1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.618 2010/06/05 10:04:43 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
8 PP/01 Date: & Message-Id: revert to normally being appended to a message,
9 only prepend for the Resent-* case. Fixes regression introduced in
10 Exim 4.70 by NM/22 for Bugzilla 607.
12 PP/02 Include check_rfc2047_length in configure.default because we're seeing
13 increasing numbers of administrators be bitten by this.
15 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
17 PP/03 Bugzilla 994: added openssl_options main configuration option.
19 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
21 PP/05 Bugzilla 834: provide a permit_codedump option for pipe transports.
26 JJ/01 installed exipick 20100104.1, adding $max_received_linelength, $data_path, and $header_path variables; fixed documentation bugs and typos
28 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow exipick to access non-standard spools, including the "frozen" queue (Finput)
30 NM/01 Bugzilla 965: Support mysql stored procedures.
31 Patch from Alain Williams
33 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
35 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
36 Patch from Andreas Metzler
38 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
39 Patch from Kirill Miazine
41 NM/05 Bugzilla 671: Added umask to procmail example.
43 JJ/03 installed exipick 20100323.0, fixing doc bug
45 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
46 directory. Notification and patch from Dan Rosenberg.
48 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
50 TK/02 Improve log output when DKIM signing operation fails.
52 MH/01 Treat the transport option dkim_domain as a colon separated
53 list, not as a single string, and sign the message with each element,
54 omitting multiple occurences of the same signer.
56 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
57 Bugzilla 985, 986. Patch by Simon Arlott
59 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
62 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
63 MBX locking. Notification from Dan Rosenberg.
69 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
71 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
73 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
75 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
77 NM/04 Bugzilla 915: Flag broken perl installation during build.
83 TK/01 Added patch by Johannes Berg that expands the main option
84 "spamd_address" if it starts with a dollar sign.
86 TK/02 Write list of recipients to X-Envelope-Sender header when building
87 the mbox-format spool file for content scanning (suggested by Jakob
90 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
91 (http://www.dcc-servers.net/) support via dccifd. Activated by
92 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
94 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
95 by Mark Daniel Reidel <mr@df.eu>.
97 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
98 When building exim an external PCRE library is now needed -
99 PCRE is a system library on the majority of modern systems.
100 See entry on PCRE_LIBS in EDITME file.
102 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
103 conversation. Added nologin parameter to request.
104 Patch contributed by Kirill Miazine
106 TF/01 Do not log submission mode rewrites if they do not change the address.
108 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
110 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
111 log files in place. Contributed by Roberto Lima
113 NM/04 Bugzilla 667: close socket used by dovecot authenticator
115 TF/03 Bugzilla 615: When checking the local_parts router precondition
116 after a local_part_suffix or local_part_prefix option, Exim now
117 does not use the address's named list lookup cache, since this
118 contains cached lookups for the whole local part.
120 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
121 Robert Millan. Documentation is in experimental-spec.txt
123 TF/04 Bugzilla 668: Fix parallel build (make -j).
125 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
127 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
128 Patch provided by Jan Srzednicki
130 TF/05 Leading white space used to be stripped from $spam_report which
131 wrecked the formatting. Now it is preserved.
133 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
134 that they are available at delivery time.
136 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
138 TF/08 TLS error reporting now respects the incoming_interface and
139 incoming_port log selectors.
141 TF/09 Produce a more useful error message if an SMTP transport's hosts
142 setting expands to an empty string.
144 NM/06 Bugzilla 744: EXPN did not work under TLS.
145 Patch provided by Phil Pennock
147 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
148 Patch provided by Richard Godbee
150 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
151 acl_smtp_notquit, added index entry.
153 NM/09 Bugzilla 787: Potential buffer overflow in string_format
154 Patch provided by Eugene Bujak
156 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
157 Patch provided by Maxim Dounin
159 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
160 Patch provided by Phil Pennock
162 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
164 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
165 Patch provided by Brad "anomie" Jorsch
167 NM/14 Bugzilla 622: Added timeout setting to transport filter.
168 Patch provided by Dean Brooks
170 TK/05 Add native DKIM support (does not depend on external libraries).
172 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
173 Patch provided by Graeme Fowler
175 NM/16 Bugzilla 851: Documentation example syntax fix.
177 NM/17 Changed NOTICE file to remove references to embedded PCRE.
179 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
181 NM/19 Bugzilla 745: TLS version reporting
182 Patch provided by Phil Pennock
184 NM/20 Bugzilla 167: bool: condition support
185 Patch provided by Phil Pennock
187 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
188 Patch provided by Phil Pennock
190 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
191 Patch provided by Brad "anomie" Jorsch
193 NM/23 Bugzilla 687: Fix misparses in eximstats
194 Patch provided by Heiko Schlittermann
196 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
197 Patch provided by Heiko Schlittermann
199 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
200 plus update to original patch.
202 NM/26 Bugzilla 799: Documentation correction for ratelimit
204 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
205 Patch provided by David Brownlee
207 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
209 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
211 NM/30 Bugzilla 888: TLS documentation bugfixes
213 NM/31 Bugzilla 896: Dovecot buffer overrun fix
215 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
216 Unlike the original bugzilla I have changed all shell scripts in src tree
218 NM/33 Bugzilla 898: Transport filter timeout fix
219 Patch by Todd Rinaldo
221 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
222 Patch by Serge Demonchaux
224 NM/35 Bugzilla 39: Base64 decode bug fixes
225 Patch by Jakob Hirsch
227 NM/36 Bugzilla 909: Correct connect() call in dcc code
229 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
231 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
233 NM/39 Bugzilla 911: Fixed MakeLinks build script
239 TK/01 Add preliminary DKIM support. Currently requires a forked version of
240 ALT-N's libdkim that I have put here:
241 http://duncanthrax.net/exim-experimental/
243 Note to Michael Haardt: I had to rename some vars in sieve.c. They
244 were called 'true' and it seems that C99 defines that as a reserved
245 keyword to be used with 'bool' variable types. That means you could
246 not include C99-style headers which use bools without triggering
247 build errors in sieve.c.
249 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
250 as mailq or other aliases. Changed the --help handling significantly
251 to do whats expected. exim_usage() emits usage/help information.
253 SC/01 Added the -bylocaldomain option to eximstats.
255 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
257 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
259 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
265 PH/01 Another patch from the Sieve maintainer.
267 PH/02 When an IPv6 address is converted to a string for single-key lookup
268 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
269 dots are used instead of colons so that keys in lsearch files need not
270 contain colons. This was done some time before quoting was made available
271 in lsearch files. However, iplsearch files do require colons in IPv6 keys
272 (notated using the quote facility) so as to distinguish them from IPv4
273 keys. This meant that lookups for IP addresses in host lists did not work
274 for iplsearch lookups.
276 This has been fixed by arranging for IPv6 addresses to be expressed with
277 colons if the lookup type is iplsearch. This is not incompatible, because
278 previously such lookups could never work.
280 The situation is now rather anomolous, since one *can* have colons in
281 ordinary lsearch keys. However, making the change in all cases is
282 incompatible and would probably break a number of configurations.
284 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
287 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
288 conversion specification without a maximum field width, thereby enabling
289 a rogue spamd server to cause a buffer overflow. While nobody in their
290 right mind would setup Exim to query an untrusted spamd server, an
291 attacker that gains access to a server running spamd could potentially
292 exploit this vulnerability to run arbitrary code as the Exim user.
294 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
295 $primary_hostname instead of what libspf2 thinks the hosts name is.
297 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
298 a directory entry by the name of the lookup key. Previously, if a
299 symlink pointed to a non-existing file or a file in a directory that
300 Exim lacked permissions to read, a lookup for a key matching that
301 symlink would fail. Now it is enough that a matching directory entry
302 exists, symlink or not. (Bugzilla 503.)
304 PH/03 The body_linecount and body_zerocount variables are now exported in the
307 PH/04 Added the $dnslist_matched variable.
309 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
310 This means they are set thereafter only if the connection becomes
313 PH/06 Added the client_condition to authenticators so that some can be skipped
314 by clients under certain conditions.
316 PH/07 The error message for a badly-placed control=no_multiline_responses left
317 "_responses" off the end of the name.
319 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
321 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
322 (without spaces) instead of just copying the configuration text.
324 PH/10 Added the /noupdate option to the ratelimit ACL condition.
326 PH/11 Added $max_received_linelength.
328 PH/12 Added +ignore_defer and +include_defer to host lists.
330 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
331 way in which PCRE > 7.0 is built.
333 PH/14 Implemented queue_only_load_latch.
335 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
336 MAIL command. The effect was to mangle the value on 64-bit systems.
338 PH/16 Another patch from the Sieve maintainer.
340 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
342 PH/18 If a system quota error occurred while trying to create the file for
343 a maildir delivery, the message "Mailbox is full" was not appended to the
344 bounce if the delivery eventually timed out. Change 4.67/27 below applied
345 only to a quota excession during the actual writing of the file.
347 PH/19 It seems that peer DN values may contain newlines (and other non-printing
348 characters?) which causes problems in log lines. The DN values are now
349 passed through string_printing() before being added to log lines.
351 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
352 and InterBase are left for another time.)
354 PH/21 Added message_body_newlines option.
356 PH/22 Guard against possible overflow in moan_check_errorcopy().
358 PH/23 POSIX allows open() to be a macro; guard against that.
360 PH/24 If the recipient of an error message contained an @ in the local part
361 (suitably quoted, of course), incorrect values were put in $domain and
362 $local_part during the evaluation of errors_copy.
368 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
369 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
372 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
373 issue a MAIL command.
375 PH/02 In an ACL statement such as
377 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
379 if a client was not listed at all, or was listed with a value other than
380 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
381 the condition was not true (as it should be), so access was not denied.
382 The bug was that the ! inversion was incorrectly passed on to the second
383 item. This has been fixed.
385 PH/03 Added additional dnslists conditions == and =& which are different from
386 = and & when the dns lookup returns more than one IP address.
388 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
389 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
391 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
392 FSYNC, which compiles an option called disable_fsync that allows for
393 bypassing fsync(). The documentation is heavily laced with warnings.
395 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
397 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
398 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
399 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
400 including adding "make clean"; (3) Added -fPIC when compiling the test
401 dynamically loaded module, to get rid of a warning.
403 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
404 message fails, move_frozen_messages = true and ignore_bounce_errors_after
405 = 0s. The bug is otherwise harmless.
407 PH/07 There was a bug in the dovecot authenticator such that the value of
408 $auth1 could be overwritten, and so not correctly preserved, after a
409 successful authentication. This usually meant that the value preserved by
410 the server_setid option was incorrect.
412 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
414 PH/09 Installed PCRE release 7.0.
416 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
417 run for batched SMTP input. It is now run at the start of every message
418 in the batch. While fixing this I discovered that the process information
419 (output by running exiwhat) was not always getting set for -bs and -bS
420 input. This is fixed, and it now also says "batched" for BSMTP.
422 PH/11 Added control=no_pipelining.
424 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
425 patch, slightly modified), and move the expansion of helo_data till after
426 the connection is made in the smtp transport (so it can use these
429 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
431 PH/14 Added log_selector = +pid.
433 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
435 PH/16 Add ${if forany and ${if forall.
437 PH/17 Added dsn_from option to vary the From: line in DSNs.
439 PH/18 Flush SMTP output before performing a callout, unless control =
440 no_callout_flush is set.
442 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
443 was true (the default) a successful delivery failed to delete the retry
444 item, thus causing premature timeout of the address. The bug is now
447 PH/20 Added hosts_avoid_pipelining to the smtp transport.
449 PH/21 Long custom messages for fakedefer and fakereject are now split up
450 into multiline reponses in the same way that messages for "deny" and
451 other ACL rejections are.
