), making all aliases non-functional 1219 after exceeding this point. 1220 1221 What I did was to get the gnu-dbm-1.6 package, compile it, and 1222 then re-compile sendmail with "-lgdbm", "-DNDBM", and using the 1223 ndbm.h header file that comes with the gnu-package. This makes 1224 things behave properly. 1225 [NOTE: see comment above about GDBM] 1226 1227 I suppose porting the New Berkeley DB package is another route, 1228 however, I made a quick attempt at it, and found it difficult 1229 (not easy at least); the gnu-dbm package "configured" and 1230 compiled easily. 1231 1232 [NOTE: Berkeley DB version 2.X runs on A/UX and can be used for 1233 database maps.] 1234 1235SCO Unix 1236 From: Thomas Essebier <tom@stallion.oz.au> 1237 Organisation: Stallion Technologies Pty Ltd. 1238 1239 It will probably help those who are trying to configure sendmail 8.6.9 1240 to know that if they are on SCO, they had better set 1241 OI-dnsrch 1242 or they will core dump as soon as they try to use the resolver. 1243 i.e., although SCO has _res.dnsrch defined, and is kinda BIND 4.8.3, 1244 it does not inititialise it, nor does it understand 'search' in 1245 /etc/named.boot. 1246 - sigh - 1247 1248 According to SCO, the m4 which ships with UnixWare 2.1.2 is broken. 1249 We recommend installing GNU m4 before attempting to build sendmail. 1250 1251 On some versions a bogus error value is listed if connections 1252 time out (large negative number). To avoid this explicitly set 1253 Timeout.connect to a reasonable value (several minutes). 1254 1255DG/UX 1256 Doug Anderson <dlander@afterlife.ncsc.mil> has successfully run 1257 V8 on the DG/UX 5.4.2 and 5.4R3.x platforms under heavy usage. 1258 Originally, the DG /bin/mail program wasn't compatible with 1259 the V8 sendmail, since the DG /bin/mail requires the environment 1260 variable "_FORCE_MAIL_LOCAL_=yes" be set. Version 8.7 now includes 1261 this in the environment before invoking the local mailer. Some 1262 have used procmail to avoid this problem in the past. It works 1263 but some have experienced file locking problems with their DG/UX 1264 ports of procmail. 1265 1266Apollo DomainOS 1267 If you are compiling on Apollo, you will have to create an empty 1268 file "unistd.h" (for DomainOS 10.3 and earlier) and create a file 1269 "dirent.h" containing: 1270 1271 #include <sys/dir.h> 1272 #define dirent direct 1273 1274 (devtools/OS/DomainOS will attempt to do both of these for you.) 1275 1276HP-UX 8.00 1277 Date: Mon, 24 Jan 1994 13:25:45 +0200 1278 From: Kimmo Suominen <Kimmo.Suominen@lut.fi> 1279 Subject: 8.6.5 w/ HP-UX 8.00 on s300 1280 1281 Just compiled and fought with sendmail 8.6.5 on a HP9000/360 (i.e., 1282 a series 300 machine) running HP-UX 8.00. 1283 1284 I was getting segmentation fault when delivering to a local user. 1285 With debugging I saw it was faulting when doing _free@libc... *sigh* 1286 It seems the new implementation of malloc on s300 is buggy as of 8.0, 1287 so I tried out the one in -lmalloc (malloc(3X)). With that it seems 1288 to work just dandy. 1289 1290 When linking, you will get the following error: 1291 1292 ld: multiply defined symbol _freespace in file /usr/lib/libmalloc.a 1293 1294 but you can just ignore it. You might want to add this info to the 1295 README file for the future... 1296 1297Linux 1298 Something broke between versions 0.99.13 and 0.99.14 of Linux: the 1299 flock() system call gives errors. If you are running .14, you must 1300 not use flock. You can do this with -DHASFLOCK=0. We have also 1301 been getting complaints since version 2.4.X was released. Unless 1302 the bug is fixed before sendmail 8.13 is shipped, 8.13 will change 1303 the default locking method to fcntl() for Linux kernel version 2.4 1304 and later. Be sure to update other sendmail related programs to 1305 match locking techniques (some examples, besides makemap and 1306 mail.local, include procmail, mailx, mutt, elm, etc). 