453 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
454 with slight modification.
456 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
457 draft, changing the syntax and factoring some duplicate code".
459 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
460 for deliveries of the second and subsequent messages over the same SMTP
463 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
464 ${reduce, with only minor "tidies".
466 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
468 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
469 expansion side effects.
471 PH/27 When a message times out after an over-quota error from an Exim-imposed
472 quota, the bounce message says "mailbox is full". This message was not
473 being given when it was a system quota that was exceeded. It now should
476 MH/03 Made $recipients available in local_scan(). local_scan() already has
477 better access to the recipient list through recipients_list[], but
478 $recipients can be useful in postmaster-provided expansion strings.
480 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
481 in the case of a MAIL command with additional options following the
482 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
483 were accidentally chopped off.
485 PH/29 SMTP synchronization checks are implemented when a command is read -
486 there is a check that no more input is waiting when there shouldn't be
487 any. However, for some commands, a delay in an ACL can mean that it is
488 some time before the response is written. In this time, more input might
489 arrive, invalidly. So now there are extra checks after an ACL has run for
490 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
491 pipelining has not been advertised.
493 PH/30 MH's patch to allow iscntrl() characters to be list separators.
495 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
496 returned in the SMTP response when smtp_return_error_details was false.
499 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
500 instead of the C I/O that was originally supplied, because problems were
503 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
504 Exim which did not show up earlier: it was assuming that a call to
505 SSL_CTX_set_info_callback() might give an error value. In fact, there is
506 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
507 was a macro that became an assignment, so it seemed to work. This has
508 changed to a proper function call with a void return, hence the compile
509 error. Exim's code has been fixed.
511 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
514 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
516 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
518 PH/37 If a message is not accepted after it has had an id assigned (e.g.
519 because it turns out to be too big or there is a timeout) there is no
520 "Completed" line in the log. When some messages of this type were
521 selected by exigrep, they were listed as "not completed". Others were
522 picked up by some special patterns. I have improved the selection
523 criteria to be more general.
525 PH/38 The host_find_failed option in the manualroute router can now be set
526 to "ignore", to completely ignore a host whose IP address cannot be
527 found. If all hosts are ignored, the behaviour is controlled by the new
528 host_all_ignored option.
530 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
531 homing or because of multiple MX records with /mx) generated more than
532 one IP address, and the following item turned out to be the local host,
533 all the secondary addresses of the first item were incorrectly removed
534 from the list, along with the local host and any following hosts (which
535 is what is supposed to happen).
537 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
538 whoever called Exim into the -H file. In the case of the daemon it was
539 behaving confusingly. When first started, it used values for whoever
540 started the daemon, but after a SIGHUP it used the Exim user (because it
541 calls itself on a restart). I have changed the code so that it now always
544 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
545 message are rejected with the same error (e.g. no authentication or bad
546 sender address), and a DATA command is nevertheless sent (as can happen
547 with PIPELINING or a stupid MUA), the error message that was given to the
548 RCPT commands is included in the rejection of the DATA command. This is
549 intended to be helpful for MUAs that show only the final error to their
552 PH/42 Another patch from the Sieve maintainer.
554 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
555 Eximstats - Fixed some broken HTML links and added missing column headers
557 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
558 columns for Rejects, Temp Rejects, Ham, and Spam rows.
560 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
562 PH/43 Yet another patch from the Sieve maintainer.
564 PH/44 I found a way to check for a TCP/IP connection going away before sending
565 the response to the final '.' that terminates a message, but only in the
566 case where the client has not sent further data following the '.'
567 (unfortunately, this is allowed). However, in many cases there won't be
568 any further data because there won't be any more messages to send. A call
569 to select() can be used: if it shows that the input is "ready", there is
570 either input waiting, or the socket has been closed. An attempt to read
571 the next input character can distinguish the two cases. Previously, Exim
572 would have sent an OK response which the client would never have see.
573 This could lead to message repetition. This fix should cure that, at
574 least in a lot of common cases.
576 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
577 advertised in response to EHLO.
583 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
584 fixed by 4.65/MH/01 (is this a record?) are fixed:
586 (i) An empty string was always treated as zero by the numeric comparison
587 operators. This behaviour has been restored.
589 (ii) It is documented that the numeric comparison operators always treat
590 their arguments as decimal numbers. This was broken in that numbers
591 starting with 0 were being interpreted as octal.
593 While fixing these problems I realized that there was another issue that
594 hadn't been noticed. Values of message_size_limit (both the global option
595 and the transport option) were treated as octal if they started with 0.
596 The documentation was vague. These values are now always treated as
597 decimal, and I will make that clear in the documentation.
603 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
604 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
607 MH/01 Don't check that the operands of numeric comparison operators are
608 integers when their expansion is in "skipping" mode (fixes bug
609 introduced by 4.64-PH/07).
611 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
612 child addresses, Exim now panics and dies. Previously, because the count
613 is held in a short int, deliveries were likely to be lost. As such a
614 large number of recipients for a single message is ridiculous
615 (performance will be very, very poor), I have chosen to impose a limit
616 rather than extend the field.
622 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
623 leftover -K file (the existence of which was triggered by #402).
624 While we were at it, introduced process PID as part of the -K
625 filename. This should rule out race conditions when creating
628 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
629 processing considerably. Previous code took too long for large mails,
630 triggering a timeout which in turn triggers #401.
632 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
633 in the DK code in transports.c. sendfile() is not really portable,
634 hence the _LINUX specificness.
636 TF/01 In the add_headers option to the mail command in an Exim filter,
637 there was a bug that Exim would claim a syntax error in any
638 header after the first one which had an odd number of characters
641 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
642 callout verification, Exim cached a "reject" for the entire domain. This
643 is correct for most verifications, but it is not correct for a recipient
644 verification with use_sender or use_postmaster set, because in that case
645 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
646 case of MAIL FROM:<> rejection from other early rejections (e.g.
647 rejection of HELO). When verifying a recipient using a non-null MAIL
648 address, the cache is ignored if it shows MAIL FROM:<> rejection.
649 Whatever the result of the callout, the value of the domain cache is
650 left unchanged (for any other kind of callout, getting as far as trying
651 RCPT means that the domain itself is ok).
653 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
656 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
657 manifest itself as EPIPE rather than ECONNECT. When tidying away a
658 session, the daemon ignores ECONNECT errors and logs others; it now
659 ignores EPIPE as well.
661 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
662 (quoted-printable decoding).
664 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
665 later the small subsequent patch to fix an introduced bug.
667 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
669 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
671 PH/08 An error is now given if message_size_limit is specified negative.
673 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
674 to be given (somewhat) arbitrary names.
676 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
679 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
680 miscellaneous code fixes
682 PH/10 Added the log_reject_target ACL modifier to specify where to log
685 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
686 hostname. This is wrong, because it relates to the incoming message (and
687 probably the interface on which it is arriving) and not to the outgoing
688 callout (which could be using a different interface). This has been
689 changed to use the value of the helo_data option from the smtp transport
690 instead - this is what is used when a message is actually being sent. If
691 there is no remote transport (possible with a router that sets up host
692 addresses), $smtp_active_hostname is used.
694 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
695 tweaks were necessary in order to get it to work (see also 21 below):
696 (a) The code assumed that strncpy() returns a negative number on buffer
697 overflow, which isn't the case. Replaced with Exim's string_format()
699 (b) There were several signed/unsigned issues. I just did the minimum
700 hacking in of casts. There is scope for a larger refactoring.
701 (c) The code used strcasecmp() which is not a standard C function.
702 Replaced with Exim's strcmpic() function.
703 (d) The code set only $1; it now sets $auth1 as well.
704 (e) A simple test gave the error "authentication client didn't specify
705 service in request". It would seem that Dovecot has changed its
706 interface. Fortunately there's a specification; I followed it and
707 changed what the client sends and it appears to be working now.
709 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
712 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
713 address A is aliased to B and C, where B exists and C does not. Without
714 -v the output is "A verified" because verification stops after a
715 successful redirection if more than one address is generated. However,
716 with -v the child addresses are also verified. Exim was outputting "A
717 failed to verify" and then showing the successful verification for C,
718 with its parentage. It now outputs "B failed to verify", showing B's
719 parentage before showing the successful verification of C.
721 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
722 look up a TXT record in a specific list after matching in a combined
725 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
726 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
727 they consult the DNS. I had assumed they would set it the way they
728 wanted; and indeed my experiments on Linux seem to show that in some
729 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
730 To be on the safe side, however, I have now made the interface to
731 host_find_byname() similar to host_find_bydns(), with an argument
732 containing the DNS resolver options. The host_find_byname() function now
733 sets these options at its start, just as host_find_bydns() does. The smtp
734 transport options dns_qualify_single and dns_search_parents are passed to
735 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
736 of host_find_byname() use the default settings of RES_DEFNAMES
737 (qualify_single) but not RES_DNSRCH (search_parents).
739 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
740 spool_read_header() do less string testing, by means of a preliminary
741 switch on the second character of optional "-foo" lines. (This is
742 overdue, caused by the large number of possibilities that now exist.
743 Originally there were few.) While I was there, I also converted the
744 str(n)cmp tests so they don't re-test the leading "-" and the first
745 character, in the hope this might squeeze out yet more improvement.
747 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
748 flag allowing group syntax was set by the header_syntax check but not
749 turned off, possible causing trouble later; (2) The flag was not being
750 set at all for the header_verify test, causing "group"-style headers to
751 be rejected. I have now set it in this case, and also caused header_
752 verify to ignore an empty address taken from a group. While doing this, I
753 came across some other cases where the code for allowing group syntax
754 while scanning a header line wasn't quite right (mostly, not resetting
755 the flag correctly in the right place). These bugs could have caused
756 trouble for malformed header lines. I hope it is now all correct.
758 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
759 with the "reply" argument non-NULL. The code, however (which originally
760 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
761 but it didn't always do it. This confused somebody who was copying the
762 code for some other use. I have removed all the tests.
764 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
765 feature that was used to support insecure browsers during the U.S. crypto
766 embargo. It requires special client support, and Exim is probably the
767 only MTA that supported it -- and would never use it because real RSA is
768 always available. This code has been removed, because it had the bad
769 effect of slowing Exim down by computing (never used) parameters for the
770 RSA_EXPORT functionality.
772 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
773 authenticator to fail if there's a tab character in the incoming data
774 (there should never be unless someone is messing about, as it's supposed
775 to be base64-encoded). Also added, on Timo's advice, the "secured" option
776 if the connection is using TLS or if the remote IP is the same as the
777 local IP, and the "valid-client-cert option" if a client certificate has
780 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
781 authenticators. This can be used for authorization after authentication
782 succeeds. (In the case of plaintext, it servers for both authentication
785 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
786 if any retry times were supplied.
788 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
789 connection, where there is no client IP address to check. In this
790 situation, the verify now always succeeds.
792 PH/25 Applied John Jetmore's -Mset patch.
794 PH/26 Added -bem to be like -Mset, but loading a message from a file.
796 PH/27 In a string expansion for a processed (not raw) header when multiple
797 headers of the same name were present, leading whitespace was being
798 removed from all of them, but trailing whitespace was being removed only
799 from the last one. Now trailing whitespace is removed from each header
800 before concatenation. Completely empty headers in a concatenation (as
803 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
804 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
806 PH/29 [Removed. This was a change that I later backed out, and forgot to
807 correct the ChangeLog entry (that I had efficiently created) before
808 committing the later change.]
810 PH/30 Exim was sometimes attempting to deliver messages that had suffered
811 address errors (4xx response to RCPT) over the same connection as other
812 messages routed to the same hosts. Such deliveries are always "forced",
813 so retry times are not inspected. This resulted in far too many retries
814 for the affected addresses. The effect occurred only when there were more
815 hosts than the hosts_max_try setting in the smtp transport when it had
816 the 4xx errors. Those hosts that it had tried were not added to the list
817 of hosts for which the message was waiting, so if all were tried, there
818 was no problem. Two fixes have been applied:
820 (i) If there are any address or message errors in an SMTP delivery, none
821 of the hosts (tried or untried) are now added to the list of hosts
822 for which the message is waiting, so the message should not be a
823 candidate for sending over the same connection that was used for a
824 successful delivery of some other message. This seems entirely
825 reasonable: after all the message is NOT "waiting for some host".
826 This is so "obvious" that I'm not sure why it wasn't done
827 previously. Hope I haven't missed anything, but it can't do any
828 harm, as the worst effect is to miss an optimization.
830 (ii) If, despite (i), such a delivery is accidentally attempted, the
831 routing retry time is respected, so at least it doesn't keep
832 hammering the server.
834 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
835 in ${readsocket because some servers need this prod.
837 PH/32 Added some extra debug output when updating a wait-xxx database.
839 PH/33 The hint "could be header name not terminated by colon", which has been
840 given for certain expansion errors for a long time, was not being given
841 for the ${if def:h_colon_omitted{... case.
843 PH/34 The spec says: "With one important exception, whenever a domain list is
844 being scanned, $domain contains the subject domain." There was at least
845 one case where this was not true.
847 PH/35 The error "getsockname() failed: connection reset by peer" was being
848 written to the panic log as well as the main log, but it isn't really
849 panic-worthy as it just means the connection died rather early on. I have
850 removed the panic log writing for the ECONNRESET error when getsockname()
853 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
854 runs only) independently of the message's sender address. This meant
855 that, if the 4xx error was in fact related to the sender, a different
856 message to the same recipient with a different sender could confuse
857 things. In particualar, this can happen when sending to a greylisting
858 server, but other circumstances could also provoke similar problems.
859 I have changed the default so that the retry time for these errors is now
860 based a combination of the sender and recipient addresses. This change
861 can be overridden by setting address_retry_include_sender=false in the
864 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
865 remote server are returned as part of bounce messages. This was not
866 happening for LMTP over a pipe (the lmtp transport), but now it is the
867 same for both kinds of LMTP.
869 PH/38 Despite being documented as not happening, Exim was rewriting addresses
870 in header lines that were in fact CNAMEs. This is no longer the case.
872 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
873 and queue runs started by the daemon processed all messages. This has
874 been fixed so that -R and -S can now usefully be given with -q<time>.
876 PH/40 Import PCRE release 6.7 (fixes some bugs).
878 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
880 PH/42 Give an error if -q is specified more than once.
882 PH/43 Renamed the variables $interface_address and $interface_port as
883 $received_ip_address and $received_port, to make it clear that these
884 values apply to message reception, and not to the outgoing interface when
885 a message is delivered. (The old names remain recognized, of course.)
887 PH/44 There was no timeout on the connect() call when using a Unix domain
888 socket in the ${readsocket expansion. There now is.
890 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
891 be meaningful with "accept".
893 SC/01 Eximstats V1.43
894 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
896 SC/02 Eximstats V1.44
897 Use a glob alias rather than an array ref in the generated
898 parser. This improves both readability and performance.
900 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
901 Collect SpamAssassin and rejection statistics.
902 Don't display local sender or destination tables unless
903 there is data to show.
904 Added average volumes into the top table text output.
906 SC/04 Eximstats V1.46
907 Collect data on the number of addresses (recipients)
908 as well as the number of messages.
910 SC/05 Eximstats V1.47
911 Added 'Message too big' to the list of mail rejection
912 reasons (thanks to Marco Gaiarin).
914 SC/06 Eximstats V1.48
915 Mainlog lines which have GMT offsets and are too short to
916 have a flag are now skipped.
918 SC/07 Eximstats V1.49 (Alain Williams)
919 Added the -emptyok flag.
921 SC/08 Eximstats V1.50
922 Fixes for obtaining the IP address from reject messages.
924 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
925 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
926 whitesspace changes from 4.64-PH/27
928 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
931 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
932 are found, allow negative numbers in numeric criteria)
934 JJ/06 exipick.20061117.2, added new $message_body_missing variable
936 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
937 to match changes made in 4.64-PH/43
939 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
941 PH/47 Put in an explicit test for a DNS lookup of an address record where the
942 "domain" is actually an IP address, and force a failure. This locks out
943 those revolvers/nameservers that support "A-for-A" lookups, in
944 contravention of the specifications.
946 PH/48 When a host name was looked up from an IP address, and the subsequent
947 forward lookup of the name timed out, the host name was left in
948 $sender_host_name, contrary to the specification.
950 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
951 restricted to single-key lookups, Exim was not diagnosing an error if
952 * or *@ was used with a query-style lookup.
954 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
956 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
957 long ago, but noone interested enough thought of it. Let's just say that
958 the "1.1" means that there are some new functions that weren't there at
959 some point in the past.
961 PH/51 Error processing for expansion failure of helo_data from an smtp
962 transport during callout processing was broken.
964 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
965 tested/used via the -bh/-bhc/-bs options.
967 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
968 bug, fixed in subsequent PCRE releases).
970 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
971 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
973 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
979 SC/01 Use a glob alias rather than an array ref in eximstats generated
980 parser. This improves both readability and performance.
982 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
983 Don't display local sender or destination tables in eximstats unless
984 there is data to show.
985 Added average volumes into the eximstats top table text output.
987 SC/03 Collect data on the number of addresses (recipients) as well
988 as the number of messages in eximstats.
990 TF/01 Correct an error in the documentation for the redirect router. Exim
991 does (usually) call initgroups() when daemonizing.
993 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
994 with consistent privilege compared to when running as a daemon.
996 TF/03 Note in the spec that $authenticated_id is not set for local
997 submissions from trusted users.
999 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1000 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1002 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1003 by adding some example configuration directives to the default
1004 configuration file. A little bit of work is required to uncomment the
1005 directives and define how usernames and passwords are checked, but
1006 there is now a framework to start from.
1008 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1009 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1010 without this. I don't know how relevant this is to other LDAP libraries.
1012 PH/02 Add the verb name to the "unknown ACL verb" error.
1014 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1016 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1018 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1019 directory not expanded when it should be if an expanded home directory
1020 was set for the address (which is overridden by the transport).
1022 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1025 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1026 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1027 because it is too late at that time, and has no effect.
1029 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1030 security issue with \' (bugzilla #107). I could not use the
1031 PQescapeStringConn() function, because it needs a PGconn value as one of
1034 PH/08 When testing addresses using -bt, indicate those final addresses that
1035 are duplicates that would not cause an additional delivery. At least one
1036 person was confused, thinking that -bt output corresponded to deliveries.
1037 (Suppressing duplicates isn't a good idea as you lose the information
1038 about possibly different redirections that led to the duplicates.)
1040 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1041 systems where poll() doesn't work, in particular OS X.
1043 PH/10 Added more information to debugging output for retry time not reached.
1045 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1046 operations in malware.c.
1048 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1051 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1052 syslog_duplication set false, log lines that would normally be written
1053 both the the main log and to the reject log were not written to syslog at
1056 PH/14 In the default configuration, change the use of "message" in ACL warn
1057 statements to "add_header".
1059 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1060 not followed by a command (e.g. "seen endif").
1062 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1063 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1066 PH/17 Added extra conditions to the default value of delay_warning_condition
1070 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1071 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1072 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1075 The Auto-Submitted: and various List- headers are standardised, whereas I
1076 don't think Precedence: ever was.
1078 PH/18 Refactored debugging code in route_finduser() to show more information,
1079 in particular, the error code if getpwnam() issues one.
1081 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1082 This is apparently needed in addition to the PH/07 change above to avoid
1083 any possible encoding problems.
1085 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1086 but not after initializing Perl.
1088 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1089 output them only if debugging. By default they are written stderr,
1090 apparently, which is not desirable.
1092 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1095 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1098 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1100 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1101 authenticated or an ident call has been made. Suppress the default
1102 values for $authenticated_id and $authenticated_sender (but permit -oMai
1103 and -oMas) when testing with -bh.
1105 PH/24 Re-jigged the order of the tests in the default configuration so that the
1106 tests for valid domains and recipients precede the DNS black list and CSA
1107 tests, on the grounds that those ones are more expensive.
1109 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1110 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1111 This bug exists in every version of Exim that I still have, right back to
1114 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1115 However, an attempt to turn on case-sensitivity in a regex key by
1116 including (?-i) didn't work because the subject string was already
1117 lowercased, and the effects were non-intuitive. It turns out that a
1118 one-line patch can be used to allow (?-i) to work as expected.
1124 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1125 other effects) broke the use of negated acl sub-conditions.
1127 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1130 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1131 "Deny" causes Exim to reject the incoming connection with a 554 error.
1132 Unfortunately, if there is a major crisis, such as a disk failure,
1133 tcp-wrappers gives "deny", whereas what one would like would be some
1134 kind of temporary error. A kludge has been added to help with this.
1135 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1136 554 error is used if errno is still zero or contains ENOENT (which occurs
1137 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1140 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1142 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1143 errors. Otherwise a message that provokes a temporary error (when other
1144 messages do not) can cause a whole host to time out.
1146 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1147 addresses were routed directly to files or pipes from a redirect router.
1148 File deliveries just didn't batch; pipe deliveries might have suffered
1151 PH/06 A failure to get a lock for a hints database would erroneously always say
1152 "Failed to get write lock", even when it was really a read lock.
1154 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1155 of 0600. This has been changed to use the value of the lockfile_mode
1156 option (which defaults to 0600).
1158 PH/08 Applied small patch from the Sieve maintainer.
1160 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1161 folder from quota calculations, a direct delivery into this folder messed
1162 up the contents of the maildirsize file. This was because the regex was
1163 used only to exclude .Trash (or whatever) when the size of the mailbox
1164 was calculated. There was no check that a delivery was happening into an
1165 excluded directory. This bug has been fixed by ignoring all quota
1166 processing for deliveries into excluded directories.
1168 PH/10 Added the maildirfolder_create_regex option to appendfile.
1174 PH/01 The code for finding all the local interface addresses on a FreeBSD
1175 system running IPv6 was broken. This may well have applied to all BSD
1176 systems, as well as to others that have similar system calls. The broken
1177 code found IPv4 interfaces correctly, but gave incorrect values for the
1178 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1179 that it would not match correctly against @[] and not recognize the IPv6
1182 PH/02 The ipliteral router was not recognizing addresses of the form user@
1183 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1185 PH/03 Added disable_ipv6.
1187 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1188 decoded form, because it is most often used to construct To: headers
1189 lines in autoreplies, and the decoded form may well be syntactically
1190 invalid. However, $reply_address has leading white space removed, and all
1191 newlines turned into spaces so that the autoreply transport does not
1194 PH/05 If group was specified without a user on a router, and no group or user
1195 was specified on a transport, the group from the router was ignored.
1197 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1198 for visible compile-time settings that can be used to change these
1199 numbers, for those that want even more. Backwards compatibility with old
1200 spool files has been maintained. However, going back to a previous Exim
1201 release will lost any variables that are in spool files.
1203 PH/07 Two small changes when running in the test harness: increase delay when
1204 passing a TCP/IP connection to a new process, in case the original
1205 process has to generate a bounce, and remove special handling of
1206 127.0.0.2 (sic), which is no longer necessary.
1208 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1209 be the same on different OS.