1307 1308 Around the inclusion of bind-4.9.3 & Linux libc-4.6.20, the 1309 initialization of the _res structure changed. If /etc/hosts.conf 1310 was configured as "hosts, bind" the resolver code could return 1311 "Name server failure" errors. This is supposedly fixed in 1312 later versions of libc (>= 4.6.29?), and later versions of 1313 sendmail (> 8.6.10) try to work around the problem. 1314 1315 Some older versions (< 4.6.20?) of the libc/include files conflict 1316 with sendmail's version of cdefs.h. Deleting sendmail's version 1317 on those systems should be non-harmful, and new versions don't care. 1318 1319 NOTE ON LINUX & BIND: By default, the Makefile generated for Linux 1320 includes header files in /usr/local/include and libraries in 1321 /usr/local/lib. If you've installed BIND on your system, the header 1322 files typically end up in the search path and you need to add 1323 "-lresolv" to the LIBS line in your Makefile. Really old versions 1324 may need to include "-l44bsd" as well (particularly if the link phase 1325 complains about missing strcasecmp, strncasecmp or strpbrk). 1326 Complaints about an undefined reference to `__dn_skipname' in 1327 domain.o are a sure sign that you need to add -lresolv to LIBS. 1328 Newer versions of Linux are basically threaded BIND, so you may or 1329 may not see complaints if you accidentally mix BIND 1330 headers/libraries with virginal libc. If you have BIND headers in 1331 /usr/local/include (resolv.h, etc) you *should* be adding -lresolv 1332 to LIBS. Data structures may change and you'd be asking for a 1333 core dump. 1334 1335 A number of problems have been reported regarding the Linux 2.2.0 1336 kernel. So far, these problems have been tracked down to syslog() 1337 and DNS resolution. We believe the problem is with the poll() 1338 implementation in the Linux 2.2.0 kernel and poll()-aware versions 1339 of glib (at least up to 2.0.111). 1340 1341glibc 1342 glibc 2.2.1 (and possibly other versions) changed the value of 1343 __RES in resolv.h but failed to actually provide the IPv6 API 1344 changes that the change implied. Therefore, compiling with 1345 -DNETINET6 fails. 1346 1347 Workarounds: 1348 1) Compile without -DNETINET6 1349 2) Build against a real BIND 8.2.2 include/lib tree 1350 3) Wait for glibc to fix it 1351 1352AIX 4.X 1353 The AIX 4.X linker uses library paths specified during compilation 1354 using -L for run-time shared library searches. Therefore, it is 1355 vital that relative and unsafe directory paths not be using when 1356 compiling sendmail. Because of this danger, by default, compiles 1357 on AIX use the -blibpath option to limit shared libraries to 1358 /usr/lib and /lib. If you need to allow more directories, such as 1359 /usr/local/lib, modify your devtools/Site/site.AIX.4.2.m4, 1360 site.AIX.4.3.m4, and/or site.AIX.4.x.m4 file(s) and set confLDOPTS 1361 appropriately. For example: 1362 1363 define(`confLDOPTS', `-blibpath:/usr/lib:/lib:/usr/local/lib') 1364 1365 Be sure to only add (safe) system directories. 1366 1367 The AIX version of GNU ld also exhibits this problem. If you are 1368 using that version, instead of -blibpath, use its -rpath option. 