1211 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1214 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1215 whether --show-vars was specified or not
1217 JJ/02 exipick: Added support for new ACL variable spool format introduced
1220 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1221 syntactically invalid From: or Reply-to: line, and a filter used this to
1222 generate an autoreply, and therefore failed to obtain an address for the
1223 autoreply, Exim could try to deliver to a non-existent relative file
1224 name, causing unrelated and misleading errors. What now happens is that
1225 it logs this as a hard delivery error, but does not attempt to create a
1228 PH/11 The exinext utility has a -C option for testing purposes, but although
1229 the given file was scanned by exinext itself; it wasn't being passed on
1230 when Exim was called.
1232 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1233 an end-of-file indication when reading a command response.
1235 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1236 compiled. In many other places in Exim, IPv6 addresses are always
1237 recognized, so I have changed this. It also means that IPv4 domain
1238 literals of the form [IPV4:n.n.n.n] are now always recognized.
1240 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1241 used if the router is not running as root, for example, when verifying at
1242 ACL time, or when using -bh. The debugging output from this situation was
1243 non-existent - all you got was a failure to exec. I have made two
1246 (a) Failures to set uid/gid, the current directory, or a process leader
1247 in a subprocess such as that created by queryprogram now generate
1248 suitable debugging ouput when -d is set.
1250 (b) The queryprogram router detects when it is not running as root,
1251 outputs suitable debugging information if -d is set, and then runs
1252 the subprocess without attempting to change uid/gid.
1254 PH/15 Minor change to Makefile for building test_host (undocumented testing
1257 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1258 additional section of a DNS packet that returns MX or SRV records.
1259 Instead, it always explicitly searches for A/AAAA records. This avoids
1260 major problems that occur when a DNS server includes only records of one
1261 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1262 fixed another bug: if SRV records were looked up and the corresponding
1263 address records were *not* found in the additional section, the port
1264 values from the SRV records were lost.
1266 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1267 using the correct key (the original address) when searching the retry
1268 rules in order to find which one to use for generating the retry hint.
1270 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1271 adding the default one. Similarly, if it contains a Reply-To: header, the
1272 errors_reply_to option, if set, is not used.
1274 PH/19 When calculating a retry time, Exim used to measure the "time since
1275 failure" by looking at the "first failed" field in the retry record. Now
1276 it does not use this if it is later than than the arrival time of the
1277 message. Instead it uses the arrival time. This makes for better
1278 behaviour in cases where some deliveries succeed, thus re-setting the
1279 "first failed" field. An example is a quota failure for a huge message
1280 when small messages continue to be delivered. Without this change, the
1281 "time since failure" will always be short, possible causing more frequent
1282 delivery attempts for the huge message than are intended.
1283 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1285 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1286 $1, $2, $3) because the numerical variables can be reset during some
1287 expansion items (e.g. "match"), thereby losing the authentication data.
1289 PH/21 Make -bV show the size of off_t variables so that the test suite can
1290 decide whether to run tests for quotas > 2G.
1292 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1293 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1294 filecount value is greater than 2G or if a quota value is greater than 2G
1295 on a system where the size of off_t is not greater than 4, a panic error
1298 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1299 never match. The debug and -bh output now contains an explicit error
1300 message indicating a malformed IPv4 address or mask.
1302 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1303 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1304 PH/23 above applies.
1306 PH/25 Do not write to syslog when running in the test harness. The only
1307 occasion when this arises is a failure to open the main or panic logs
1308 (for which there is an explicit test).
1310 PH/26 Added the /no_tell option to "control=freeze".
1312 PH/27 If a host name lookup failed very early in a connection, for example, if
1313 the IP address matched host_lookup and the reverse lookup yielded a name
1314 that did not have a forward lookup, an error message of the form "no IP
1315 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1316 could be logged. Now it outputs the IP address instead of "NULL".
1318 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1319 allows the sender and the authenticated sender to be set when
1320 submitting a message from within Exim. Since child_open_exim() is
1321 documented for local_scan(), the new function should be too.
1323 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1324 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1325 results in an empty string is now treated as unset.
1327 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1329 PH/31 Added sender_verify_fail logging option.
1331 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1332 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1333 tidied the source and removed it altogether.
1335 PH/33 When a queue run was abandoned because the load average was too high, a
1336 log line was always written; now it is written only if the queue_run log
1337 selector is set. In addition, the log line for abandonment now contains
1338 information about the queue run such as the pid. This is always present
1339 in "start" and "stop" lines but was omitted from the "abandon" line.
1341 PH/34 Omit spaces between a header name and the colon in the error message that
1342 is given when verify = headers_syntax fails (if there are lots of them,
1343 the message gets confusing).
1345 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1346 names, as there are now some PTR records that contain slashes. This check
1347 is only to protect against broken name servers that fall over on strange
1348 characters, so the fact that it applies to all lookups doesn't matter.
1350 PH/36 Now that the new test suite is complete, we can remove some of the
1351 special code in Exim that was needed for the old test suite. For example,
1352 sorting DNS records because real resolvers return them in an arbitrary
1353 order. The new test suite's fake resolver always returns records in the
1356 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1357 bounces) except when queue_only is set, to avoid logging races between
1358 the different processes.
1360 PH/38 Panic-die if .include specifies a non-absolute path.
1362 PH/39 A tweak to the "H" retry rule from its user.
1364 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1365 a label. They prevented compilation on older perls.
1367 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1368 a warning to be raised on newish perls.
1370 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1371 on queue. Changes to match documented behaviour of showing count of
1372 messages matching specified criteria.
1374 PH/40 Changed the default ident timeout from 30s to 5s.
1376 PH/41 Added support for the use of login_cap features, on those BSD systems
1377 that have them, for controlling the resources used by pipe deliveries.
1379 PH/42 The content-scanning code uses fopen() to create files in which to put
1380 message data. Previously it was not paying any attention to the mode of
1381 the files. Exim runs with umask(0) because the rest of the code creates
1382 files with open(), and sets the required mode explicitly. Thus, these
1383 files were ending up world-writeable. This was not a big issue, because,
1384 being within the spool directory, they were not world-accessible. I have
1385 created a function called modefopen, which takes an additional mode
1386 argument. It sets umask(777), creates the file, chmods it to the required
1387 mode, then resets the umask. All the relevant calls to fopen() in the
1388 content scanning code have been changed to use this function.
1390 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1391 to 24 hours. This avoids potential overflow problems when processing G
1392 and H retry rules. I suspect nobody ever tinkers with this value.
1394 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1396 PH/45 When the plaintext authenticator is running as a client, the server's
1397 challenges are checked to ensure they are valid base64 strings. By
1398 default, the authentication attempt is cancelled if an invalid string is
1399 received. Setting client_ignore_invalid_base64 true ignores these errors.
1400 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1401 they are received. Thus, the responses can be made to depend on the
1402 challenges. If an invalid string is ignored, an empty string is placed in
1405 PH/46 Messages that are created by the autoreply transport now contains a
1406 References: header, in accordance with RFCs 2822 and 3834.
1408 PH/47 Added authenticated_sender_force to the smtp transport.
1410 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1412 PH/49 Installed latest patch from the Sieve maintainer.
1414 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1415 was also set, the appendfile transport was unnecessarily scanning a
1416 directory of message files (e.g. for maildir delivery) to find the count
1417 of files (along with the size), even though it did not need this
1418 information. It now does the scan only if it needs to find either the
1419 size of the count of files.
1421 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1423 PH/52 Two bugs concerned with error handling when the smtp transport is
1426 (i) Exim was not creating retry information for temporary errors given
1427 for individual recipients after the DATA command when the smtp transport
1428 was used in LMTP mode. This meant that they could be retried too
1429 frequently, and not timed out correctly.
1431 (ii) Exim was setting the flag that allows error details to be returned
1432 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1433 recipients that were returned after the DATA command.
1435 PH/53 This is related to PH/52, but is more general: for any failing address,
1436 when detailed error information was permitted to be returned to the
1437 sender, but the error was temporary, then after the final timeout, only
1438 "retry timeout exceeded" was returned. Now it returns the full error as
1439 well as "retry timeout exceeded".
1441 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1442 do this, and (what is worse) MTAs that accept it.
1444 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1445 will now be deprecated.
1447 PH/56 New os.c-cygwin from the Cygwin maintainer.
1449 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1450 formats (previously only available in exim formats via -bpr, -bpru,
1451 and -bpra. Now also available in native and exiqgrep formats)
1453 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1454 with very large, slow to parse queues
1456 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1458 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1460 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1461 responses to authentication challenges, though it was showing the
1462 challenges; (ii) I've removed the CR characters from the debug output for
1465 PH/58 Allow for the insertion of a newline as well as a space when a string
1466 is turned into more than one encoded-word during RFC 2047 encoding. The
1467 Sieve code now uses this.
1469 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1470 data_4xx, lost_connection, tls_required.
1472 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1473 message was being sent as an SMTP response.
1475 PH/61 Add -l and -k options to exicyclog.
1477 PH/62 When verifying, if an address was redirected to one new address, so that
1478 verification continued, and the new address failed or deferred after
1479 having set something in $address_data, the value of $address_data was not
1480 passed back to the ACL. This was different to the case when no
1481 redirection occurred. The value is now passed back in both cases.
1483 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1484 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1485 use login_cap.h, so on its own it isn't the distinguishing feature. The
1486 new name refers directly to the setclassresources() function.
1488 PH/65 Added configuration files for NetBSD3.
1490 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1492 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1493 is preferred over IPv4.
1495 PH/68 The bounce_return_message and bounce_return_body options were not being
1496 honoured for bounces generated during the reception of non-SMTP messages.
1497 In particular, this applied to messages rejected by the ACL. This bug has
1498 been fixed. However, if bounce_return_message is true and bounce_return_
1499 body is false, the headers that are returned for a non-SMTP message
1500 include only those that have been read before the error was detected.
1501 (In the case of an ACL rejection, they have all been read.)
1503 PH/69 The HTML version of the specification is now built in a directory called
1504 spec_html instead of spec.html, because the latter looks like a path with
1505 a MIME-type, and this confuses some software.
1507 PH/70 Catch two compiler warnings in sieve.c.
1509 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1510 function verify_get_ident() calls ip_connect() to connect a socket, but
1511 if the "connect()" function timed out, ip_connect() used to close the
1512 socket. However, verify_get_ident() also closes the socket later, and in
1513 between Exim writes to the log, which may get opened at this point. When
1514 the socket was closed in ip_connect(), the log could get the same file
1515 descriptor number as the socket. This naturally causes chaos. The fix is
1516 not to close the socket in ip_connect(); the socket should be closed by
1517 the function that creates it. There was only one place in the code where
1518 this was missing, in the iplookup router, which I don't think anybody now
1519 uses, but I've fixed it anyway.
1521 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1522 well as to direct DNS lookups. Otherwise the handling of names in host
1523 lists is inconsistent and therefore confusing.
1529 PH/01 Two changes to the default runtime configuration:
1531 (1) Move the checks for relay_from_hosts and authenticated clients from
1532 after to before the (commented out) DNS black list checks.
1534 (2) Add control=submission to the relay_from_hosts and authenticated
1535 clients checks, on the grounds that messages accepted by these
1536 statements are most likely to be submissions.
1538 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1540 (1) Generate an error if the third argument for the ${prvs expansion is
1543 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1546 (3) Reset the variables that are obtained from the first argument of
1547 ${prvscheck and used in the second argument before leaving the code,
1548 because their memory is reclaimed, so using them afterwards may do
1551 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1552 one (it's much easier than Tom thought :-).
1554 (5) Because of (4), we can now allow for the use of $prvscheck_result
1555 inside the third argument.
1557 PH/03 For some reason, the default setting of PATH when running a command from
1558 a pipe transport was just "/usr/bin". I have changed it to
1561 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1562 anything to be listed in the output from -bV.
1564 PH/05 When a filter generated an autoreply, the entire To: header line was
1565 quoted in the delivery log line, like this:
1567 => >A.N.Other <ano@some.domain> <original@ddress> ...
1569 This has been changed so that it extracts the operative address. There
1570 may be more than one such address. If so, they are comma-separated, like
1573 => >ano@some.domain,ona@other.domain <original@ddress> ...
1575 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1576 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1577 not being looked up in the rDNS to get a host name, Exim was showing the
1578 IP address twice in Received: lines, even though the IP addresses were
1579 identical. For example:
1581 Received: from [1.2.3.4] (helo=[1.2.3.4])
1583 However, if the real host name was known, it was omitting the HELO data
1584 if it matched the actual IP address. This has been tidied up so that it
1585 doesn't show the same IP address twice.
1587 PH/07 When both +timestamp and +memory debugging was on, the value given by
1588 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1589 called by the debug printing, thereby overwriting the timestamp buffer.
1590 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1592 PH/08 When the original message was included in an autoreply transport, it
1593 always said "this is a copy of the message, including all the headers",
1594 even if body_only or headers_only was set. It now gives an appropriate
1597 PH/09 Applied a patch from the Sieve maintainer which:
1599 o fixes some comments
1600 o adds the (disabled) notify extension core
1601 o adds some debug output for the result of if/elsif tests
1602 o points to the current vacation draft in the documentation
1603 and documents the missing references header update
1607 o fixes a bug in processing the envelope test (when testing
1608 multiple envelope elements, the last element determinted the
1611 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1612 Electronic Mail") by including:
1614 Auto-submitted: auto-generated
1616 in the messages that it generates (bounce messages and others, such as
1617 warnings). In the case of bounce messages for non-SMTP mesages, there was
1618 also a typo: it was using "Auto_submitted" (underscore instead of
1619 hyphen). Since every message generated by Exim is necessarily in response
1620 to another message, thes have all been changed to:
1622 Auto-Submitted: auto-replied
1624 in accordance with these statements in the RFC:
1626 The auto-replied keyword:
1628 - SHOULD be used on messages sent in direct response to another
1629 message by an automatic process,
1631 - MUST NOT be used on manually-generated messages,
1633 - MAY be used on Delivery Status Notifications (DSNs) and Message
1634 Disposition Notifications (MDNs),
1636 - MUST NOT be used on messages generated by automatic or periodic
1637 processes, except for messages which are automatic responses to
1640 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1641 to the default Received: header definition.
1643 PH/12 Added log selector acl_warn_skipped (default on).
1645 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1646 variables because (a) they are in the wrong storage pool and (b) even if
1647 they were copied, it wouldn't work properly because of the caching.
1649 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1650 checking when decoding. Apparently there are clients that generate
1651 overlong encoded strings. Why am I not surprised?
1653 PH/15 If the first argument of "${if match_address" was not empty, but did not
1654 contain an "@" character, Exim crashed. Now it writes a panic log message
1655 and treats the condition as false.
1657 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1659 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1660 extension "envelope-auth". The code is finished and in agreement with
1661 other implementations, but there is no documentation so far and in fact,
1662 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1663 not changing the active code.
1665 Print executed "if" and "elsif" statements when debugging is used. This
1666 helps a great deal to understand what a filter does.
1668 Document more things not specified clearly in RFC3028. I had all this
1669 sorted out, when out of a sudden new issues came to my mind. Oops."
1671 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1674 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1675 canonical form (as documented). However, after a host name lookup from
1676 the IP address, check_host() was doing a simple string comparison with
1677 addresses acquired from the DNS when checking that the found name did
1678 have the original IP as one of its addresses. Since any found IPv6
1679 addresses are likely to be in abbreviated form, the comparison could
1680 fail. Luckily, there already exists a function for doing the comparison
1681 by converting both addresses to binary, so now that is used instead of
1682 the text comparison.
1684 PH/20 There was another similar case to PH/19, when a complete host name was
1685 given in a host list; looking up its IP address could give an abbreviated
1686 form, whereas the current host's name might or might not be abbreviated.
1687 The same fix has been applied.
1693 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1694 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1697 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1698 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1700 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1702 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1703 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1704 Also, for "Auto-Submitted", treat anything other than "no" as
1705 non-personal, in accordance with RFC 3834. (Previously it treated
1706 anything starting "auto-" as non-personal.)
1708 TF/01 The control=submission/name=... option had a problem with syntax
1709 errors if the name included a slash character. The /name= option
1710 now slurps the rest of the string, so it can include any characters
1711 but it must come last in the list of options (after /sender_retain
1714 PH/05 Some modifications to the interface to the fake nameserver for the new
1722 TK/01 Added the "success_on_redirect" address verification option. See
1723 NewStuff for rationale and an example.
1725 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1727 PH/02 Patch to exigrep to allow it to work on syslog lines.
1729 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1730 fread() to skip over the body file's header line, because in Cygwin the
1731 header line is locked and is inaccessible.
1733 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1734 co-exist for some time) to make it clear that it is the Exim ID that is
1735 referenced, not the Message-ID: header line.
1737 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1738 string_format() function, because snprintf() does not exist on all
1741 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1742 ${stat: expansion item.
1744 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1745 protocol synchronization error", to keep the pedants happy.
1747 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1748 well as for IRIX systems, when gcc is being used. See the host.c source
1751 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1753 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1756 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1757 transport and to the smtp transport in LMTP mode.
1759 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1761 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1762 run a filter in a subprocess. This could lead to confusion in subsequent
1763 lookups in the parent process. There should also be a search_tidyup() at
1764 the end of the subprocess.
1766 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1767 only if the host matched helo_try_verify_hosts, which caused the
1768 verification to occur when the EHLO/HELO command was issued. The ACL just
1769 tested the remembered result. Now, if a previous verification attempt has
1770 not happened, "verify = helo" does it there and then.
1772 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1774 TK/03 Fix log output including CR from clamd.
1776 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1777 exist provoked a memory error which could cause a segfault.
1779 PH/15 Installed PCRE 6.2
1781 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1783 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1784 of the problem. Specifically, suggested +O2 rather than +O1 for the
1787 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1789 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1790 record, the port was not correctly logged when the outgoing_port log
1791 selector was set (it logged the transort's default port).
1793 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1794 fallback_hosts, and "hosts" in the smtp transport.
1796 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1797 host errors such as "Connection refused".
1799 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1800 authentication with radiusclient 0.4.9:
1802 - Error returned from rc_read_config was caught wrongly
1803 - Username/password not passed on to radius server due to wrong length.
1805 The presumption is that some radiusclient API changes for 4.51/PH/17
1806 were not taken care of correctly. The code is still untested by me (my
1807 Linux distribution still has 0.3.2 of radiusclient), but it was
1808 contributed by a Radius user.
1810 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1811 expanding the "port" option of the smtp transport.
1813 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1814 while reading a MIME header. Thanks to Tom Hughes for a patch.
1816 PH/24 Include config.h inside local_scan.h so that configuration settings are
1819 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1820 that in an ACL for RCPT (for example), you can examine exactly what was
1823 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1824 commands, but it was not correctly comparing the address with the actual
1825 client host address. Thus, it would show the EHLO address in Received:
1826 header lines when this was not necessary.
1828 PH/27 Added the % operator to ${eval:}.
1830 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1831 it should be ignoring failures (because with -C, for example, it has lost
1832 privilege). It wasn't ignoring creation failures other than "already
1835 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1838 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1839 because an input line was too long, either on its own, or by virtue of
1840 too many continuations, the temporary file was not being removed, and the
1841 return code was incorrect.
1843 PH/31 Missing "BOOL" in function definition in filtertest.c.
1845 PH/32 Applied Sieve patches from the maintainer.
1847 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1849 PH/33 Added "verify = not_blind".
1851 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1852 Local/Makefile (with some defaults set). These are used in built scripts
1853 such as exicyclog, but they have never been used in the exim_install
1854 script (though there are many overriding facilities there). I have
1855 arranged that the exim_install script now takes note of these two
1858 PH/35 Installed configuration files for Dragonfly.
1860 PH/36 When a locally submitted message by a trusted user did not contain a
1861 From: header, and the sender address was obtained from -f or from an SMTP
1862 MAIL command, and the trusted user did not use -F to supply a sender
1863 name, $originator_name was incorrectly used when constructing a From:
1864 header. Furthermore, $originator_name was used for submission mode
1865 messages from external hosts without From: headers in a similar way,
1866 which is clearly wrong.
1868 PH/37 Added control=suppress_local_fixups.
1870 PH/38 When log_selector = +received_sender was set, and the addition of the
1871 sender made the log line's construction buffer exactly full, or one byte
1872 less than full, an overflow happened when the terminating "\n" was
1875 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1876 when the result of a list match is failure because a DNS lookup failed.
1878 PH/40 RM_COMMAND is now used in the building process.
1880 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1881 the "build-* directories that it finds.
1883 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1884 address in a domain literal was a prefix of an interface address.
1886 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1887 when verifying a sender address, unless rewrite_headers is false.
1889 PH/44 Wrote a long comment about why errors_to addresses are verified as
1890 recipients, not senders.
1892 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1893 the ratelimit ACL was added.
1895 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1897 PH/46 Added extra information about PostgreSQL errors to the error string.
1899 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1900 avoiding the need to install special zones in a real server. This is
1901 backwards compatible; if it can't find the fake resolver, it drops back.
1902 Thus, both old and new test suites can be run.
1904 TF/02 Added util/ratelimit.pl
1906 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1907 clock is set back in time.
1909 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1910 Candler <B.Candler@pobox.com>.
1912 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1913 OK for addresses that are the result of redirections.
1915 PH/48 A number of further additions for the benefit of the new test suite,
1916 including a fake gethostbyname() that interfaces to the fake DNS resolver
1919 TF/06 The fix for widen_domains has also been applied to qualify_single and
1920 search_parents which are the other dnslookup options that can cause
1923 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1926 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1928 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1929 service" if it encountered an MX record with an empty target hostname.
1930 The message is now "an MX or SRV record indicated no SMTP service".
1932 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1933 if the DNS of the sending site is misconfigured. This is quite a
1934 common situation. This change restores the behaviour of treating a
1935 helo verification defer as a failure.
1937 PH/51 If self=fail was set on a router, the bounce message did not include the
1938 actual error message.
1944 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1946 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1947 command itself ended in error, the underlying message about the transport
1948 filter timeout was being overwritten with the pipe command error. Now the
1949 underlying error message should be appended to the second error message.
1951 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1953 PH/02 Reduce the amount of output that "make" produces by default. Full output
1954 can still be requested.
1956 PH/03 The warning log line about a condition test deferring for a "warn" verb
1957 was being output only once per connection, rather than after each
1958 occurrence (because it was using the same function as for successful
1959 "warn" verbs). This seems wrong, so I have changed it.
1961 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1962 it should not have, which might have caused a crash in the right
1963 circumstances, but probably never did.
1965 PH/04 Installed a modified version of Tony Finch's patch to make submission
1966 mode fix the return path as well as the Sender: header line, and to
1967 add a /name= option so that you can make the user's friendly name appear
1970 TF/03 Added the control = fakedefer ACL modifier.
1972 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1973 Mark Lowes for thorough testing.
1975 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1977 TK/03 Merged latest SRS patch from Miles Wilton.
1979 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1980 with the definition in sysexits.h (which is #included earlier).
1981 Fortunately, Exim does not actually use EX_OK. The code used to try to
1982 preserve the sysexits.h value, by assumimg that macro definitions were
1983 scanned for macro replacements. I have been disabused of this notion,
1984 so now the code just undefines EX_OK before #including unistd.h.