1369 For example: 1370 1371 gcc -Wl,-rpath /usr/lib -Wl,-rpath /lib -Wl,-rpath /usr/local/lib 1372 1373AIX 4.X If the test program t-event (and most others) in libsm fails, 1374 check your compiler settings. It seems that the flags -qnoro or 1375 -qnoroconst on some AIX versions trigger a compiler bug. Check 1376 your compiler settings or use cc instead of xlc. 1377 1378AIX 4.0-4.2, maybe some AIX 4.3 versions 1379 The AIX m4 implements a different mechanism for ifdef which is 1380 inconsistent with other versions of m4. Therefore, it will not 1381 work properly with the sendmail Build architecture or m4 1382 configuration method. To work around this problem, please use 1383 GNU m4 from ftp://ftp.gnu.org/pub/gnu/. 1384 The problem seems to be solved in AIX 4.3.3 at least. 1385 1386AIX 4.3.3 1387 From: Valdis.Kletnieks@vt.edu 1388 Date: Sun, 02 Jul 2000 03:58:02 -0400 1389 1390 Under AIX 4.3.3, after applying bos.adt.include 4.3.3.12 to close the 1391 BIND 8.2.2 security holes, you can no longer build with -DNETINET6 1392 because they changed the value of __RES in resolv.h but failed to 1393 actually provide the API changes that the change implied. 1394 1395 Workarounds: 1396 1) Compile without -DNETINET6 1397 2) Build against a real BIND 8.2.2 include/lib tree 1398 3) Wait for IBM to fix it 1399 1400AIX 3.x 1401 This version of sendmail does not support MB, MG, and MR resource 1402 records, which are supported by AIX sendmail. 1403 1404 Several people have reported that the IBM-supplied named returns 1405 fairly random results -- the named should be replaced. It is not 1406 necessary to replace the resolver, which will simplify installation. 1407 A new BIND resolver can be found at http://www.isc.org/isc/. 1408 1409AIX 3.1.x 1410 The supplied load average code only works correctly for AIX 3.2.x. 1411 For 3.1, use -DLA_TYPE=LA_SUBR and get the latest ``monitor'' 1412 package by Jussi Maki <jmaki@hut.fi> from ftp.funet.fi in the 1413 directory pub/unix/AIX/rs6000/monitor-1.12.tar.Z; use the loadavgd 1414 daemon, and the getloadavg subroutine supplied with that package. 1415 If you don't care about load average throttling, just turn off 1416 load average checking using -DLA_TYPE=LA_ZERO. 1417 1418RISC/os 1419 RISC/os from MIPS is a merged AT&T/Berkeley system. When you 1420 compile on that platform you will get duplicate definitions 1421 on many files. You can ignore these. 1422 1423System V Release 4 Based Systems 1424 There is a single devtools OS that is intended for all SVR4-based 1425 systems (built from devtools/OS/SVR4). It defines __svr4__, 1426 which is predefined by some compilers. If your compiler already 1427 defines this compile variable, you can delete the definition from 1428 the generated Makefile or create a devtools/Site/site.config.m4 1429 file. 1430 1431 It's been tested on Dell Issue 2.2. 1432 1433DELL SVR4 1434 Date: Mon, 06 Dec 1993 10:42:29 EST 1435 From: "Kimmo Suominen" <kim@grendel.lut.fi> 1436 Message-ID: <2d0352f9.lento29@lento29.UUCP> 1437 To: eric@cs.berkeley.edu 1438 Cc: sendmail@cs.berkeley.edu 1439 Subject: Notes for DELL SVR4 1440 1441 Eric, 1442 1443 Here are some notes for compiling Sendmail 8.6.4 on DELL SVR4. I ran 1444 across these things when helping out some people who contacted me by 1445 e-mail. 1446 1447 1) Use gcc 2.4.5 (or later?). Dell distributes gcc 2.1 with their 1448 Issue 2.2 Unix. It is too old, and gives you problems with 1449 clock.c, because sigset_t won't get defined in <sys/signal.h>. 1450 This is due to a problematic protection rule in there, and is 1451 fixed with gcc 2.4.5. 1452 1453 2) If you don't use the new Berkeley DB (-DNEWDB), then you need 1454 to add "-lc -lucb" to the libraries to link with. This is because 1455 the -ldbm distributed by Dell needs the bcopy, bcmp and bzero 1456 functions. It is important that you specify both libraries in 1457 the given order to be sure you only get the BSTRING functions 1458 from the UCB library (and not the signal routines etc.). 