1986 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1987 in the smtp transport. When a block could not be written in a single
1988 write() function, the timeout was being re-applied to each part-write.
1989 This seems wrong - if the receiver was accepting one byte at a time it
1990 would take for ever. The timeout is now adjusted when this happens. It
1991 doesn't have to be particularly precise.
1993 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1994 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1996 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1997 without a domain if the check to <postmaster@domain> fails.
1999 SC/01 Eximstats: added -xls and the ability to specify output files
2000 (patch written by Frank Heydlauf).
2002 SC/02 Eximstats: use FileHandles for outputing results.
2004 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2006 SC/04 Eximstats: fixed display of large numbers with -nvr option
2008 SC/05 Eximstats: fixed merging of reports with empty tables.
2010 SC/06 Eximstats: added the -include_original_destination flag
2012 SC/07 Eximstats: removed tabs and trailing whitespace.
2014 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2016 TK/06 MBOX spool code: Add real "From " MBOX separator line
2017 so the .eml file is really in mbox format (even though
2018 most programs do not really care). Patch from Alex Miller.
2020 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2021 The latter is generated from $received_to and is only set if the
2022 message has one envelope recipient. SA can use these headers,
2023 obviously out-of-the-box. Patch from Alex Miller.
2025 PH/08 The ${def test on a variable was returning false if the variable's
2026 value was "0", contrary to what the specification has always said!
2027 The result should be true unless the variable is empty.
2029 PH/09 The syntax error of a character other than { following "${if
2030 def:variable_name" (after optional whitespace) was not being diagnosed.
2031 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2032 accidental colon was present, for example, could give incorrect results.
2034 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2035 result is used (not including appendfile, where other changes are about
2038 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2039 This involved changing a lot of size variables from int to off_t. It
2040 should work with maildirs and everything.
2042 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2043 spamd dying while we are connected to it.
2045 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2048 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2049 function for BDB 4.3.
2051 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2053 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2054 the sources for PCRE have been re-arranged and more files are now
2057 PH/15 The code I had for printing potentially long long variables in PH/11
2058 above was not the best (it lost precision). The length of off_t variables
2059 is now inspected at build time, and an appropriate printing format (%ld
2060 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2061 to be "long long int" or "long int". This is needed for the internal
2062 formatting function string_vformat().
2064 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2065 the configuration file to be ":syslog", then the script "guesses" where
2066 the logs files are, rather than using the compiled in default. In our
2067 case the guess is not the same as the compiled default, so the script
2068 suddenly stopped working when I started to use syslog. The patch checks
2069 to see if log_file_path is "". If so, it attempts to read it from exim
2070 with no configuration file to get the compiled in version, before it
2071 falls back to the previous guessing code."
2073 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2074 implementing BATV in an Exim configuration. See NewStuff for the gory
2077 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2078 Makefile that are specific to HP-UX.
2080 PH/18 If the "use_postmaster" option was set for a recipient callout together
2081 with the "random" option, the postmaster address was used as the MAIL
2082 FROM address for the random test, but not for the subsequent recipient
2083 test. It is now used for both.
2085 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2086 patch removes a few documentation additions to RFC 3028, because the
2087 latest draft now contains them. It adds the new en;ascii-case comparator
2088 and a new error check for 8bit text in MIME parts. Comparator and
2089 require names are now matched exactly. I enabled the subaddress
2090 extension, but it is not well tested yet (read: it works for me)."
2092 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2093 rework some of the code of TK/09 above to avoid the hardwired use of
2094 "%lld" and "long long". Replaced the call to snprintf() with a call to
2097 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2098 records point to non-existent hosts", "retry timeout exceeded", and
2099 "retry time not reached for any host after a long failure period".
2101 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2102 experimental DomainKeys support:
2104 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2105 (2) On an error such as an illegally used "control", the wrong name for
2106 the control was given.
2108 These problems did NOT occur unless DomainKeys support was compiled.
2110 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2112 PH/24 Added ${if match_ip condition.
2114 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2115 databases so that it will be absolutely obvious if a crash occurs in the
2116 DB library. This is a regular occurrence (often caused by mis-matched
2119 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2120 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2121 picked up on a user's system that detects such things. There doesn't seem
2122 to be a gcc warning option for this - only an attribute that has to be
2123 put on the function's prototype. It seems that in Fedora Core 4 they have
2124 set this on a number of new functions. No doubt there will be more in due
2127 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2128 specify a transport. However, if an address that was verified by such a
2129 router was the subject of a callout, Exim crashed because it tried to
2130 read the rcpt_include_affixes from the non-existent transport. Now it
2131 just assumes that the setting of that option is false. This bug was
2132 introduced by 4.51/PH/31.
2134 PH/28 Changed -d+all to exclude +memory, because that information is very
2135 rarely of interest, but it makes the output a lot bigger. People tend to
2136 do -d+all out of habit.
2138 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2139 code in os-type was giving problems when libc.so lives in lib64, like on
2142 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2143 aren't the modern standard, and it seems that some systems' include files
2144 don't always have them. Exim was already checking for some of the newer
2145 ones like T_AAAA, and defining it itself. I've added checks for all the
2146 record types that Exim uses.
2148 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2149 not automatically generating a new one, as it is supposed to. This
2150 prevented TLS from working. If the file did exist, but contained invalid
2151 data, a new version was generated, as expected. It was only the case of a
2152 non-existent file that was broken.
2154 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2155 with a change in libdomainkeys > 0.64.
2157 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2158 from DNS. If the selector record carries the flag, it now has
2159 precedence over the domain-wide flag.
2161 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2163 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2164 the use of a port name that isn't defined in /etc/services) occurred, the
2165 message was deferred as in a normal delivery, and thus remained on the
2166 spool, instead of being failed because of the mua_wrapper setting. This
2167 is now fixed, and I tidied up some of the mua_wrapper messages at the
2170 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2171 the messages in a hash of arrays rather than using individual hashes.
2172 This is a bit cleaner and results in dramatic memory savings, albeit
2173 at a slight CPU cost.
2175 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2176 as requested by Marc Sherman.
2178 SC/10 Eximstats: added histograms for user specified patterns as requested
2181 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2183 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2184 fopen() in the content-scanning modules that did not already have it.
2190 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2191 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2193 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2195 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2197 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2198 to test Sieve filters that use "vacation".
2200 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2201 that changes the way the GnuTLS parameters are stored in the cache file.
2202 The new format can be generated externally. For backward compatibility,
2203 if the data in the cache doesn't make sense, Exim assumes it has read an
2204 old-format file, and it generates new data and writes a new file. This
2205 means that you can't go back to an older release without removing the
2208 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2209 work if there are any delivery delays because "one_time" forces the
2210 parent to be marked "delivered", so its unseen clone is never tried
2211 again. For this reason, Exim now forbids the simultaneous setting of
2214 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2215 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2216 entry if you want to know the details. The fix, however, neglected to
2217 consider the case when local delivery batching is involved. The test for
2218 "previously delivered" was not happening when checking to see if an
2219 address could be batched with a previous (undelivered) one; under
2220 certain circumstances this could lead to multiple deliveries to the same
2223 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2224 in its include files, and this causes problems building Exim.
2226 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2227 header_syntax) but Exim was just ignoring anything given after a slash.
2228 In particular, this caused confusion with an attempt to use "verify =
2229 reverse_host_lookup/defer_ok". An error is now given when options are
2230 supplied for verify items that do not have them. (Maybe reverse_host_
2231 lookup should have a defer_ok option, but that's a different point.)
2233 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2234 defined by RFC 821) to 2048, because there were problems with some AUTH
2235 commands, and RFC 1869 says the size should be increased for extended
2236 SMTP commands that take arguments.
2238 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2241 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2242 "unknown" error; now it says that the functionality isn't in the binary.
2244 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2245 an address' error message when a string expansion fails (syntax or
2246 whatever). Otherwise the password may appear in the log. Following change
2247 PH/42 below, there is no longer a chance of it appearing in a bounce
2250 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2252 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2253 of its addresses was ever tried. (Bugzilla bug #2.)
2255 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2256 the result incorrectly in the debug output. (It correctly added a newline
2257 to what was transported.)
2259 TF/01 Added $received_time.
2261 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2262 commented out examples of how to interface to a virus scanner and to
2263 SpamAssassin. Also added commented examples of av_scanner and
2264 spamd_address settings.
2266 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2267 and controls are allowed in which ACLs. There were a couple of minor
2268 errors. Some of the entries in the conditions table (which is a table of
2269 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2270 negation of where the condition IS allowed.
2272 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2274 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2275 header file does not have a version number, so I've had to invent a new
2276 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2277 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2278 radiusclient), but it was contributed by a Radius user.
2280 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2281 files or directories, for OpenSSL.
2283 PH/19 When an Exim process that is running as root has to create an Exim log
2284 file, it does so in a subprocess that runs as exim:exim so as to get the
2285 ownership right at creation (otherwise, other Exim processes might see
2286 the file with the wrong ownership). There was no test for failure of this
2287 fork() call, which would lead to the process getting stuck as it waited
2288 for a non-existent subprocess. Forks do occasionally fail when resources
2289 run out. I reviewed all the other calls to fork(); they all seem to check
2292 PH/20 When checking for unexpected SMTP input at connect time (before writing
2293 the banner), Exim was not dealing correctly with a non-positive return
2294 from the read() function. If the client had disconnected by this time,
2295 the result was a log entry for a synchronization error with an empty
2296 string after "input=" when read() returned zero. If read() returned -1
2297 (an event I could not check), uninitialized data bytes were printed.
2298 There were reports of junk text (parts of files, etc) appearing after
2301 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2303 PH/22 Added support for macro redefinition, and (re)definition in between
2304 driver and ACL definitions.
2306 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2307 forgetting to use the resulting value; it was using the unexpanded value.
2309 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2310 hadn't been configured. The fix is from Juergen Kreileder, who
2311 understands it better than I do:
2313 "Here's what I see happening with three configured cyrus_sasl
2314 authenticators configured (plain, login, cram-md5):
2316 On startup auth_cyrus_sasl_init() gets called for each of these.
2317 This means three calls to sasl_listmech() without a specified mech_list.
2318 => SASL tests which mechs of all available mechs actually work
2319 => three warnings about OTP not working
2320 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2322 With the patch, sasl_listmech() also gets called three times. But now
2323 SASL's mech_list option is set to the server_mech specified in the the
2324 authenticator. Or in other words, the answer from sasl_listmech()
2325 gets limited to just the mech you're testing for (which is different
2327 => the return list contains just 'plain' or 'login', 'cram-md5' or
2328 nothing depending on the value of ob->server_mech.
2330 I've just tested the patch: Authentication still works fine,
2331 unavailable mechs specified in the exim configuration are still
2332 caught, and the auth.log warnings about OTP are gone."
2334 PH/25 When debugging is enabled, the contents of the command line are added
2335 to the debugging output, even when log_selector=+arguments is not
2338 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2339 answer is "GNU", and only if the return is "GNU/something" is the answer
2342 PH/27 $acl_verify_message is now set immediately after the failure of a
2343 verification in an ACL, and so is available in subsequent modifiers. In
2344 particular, the message can be preserved by coding like this:
2346 warn !verify = sender
2347 set acl_m0 = $acl_verify_message
2349 Previously, $acl_verify_message was set only while expanding "message"
2350 and "log_message" when a very denied access.
2352 PH/28 Modified OS/os.c-Linux with
2354 -#ifndef OS_LOAD_AVERAGE
2355 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2357 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2358 nomenclature these days.)