1459 1460 3) Don't leave out "-lelf" even if compiling with "-lc -lucb". 1461 The UCB library also has another copy of the nlist routines, 1462 but we do want the ones from "-lelf". 1463 1464 If anyone needs a compiled gcc 2.4.5 and/or a ported DB library, they 1465 can use anonymous ftp to fetch them from lut.fi in the /kim directory. 1466 They are copies of what I use on grendel.lut.fi, and offering them 1467 does not imply that I would also support them. I have sent the DB 1468 port for SVR4 back to Keith Bostic for inclusion in the official 1469 distribution, but I haven't heard anything from him as of today. 1470 1471 - gcc-2.4.5-svr4.tar.gz (gcc 2.4.5 and the corresponding libg++) 1472 - db-1.72.tar.gz (with source, objects and a installed copy) 1473 1474 Cheers 1475 + Kim 1476 -- 1477 * Kimmo.Suominen@lut.fi * SysVr4 enthusiast at GRENDEL.LUT.FI * 1478 * KIM@FINFILES.BITNET * Postmaster and Hostmaster at LUT.FI * 1479 * + 358 200 865 718 * Unix area moderator at NIC.FUNET.FI * 1480 1481ConvexOS 10.1 and below 1482 In order to use the name server, you must create the file 1483 /etc/use_nameserver. If this file does not exist, the call 1484 to res_init() will fail and you will have absolutely no 1485 access to DNS, including MX records. 1486 1487Amdahl UTS 2.1.5 1488 In order to get UTS to work, you will have to port BIND 4.9. 1489 The vendor's BIND is reported to be ``totally inadequate.'' 1490 See sendmail/contrib/AmdahlUTS.patch for the patches necessary 1491 to get BIND 4.9 compiled for UTS. 1492 1493UnixWare 1494 According to Alexander Kolbasov <sasha@unitech.gamma.ru>, 1495 the m4 on UnixWare 2.0 (still in Beta) will core dump on the 1496 config files. GNU m4 and the m4 from UnixWare 1.x both work. 1497 1498 According to Larry Rosenman <ler@lerami.lerctr.org>: 1499 1500 UnixWare 2.1.[23]'s m4 chokes (not obviously) when 1501 processing the 8.9.0 cf files. 1502 1503 I had a LOCAL_RULE_0 that wound up AFTER the 1504 SBasic_check_rcpt rules using the SCO supplied M4. 1505 GNU M4 works fine. 1506 1507UNICOS 8.0.3.4 1508 Some people have reported that the -O flag on UNICOS can cause 1509 problems. You may want to turn this off if you have problems 1510 running sendmail. Reported by Jerry G. DeLapp <jgd@acl.lanl.gov>. 1511 1512Darwin/Mac OS X (10.X.X) 1513 The linker errors produced regarding getopt() and its associated 1514 variables can safely be ignored. 1515 1516 From Mike Zimmerman <zimmy@torrentnet.com>: 1517 1518 From scratch here is what Darwin users need to do to the standard 1519 10.0.0, 10.0.1 install to get sendmail working. 1520 From http://www.macosx.com/forums/showthread.php?s=6dac0e9e1f3fd118a4870a8a9b559491&threadid=2242: 1521 1. chmod g-w / /private /private/etc 1522 2. Properly set HOSTNAME in /etc/hostconfig to your FQDN: 1523 HOSTNAME=-my.domain.com- 1524 3. Edit /etc/rc.boot: 1525 hostname my.domain.com 1526 domainname domain.com 1527 4. Edit /System/Library/StartupItems/Sendmail/Sendmail: 1528 Remove the "&" after the sendmail command: 1529 /usr/sbin/sendmail -bd -q1h 1530 1531 From Carsten Klapp <carsten.klapp@home.com>: 1532 1533 The easiest workaround is to remove the group-writable permission 1534 for the root directory and the symbolic /etc inherits this 1535 change. While this does fix sendmail, the unfortunate side-effect 1536 is the OS X admin will no longer be able to manipulate icons in the 1537 top level of the Startup disk unless logged into the GUI as the 1538 superuser. 1539 1540 In applying the alternate workaround, care must be taken while 1541 swapping the symlink /etc with the directory /private/etc. In all 1542 likelihood any admin who is concerned with this sendmail error has 1543 enough experience to not accidentally harm anything in the process. 1544 1545 a. Swap the /etc symlink with /private/etc (as superuser): 1546 rm /etc 1547 mv /private/etc /etc 1548 ln -s /etc /private/etc 1549 1550 b. Set / to group unwritable (as superuser): 1551 chmod g-w / 1552 1553Darwin/Mac OS X (10.1.5) 1554 Apple's upgrade to sendmail 8.12 is incorrectly configured. You 1555 will need to manually fix it up by doing the following: 1556 1557 1. chown smmsp:smmsp /var/spool/clientmqueue 1558 2. chmod 2770 /var/spool/clientmqueue 1559 3. chgrp smmsp /usr/sbin/sendmail 1560 4. chmod g+s /usr/sbin/sendmail 1561 1562 From Daniel J. Luke <dluke@geeklair.net>: 1563 1564 It appears that setting the sendmail.cf property in 1565 /locations/sendmail in NetInfo on Mac OS X 10.1.5 with sendmail 1566 8.12.4 causes 'bad things' to happen. 1567 1568 Specifically sendmail instances that should be getting their config 1569 from /etc/mail/submit.cf don't (so mail/mutt/perl scripts which 1570 open pipes to sendmail stop working as sendmail tries to write to 1571 /var/spool/mqueue and cannot as sendmail is no longer suid root). 1572 1573 Removing the entry from NetInfo fixes this problem. 1574 1575GNU getopt 1576 I'm told that GNU getopt has a problem in that it gets confused 1577 by the double call. Use the version in conf.c instead. 1578 1579BIND 4.9.2 and Ultrix 1580 If you are running on Ultrix, be sure you read conf/Info.Ultrix 1581 in the BIND distribution very carefully -- there is information 1582 in there that you need to know in order to avoid errors of the 1583 form: 1584 1585 /lib/libc.a(gethostent.o): sethostent: multiply defined 1586 /lib/libc.a(gethostent.o): endhostent: multiply defined 1587 /lib/libc.a(gethostent.o): gethostbyname: multiply defined 1588 /lib/libc.a(gethostent.o): gethostbyaddr: multiply defined 1589 1590 during the link stage. 1591 1592BIND 8.X 1593 BIND 8.X returns HOST_NOT_FOUND instead of TRY_AGAIN on temporary 1594 DNS failures when trying to find the hostname associated with an IP 1595 address (gethostbyaddr()). This can cause problems as 1596 $&{client_name} based lookups in class R ($=R) and the access 1597 database won't succeed. 1598 1599 This will be fixed in BIND 8.2.1. For earlier versions, this can 1600 be fixed by making "dns" the last name service queried for host 1601 resolution in /etc/irs.conf: 1602 1603 hosts local continue 1604 hosts dns 1605 1606strtoul 1607 Some compilers (notably gcc) claim to be ANSI C but do not 1608 include the ANSI-required routine "strtoul". If your compiler 1609 has this problem, you will get an error in srvrsmtp.c on the 1610 code: 1611 1612 # ifdef defined(__STDC__) && !defined(BROKEN_ANSI_LIBRARY) 1613 e->e_msgsize = strtoul(vp, (char **) NULL, 10); 1614 # else 1615 e->e_msgsize = strtol(vp, (char **) NULL, 10); 1616 # endif 1617 1618 You can use -DBROKEN_ANSI_LIBRARY to get around this problem. 1619 1620Listproc 6.0c 1621 Date: 23 Sep 1995 23:56:07 GMT 1622 Message-ID: <95925101334.~INN-AUMa00187.comp-news@dl.ac.uk> 1623 From: alansz@mellers1.psych.berkeley.edu (Alan Schwartz) 1624 Subject: Listproc 6.0c + Sendmail 8.7 [Helpful hint] 1625 1626 Just upgraded to sendmail 8.7, and discovered that listproc 6.0c 1627 breaks, because it, by default, sends a blank "HELO" rather than 1628 a "HELO hostname" when using the 'system' or 'telnet' mail method. 