2360 PH/29 Installed patch from the Sieve maintainer that adds the options
2361 sieve_useraddress and sieve_subaddress to the redirect router.
2363 PH/30 In these circumstances:
2364 . Two addresses routed to the same list of hosts;
2365 . First host does not offer TLS;
2366 . First host accepts first address;
2367 . First host gives temporary error to second address;
2368 . Second host offers TLS and a TLS session is established;
2369 . Second host accepts second address.
2370 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2371 and peerdn, if requested) that were in fact used only for the second
2374 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2375 attention to any local part prefix or suffix that was matched by the
2376 router that accepted the address. It now behaves in the same way as it
2377 does for delivery: the affixes are removed from the local part unless
2378 rcpt_include_affixes is set on the transport.
2380 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2381 timeout during the DATA phase of an incoming message.
2383 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2384 applied a patch sanctioned by the Sieve maintainer.
2386 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2387 the uid or gid is negative. A case of a negative gid caused this to be
2388 noticed. The fix allows for either to be negative.
2390 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2391 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2394 PH/36 The change PH/12 above was broken. Fixed it.
2396 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2397 the grounds that routing the same address twice would always produce the
2398 same answer. This might have been true once, but it is certainly no
2399 longer true now. Routing a child address may depend on the previous
2400 routing that produced that child. Some complicated redirection strategies
2401 went wrong when messages had multiple recipients, and made Exim's
2402 behaviour dependent on the order in which the addresses were given.
2404 I have moved the duplicate checking until after the routing is complete.
2405 Exim scans the addresses that are assigned to local and remote
2406 transports, and removes any duplicates. This means that more work will be
2407 done, as duplicates will always all be routed, but duplicates are
2408 presumably rare, so I don't expect this is of any significance.
2410 For deliveries to pipes, files, and autoreplies, the duplicate checking
2411 still happens during the routing process, since they are not going to be
2414 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2415 It corrects a timeout issue with spamd. This is Ian's comment: "The
2416 background is that sometimes spamd either never reads data from a
2417 connection it has accepted, or it never writes response data. The exiscan
2418 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2419 blindly assumes that writes won't block so it may never time out."
2421 PH/39 Allow G after quota size as well as K and M.
2423 PH/40 The value set for $authenticated_id in an authenticator may not contain
2424 binary zeroes or newlines because the value is written to log lines and
2425 to spool files. There was no check on this. Now the value is run through
2426 the string_printing() function so that such characters are converted to
2427 printable escape sequences.
2429 PH/41 $message_linecount is a new variable that contains the total number of
2430 lines in the message. Compare $body_linecount, which is the count for the
2433 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2434 bounce and delay warning messages, except in certain special cases, which
2437 (a) An SMTP error message from a remote host;
2438 (b) A message specified in a :fail: redirection;
2439 (c) A message specified in a "fail" command in a system filter;
2440 (d) A message specified in a FAIL return from the queryprogram router;
2441 (e) A message specified by the cannot_route_message router option.
2443 In these cases only, Exim does include the error details in bounce and
2444 warning messages. There are also a few cases where bland messages such
2445 as "unrouteable address" or "local delivery error" are given.
2447 PH/43 $value is now also set for the "else" part of a ${run expansion.
2449 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2450 being worked on, but at least Exim now implements the latest version to
2453 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2454 process to complete was treated as a delivery failure, a timeout while
2455 writing the message to the pipe was logged, but erroneously treated as a
2456 successful delivery. Such timeouts include transport filter timeouts. For
2457 consistency with the overall process timeout, these timeouts are now
2458 treated as errors, giving rise to delivery failures by default. However,
2459 there is now a new Boolean option for the pipe transport called
2460 timeout_defer, which, if set TRUE, converts the failures into defers for
2461 both kinds of timeout. A transport filter timeout is now identified in
2464 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2465 systems where "make" and "gmake" are different, calling "gmake" at top
2466 level broke things. I've arranged for the value of $(MAKE) to be passed
2467 from the Makefile to this script so that it can call the same version of
2471 A note about Exim versions 4.44 and 4.50
2472 ----------------------------------------
2474 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2475 changes of various kinds. As a consequence, a big documentation update was
2476 needed. This delayed the release for rather longer than seemed good, especially
2477 in the light of a couple of (minor) security issues. Therefore, the changes
2478 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2479 release. So 4.44 and 4.50 are in effect two different branches that both start
2482 I have left the 4.50 change log unchanged; it contains all the changes since
2483 4.43. The change log for 4.44 is below; many of its items are identical to
2484 those for 4.50. This seems to be the most sensible way to preserve the
2485 historical information.
2491 1. Minor wording change to the doc/README.SIEVE file.
2493 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2494 computation of the current number of files was incorrect.
2496 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2497 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2500 4. Give more explanation in the error message when the command for a transport
2501 filter fails to execute.
2503 5. There are several places where Exim runs a non-Exim command in a
2504 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2505 was being done only for the command run by the queryprogram router. It is
2506 now done for all such subprocesses. The other cases are: ${run, transport
2507 filters, and the commands run by the lmtp and pipe transports.
2509 6. Added CONFIGURE_GROUP build-time option.
2511 7. Some older OS have a limit of 256 on the maximum number of file
2512 descriptors. Exim was using setrlimit() to set 1000 as a large value
2513 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2514 systems. I've change it so that if it can't get 1000, it tries for 256.
2516 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2517 was an oversight, and furthermore, ever since the addition of extra
2518 controls (e.g. 4.43/32), the checks on when to allow different forms of
2519 "control" were broken. There should now be diagnostics for all cases when a
2520 control that does not make sense is encountered.
2522 9. Added the /retain_sender option to "control=submission".
2524 10. $recipients is now available in the predata ACL (oversight).
2526 11. Tidy the search cache before the fork to do a delivery from a message
2527 received from the command line. Otherwise the child will trigger a lookup
2528 failure and thereby defer the delivery if it tries to use (for example) a
2529 cached ldap connection that the parent has called unbind on.
2531 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2532 of $address_data from the recipient verification was clobbered by the
2533 sender verification.
2535 13. The value of address_data from a sender verification is now available in
2536 $sender_address_data in subsequent conditions in the ACL statement.
2538 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2540 15. Added a new option "connect=<time>" to callout options, to set a different
2543 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2544 was its contents. (It was OK if the option was not defined at all.)
2546 17. A "Completed" log line is now written for messages that are removed from
2547 the spool by the -Mrm option.
2549 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2550 information about exactly what failed.
2552 19. Added -dd to debug only the daemon process.
2554 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2555 handles timeouts, both on the server side and network timeouts. Renamed the
2556 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2558 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2559 It is now set to "smtps".
2561 22. $host_address is now set to the target address during the checking of
2562 ignore_target_hosts.
2564 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2565 being passed; this would have caused $sender_host_name to have been used if
2566 matching the list had actually called for a host name (not very likely,
2567 since this list is usually IP addresses). A host name is now passed as
2570 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2571 code that specifies a non-restarting handler (typically sigaction() in
2572 modern systems) in an attempt to fix a rare and obscure crash bug.
2574 25. Narrowed the window for a race in the daemon that could cause it to ignore
2575 SIGCHLD signals. This is not a major problem, because they are used only to
2576 wake it up if nothing else does.
2578 26. A malformed maildirsize file could cause Exim to calculate negative values
2579 for the mailbox size or file count. Odd effects could occur as a result.
2580 The maildirsize information is now recalculated if the size or filecount
2583 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2584 support for a long time. Removed HAVE_SYS_VFS_H.
2586 28. Installed the latest version of exipick from John Jetmore.
2588 29. In an address list, if the pattern was not a regular expression, an empty
2589 subject address (from a bounce message) matched only if the pattern was an
2590 empty string. Non-empty patterns were not even tested. This was the wrong
2591 because it is perfectly reasonable to use an empty address as part of a
2592 database query. An empty address is now tested by patterns that are
2593 lookups. However, all the other forms of pattern expect the subject to
2594 contain a local part and a domain, and therefore, for them, an empty
2595 address still always fails if the pattern is not itself empty.
2597 30. Exim went into a mad DNS loop when attempting to do a callout where the
2598 host was specified on an smtp transport, and looking it up yielded more
2599 than one IP address.
2601 31. Re-factored the code for checking spool and log partition space into a
2602 function that finds that data and another that does the check. The former
2603 is then used to implement four new variables: $spool_space, $log_space,
2604 $spool_inodes, and $log_inodes.
2606 32. The RFC2047 encoding function was originally intended for short strings
2607 such as real names; it was not keeping to the 75-character limit for
2608 encoded words that the RFC imposes. It now respects the limit, and
2609 generates multiple encoded words if necessary. To be on the safe side, I
2610 have increased the buffer size for the ${rfc2047: expansion operator from
2613 33. It is now permitted to omit both strings after an "if" condition; if the
2614 condition is true, the result is "true". As before, when the second string
2615 is omitted, a false condition yields an empty string. This makes it less
2616 cumbersome to write custom ACL and router conditions.
2618 34. Failure to deliver a bounce message always caused it to be frozen, even if
2619 there was an errors_to setting on the router. The errors_to setting is now
2622 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2623 canonical form (fully expanded) before being placed in
2624 $sender_host_address.
2626 36. The table in the code that translates DNS record types into text (T_A to
2627 "A" for instance) was missing entries for NS and CNAME. It is just possible
2628 that this could have caused confusion if both these types were looked up
2629 for the same domain, because the text type is used as part of Exim's
2630 per-process caching. But the chance of anyone hitting this buglet seems
2633 37. The dnsdb lookup has been extended in a number of ways.
2635 (1) There is a new type, "zns", which walks up the domain tree until it
2636 finds some nameserver records. It should be used with care.
2638 (2) There is a new type, "mxh", which is like "mx" except that it returns
2639 just the host names, not the priorities.
2641 (3) It is now possible to give a list of domains (or IP addresses) to be
2642 looked up. The behaviour when one of the lookups defers can be
2643 controlled by a keyword.
2645 (4) It is now possible to specify the separator character for use when
2646 multiple records are returned.
2648 38. The dnslists ACL condition has been extended: it is now possible to supply
2649 a list of IP addresses and/or domains to be looked up in a particular DNS
2652 39. Added log_selector=+queue_time_overall.
2654 40. When running the queue in the test harness, wait just a tad after forking a
2655 delivery process, to get repeatability of debugging output.
2657 41. Include certificate and key file names in error message when GnuTLS fails
2658 to set them up, because the GnuTLS error message doesn't include the name
2659 of the failing file when there is a problem reading it.
2661 42. Allow both -bf and -bF in the same test run.
2663 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2665 44. The "Exiscan patch" is now merged into the mainline Exim source.
2667 45. Sometimes the final signoff response after QUIT could fail to get
2668 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2669 < 0 before doing a fflush(). This bug looks as though it goes back to the
2670 introduction of TLS in release 3.20, but "sometimes" must have been rare
2671 because the tests only now provoked it.
2673 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2674 (this can affect the format of dates).
2676 47. exim_tidydb, when checking for the continued existence of a message for
2677 which it has found a message-specific retry record, was not finding
2678 messages that were in split spool directories. Consequently, it was
2679 deleting retry records that should have stayed in existence.
2681 48. Steve fixed some bugs in eximstats.
2683 49. The SPA authentication driver was not abandoning authentication and moving
2684 on to the next authenticator when an expansion was forced to fail,
2685 contradicting the general specification for all authenticators. Instead it
2686 was generating a temporary error. It now behaves as specified.