1629 1630 The fix is to include -DZMAILER in the compilation, which will 1631 cause it to use "HELO hostname" (which Z-mail apparently requires 1632 as well. :) 1633 1634OpenSSL 1635 OpenSSL versions prior to 0.9.6 use a macro named Free which 1636 conflicts with existing macro names on some platforms, such as 1637 AIX. 1638 Do not use 0.9.3, but OpenSSL 0.9.5a or later if compatible with 1639 0.9.5a. 1640 1641PH 1642 PH support is provided by Mark Roth <roth@uiuc.edu>. The map is 1643 described at http://www-dev.cso.uiuc.edu/sendmail/ . 1644 1645 NOTE: The "spacedname" pseudo-field which was used by earlier 1646 versions of the PH map code is no longer supported! See the URL 1647 listed above for more information. 1648 1649 Please contact Mark Roth for support and questions regarding the 1650 map. 1651 1652TCP Wrappers 1653 If you are using -DTCPWRAPPERS to get TCP Wrappers support you will 1654 also need to install libwrap.a and modify your site.config.m4 file 1655 or the generated Makefile to include -lwrap in the LIBS line 1656 (make sure that INCDIRS and LIBDIRS point to where the tcpd.h and 1657 libwrap.a can be found). 1658 1659 TCP Wrappers is available at ftp://ftp.porcupine.org/pub/security/. 1660 1661 If you have alternate MX sites for your site, be sure that all of 1662 your MX sites reject the same set of hosts. If not, a bad guy whom 1663 you reject will connect to your site, fail, and move on to the next 1664 MX site, which will accept the mail for you and forward it on to you. 1665 1666Regular Expressions (MAP_REGEX) 1667 If sendmail linking fails with: 1668 1669 undefined reference to 'regcomp' 1670 1671 or sendmail gives an error about a regular expression with: 1672 1673 pattern-compile-error: : Operation not applicable 1674 1675 Your libc does not include a running version of POSIX-regex. Use 1676 librx or regex.o from the GNU Free Software Foundation, 1677 ftp://ftp.gnu.org/pub/gnu/rx-?.?.tar.gz or 1678 ftp://ftp.gnu.org/pub/gnu/regex-?.?.tar.gz. 1679 You can also use the regex-lib by Henry Spencer, 1680 ftp://ftp.funet.fi/pub/languages/C/spencer/regex.shar.gz 1681 Make sure, your compiler reads regex.h from the distribution, 1682 not from /usr/include, otherwise sendmail will dump a core. 1683 1684 1685+--------------+ 1686| MANUAL PAGES | 1687+--------------+ 1688 1689The manual pages have been written against the -man macros, and 1690should format correctly with any reasonable *roff. 1691 1692 1693+-----------------+ 1694| DEBUGGING HOOKS | 1695+-----------------+ 1696 1697As of 8.6.5, sendmail daemons will catch a SIGUSR1 signal and log 1698some debugging output (logged at LOG_DEBUG severity). The 1699information dumped is: 1700 1701 * The value of the $j macro. 1702 * A warning if $j is not in the set $=w. 1703 * A list of the open file descriptors. 1704 * The contents of the connection cache. 1705 * If ruleset 89 is defined, it is evaluated and the results printed. 1706 1707This allows you to get information regarding the runtime state of the 1708daemon on the fly. This should not be done too frequently, since 1709the process of rewriting may lose memory which will not be recovered. 1710Also, ruleset 89 may call non-reentrant routines, so there is a small 1711non-zero probability that this will cause other problems. It is 1712really only for debugging serious problems. 1713 1714A typical formulation of ruleset 89 would be: 1715 1716 R$* $@ $>0 some test address 1717 1718 1719+-----------------------------+ 1720| DESCRIPTION OF SOURCE FILES | 1721+-----------------------------+ 1722 1723The following list describes the files in this directory: 1724 1725Build Shell script for building sendmail. 1726Makefile A convenience for calling ./Build. 1727Makefile.m4 A template for constructing a makefile based on the 1728 information in the devtools directory. 1729README This file. 1730TRACEFLAGS My own personal list of the trace flags -- not guaranteed 1731 to be particularly up to date. 1732alias.c Does name aliasing in all forms. 1733aliases.5 Man page describing the format of the aliases file. 1734arpadate.c A subroutine which creates ARPANET standard dates. 1735bf.c Routines to implement memory-buffered file system using 1736 hooks provided by libsm now (formerly Torek stdio library). 1737bf.h Buffered file I/O function declarations and 1738 data structure and function declarations for bf.c. 1739collect.c The routine that actually reads the mail into a temp 1740 file. It also does a certain amount of parsing of 1741 the header, etc. 1742conf.c The configuration file. This contains information 1743 that is presumed to be quite static and non- 1744 controversial, or code compiled in for efficiency 1745 reasons. Most of the configuration is in sendmail.cf. 1746conf.h Configuration that must be known everywhere. 1747control.c Routines to implement control socket. 1748convtime.c A routine to sanely process times. 1749daemon.c Routines to implement daemon mode. 1750deliver.c Routines to deliver mail. 1751domain.c Routines that interface with DNS (the Domain Name 1752 System). 1753envelope.c Routines to manipulate the envelope structure. 1754err.c Routines to print error messages. 1755headers.c Routines to process message headers. 1756helpfile An example helpfile for the SMTP HELP command and -bt mode. 1757macro.c The macro expander. This is used internally to 1758 insert information from the configuration file. 1759mailq.1 Man page for the mailq command. 1760main.c The main routine to sendmail. This file also 1761 contains some miscellaneous routines. 1762makesendmail A convenience for calling ./Build. 1763map.c Support for database maps. 1764mci.c Routines that handle mail connection information caching. 1765milter.c MTA portions of the mail filter API. 1766mime.c MIME conversion routines. 1767newaliases.1 Man page for the newaliases command. 1768parseaddr.c The routines which do address parsing. 1769queue.c Routines to implement message queueing. 1770readcf.c The routine that reads the configuration file and 1771 translates it to internal form. 1772recipient.c Routines that manipulate the recipient list. 1773sasl.c Routines to interact with Cyrys-SASL. 1774savemail.c Routines which save the letter on processing errors. 1775sendmail.8 Man page for the sendmail command. 1776sendmail.h Main header file for sendmail. 1777sfsasl.c I/O interface between SASL/TLS and the MTA. 1778sfsasl.h Header file for sfsasl.c. 1779shmticklib.c Routines for shared memory counters. 1780sm_resolve.c Routines for DNS lookups (for DNS map type). 1781sm_resolve.h Header file for sm_resolve.c. 1782srvrsmtp.c Routines to implement server SMTP. 1783stab.c Routines to manage the symbol table. 1784stats.c Routines to collect and post the statistics. 1785statusd_shm.h Data structure and function declarations for shmticklib.c. 1786sysexits.c List of error messages associated with error codes 1787 in sysexits.h. 1788sysexits.h List of error codes for systems that lack their own. 1789timers.c Routines to provide microtimers. 1790timers.h Data structure and function declarations for timers.h. 1791tls.c Routines for TLS. 1792trace.c The trace package. These routines allow setting and 1793 testing of trace flags with a high granularity. 1794udb.c The user database interface module. 1795usersmtp.c Routines to implement user SMTP. 1796util.c Some general purpose routines used by sendmail. 1797version.c The version number and information about this 1798 version of sendmail. 1799 1800(Version $Revision: 8.355.2.14 $, last update $Date: 2003/03/27 21:39:29 $ ) 1801