2688 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2689 (the order specifies the preference for clients). The order is now AES256,
2690 AES128, 3DES, ARCFOUR128.
2692 51. Small patch to Sieve code - explicitly set From: when generating an
2695 52. Exim crashed if a remote delivery caused a very long error message to be
2696 recorded - for instance if somebody sent an entire SpamAssassin report back
2697 as a large number of 550 error lines. This bug was coincidentally fixed by
2698 increasing the size of one of Exim's internal buffers (big_buffer) that
2699 happened as part of the Exiscan merge. However, to be on the safe side, I
2700 have made the code more robust (and fixed the comments that describe what
2703 53. Now that there can be additional text after "Completed" in log lines (if
2704 the queue_time_overall log selector is set), a one-byte patch to exigrep
2705 was needed to allow it to recognize "Completed" as not the last thing in
2708 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2709 patch that reportedly fixes this has been added. I am not expert enough to
2710 create a test for it. This is what the patch creator wrote:
2712 "I found a little strange behaviour of ldap code when working with
2713 Windows 2003 AD Domain, where users was placed in more than one
2714 Organization Units. When I tried to give exim partial DN, the exit code
2715 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2716 But simultaneously result of request was absolutely normal ldap result,
2717 so I produce this patch..."
2719 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2720 REFERENCE, so I have modified the code to exclude the patch when that macro
2723 55. Some experimental protocols are using DNS PTR records for new purposes. The
2724 keys for these records are domain names, not reversed IP addresses. The
2725 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2726 leaves it alone. Component reversal etc. now happens only for IP addresses.
2729 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2731 57. Double the size of the debug message buffer (to 2048) so that more of very
2732 long debug lines gets shown.
2734 58. The exicyclog utility now does better if the number of log files to keep
2735 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2737 59. Two changes related to the smtp_active_hostname option:
2739 (1) $smtp_active_hostname is now available as a variable.
2740 (2) The default for smtp_banner uses $smtp_active_hostname instead
2741 of $primary_hostname.
2743 60. The host_aton() function is supposed to be passed a string that is known
2744 to be a valid IP address. However, in the case of IPv6 addresses, it was
2745 not checking this. This is a hostage to fortune. Exim now panics and dies
2746 if the condition is not met. A case was found where this could be provoked
2747 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2748 components; fortuitously, this particular loophole had already been fixed
2749 by change 4.50/55 above.
2751 If there are any other similar loopholes, the new check in host_aton()
2752 itself should stop them being exploited. The report I received stated that
2753 data on the command line could provoke the exploit when Exim was running as
2754 exim, but did not say which command line option was involved. All I could
2755 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2756 running as the user.
2759 61. There was a buffer overflow vulnerability in the SPA authentication code
2760 (which came originally from the Samba project). I have added a test to the
2761 spa_base64_to_bits() function which I hope fixes it.
2764 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2765 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2767 63. The daemon start-up calls getloadavg() while still root for those OS that
2768 need the first call to be done as root, but it missed one case: when
2769 deliver_queue_load_max is set with deliver_drop_privilege. This is
2770 necessary for the benefit of the queue runner, because there is no re-exec
2771 when deliver_drop_privilege is set.
2773 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2774 This has been fixed.
2776 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2777 was in use, was not putting the data itself into the right store pool;
2778 consequently, it could be overwritten for a subsequent message in the same
2779 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2782 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2784 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2785 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2786 the function were treating the return as a boolean value, which happened to
2787 work because 0=false and not-0=true, but is not correct code.
2789 68. The host_aton() function was not handling scoped IPv6 addresses (those
2790 with, for example, "%eth0" on the end) correctly.
2792 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2793 negated items (that is, ~something) in unsigned ints. Some compilers
2794 apparently mutter when there is no cast.
2796 70. If an address verification called from an ACL failed, and did not produce a
2797 user-specific message (i.e. there was only a "system" message), nothing was
2798 put in $acl_verify_message. In this situation, it now puts the system
2801 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2802 session; change 4.31/43 added the unwanted input to the log line - except
2803 that it did not do this in the start of session case. It now does.
2805 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2806 This is wrong and can cause the other end to generate a synchronization
2807 error if it is another Exim or anything else that does the synchronization
2808 check. A QUIT command is no longer sent after a timeout.
2810 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2811 during host lookups.
2813 74. The defer_ok option of callout verification was not working if it was used
2814 when verifying addresses in header lines, that is, for this case:
2816 verify = header_sender/callout=defer_ok
2818 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2819 those file descriptors could be used for SMTP connections. If anything
2820 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2821 could be sent to the SMTP client, causing chaos. The daemon now opens
2822 stdin, stdout, and stderr to /dev/null when it puts itself into the
2825 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2826 log by default. The user can override this with suitable Perl magic.
2828 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2829 the log message when discard triggers, was not working for the DATA ACL or
2830 for the non-SMTP ACL.
2832 78. Error message wording change in sieve.c.
2834 79. If smtp_accept_max_per_host was set, the number of connections could be
2835 restricted to fewer than expected, because the daemon was trying to set up
2836 a new connection before checking whether the processes handling previous
2837 connections had finished. The check for completed processes is now done
2838 earlier. On busy systems, this bug wouldn't be noticed because something
2839 else would have woken the daemon, and it would have reaped the completed
2842 80. If a message was submitted locally by a user whose login name contained one
2843 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2844 It caused a spool format error. I have fixed the spool reading code. A
2845 related problem was that the "from" clause in the Received: line became
2846 illegal because of the space(s). It is now covered by ${quote_local_part.
2848 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2851 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2853 83. Patch from Sieve maintainer to fix unterminated string problem in
2854 "vacation" handling.
2856 84. Some minor changes to the Linux configuration files to help with other
2857 OS variants using glibc.
2859 85. One more patch for Sieve to update vacation handling to latest spec.
2862 ----------------------------------------------------
2863 See the note above about the 4.44 and 4.50 releases.
2864 ----------------------------------------------------
2870 1. Change 4.43/35 introduced a bug that caused file counts to be
2871 incorrectly computed when quota_filecount was set in an appendfile
2874 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2875 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2878 3. Give more explanation in the error message when the command for a transport
2879 filter fails to execute.
2881 4. There are several places where Exim runs a non-Exim command in a
2882 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2883 was being done only for the command run by the queryprogram router. It is
2884 now done for all such subprocesses. The other cases are: ${run, transport
2885 filters, and the commands run by the lmtp and pipe transports.
2887 5. Some older OS have a limit of 256 on the maximum number of file
2888 descriptors. Exim was using setrlimit() to set 1000 as a large value
2889 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2890 systems. I've change it so that if it can't get 1000, it tries for 256.
2892 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2893 was an oversight, and furthermore, ever since the addition of extra
2894 controls (e.g. 4.43/32), the checks on when to allow different forms of
2895 "control" were broken. There should now be diagnostics for all cases when a
2896 control that does not make sense is encountered.
2898 7. $recipients is now available in the predata ACL (oversight).
2900 8. Tidy the search cache before the fork to do a delivery from a message
2901 received from the command line. Otherwise the child will trigger a lookup
2902 failure and thereby defer the delivery if it tries to use (for example) a
2903 cached ldap connection that the parent has called unbind on.
2905 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2906 of $address_data from the recipient verification was clobbered by the
2907 sender verification.
2909 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2910 was its contents. (It was OK if the option was not defined at all.)
2912 11. A "Completed" log line is now written for messages that are removed from
2913 the spool by the -Mrm option.
2915 12. $host_address is now set to the target address during the checking of
2916 ignore_target_hosts.
2918 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2919 being passed; this would have caused $sender_host_name to have been used if
2920 matching the list had actually called for a host name (not very likely,
2921 since this list is usually IP addresses). A host name is now passed as
2924 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2925 code that specifies a non-restarting handler (typically sigaction() in
2926 modern systems) in an attempt to fix a rare and obscure crash bug.
2928 15. Narrowed the window for a race in the daemon that could cause it to ignore
2929 SIGCHLD signals. This is not a major problem, because they are used only to
2930 wake it up if nothing else does.
2932 16. A malformed maildirsize file could cause Exim to calculate negative values
2933 for the mailbox size or file count. Odd effects could occur as a result.
2934 The maildirsize information is now recalculated if the size or filecount
2937 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2938 support for a long time. Removed HAVE_SYS_VFS_H.
2940 18. Updated exipick to current release from John Jetmore.
2942 19. Allow an empty sender to be matched against a lookup in an address list.
2943 Previously the only cases considered were a regular expression, or an
2946 20. Exim went into a mad DNS lookup loop when doing a callout where the
2947 host was specified on the transport, if the DNS lookup yielded more than
2950 21. The RFC2047 encoding function was originally intended for short strings
2951 such as real names; it was not keeping to the 75-character limit for
2952 encoded words that the RFC imposes. It now respects the limit, and
2953 generates multiple encoded words if necessary. To be on the safe side, I
2954 have increased the buffer size for the ${rfc2047: expansion operator from
2957 22. Failure to deliver a bounce message always caused it to be frozen, even if
2958 there was an errors_to setting on the router. The errors_to setting is now
2961 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2962 canonical form (fully expanded) before being placed in
2963 $sender_host_address.
2965 24. Updated eximstats to version 1.33
2967 25. Include certificate and key file names in error message when GnuTLS fails
2968 to set them up, because the GnuTLS error message doesn't include the name
2969 of the failing file when there is a problem reading it.
2971 26. Expand error message when OpenSSL has problems setting up cert/key files.
2974 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2975 (this can affect the format of dates).
2977 28. exim_tidydb, when checking for the continued existence of a message for
2978 which it has found a message-specific retry record, was not finding
2979 messages that were in split spool directories. Consequently, it was
2980 deleting retry records that should have stayed in existence.
2982 29. eximstats updated to version 1.35
2983 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2984 1.35 - bugfix such that pie charts by volume are generated correctly
2986 30. The SPA authentication driver was not abandoning authentication and moving
2987 on to the next authenticator when an expansion was forced to fail,
2988 contradicting the general specification for all authenticators. Instead it
2989 was generating a temporary error. It now behaves as specified.
2991 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2992 (the order specifies the preference for clients). The order is now AES256,
2993 AES128, 3DES, ARCFOUR128.
2995 31. Small patch to Sieve code - explicitly set From: when generating an
2998 32. Exim crashed if a remote delivery caused a very long error message to be
2999 recorded - for instance if somebody sent an entire SpamAssassin report back
3000 as a large number of 550 error lines. This bug was coincidentally fixed by
3001 increasing the size of one of Exim's internal buffers (big_buffer) that
3002 happened as part of the Exiscan merge. However, to be on the safe side, I
3003 have made the code more robust (and fixed the comments that describe what
3006 33. Some experimental protocols are using DNS PTR records for new purposes. The
3007 keys for these records are domain names, not reversed IP addresses. The
3008 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3009 leaves it alone. Component reversal etc. now happens only for IP addresses.
3012 34. The host_aton() function is supposed to be passed a string that is known
3013 to be a valid IP address. However, in the case of IPv6 addresses, it was
3014 not checking this. This is a hostage to fortune. Exim now panics and dies
3015 if the condition is not met. A case was found where this could be provoked
3016 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3017 components; fortuitously, this particular loophole had already been fixed
3018 by change 4.50/55 or 4.44/33 above.
3020 If there are any other similar loopholes, the new check in host_aton()
3021 itself should stop them being exploited. The report I received stated that
3022 data on the command line could provoke the exploit when Exim was running as
3023 exim, but did not say which command line option was involved. All I could
3024 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3025 running as the user.
3028 35. There was a buffer overflow vulnerability in the SPA authentication code
3029 (which came originally from the Samba project). I have added a test to the
3030 spa_base64_to_bits() function which I hope fixes it.
3033 36. The daemon start-up calls getloadavg() while still root for those OS that
3034 need the first call to be done as root, but it missed one case: when
3035 deliver_queue_load_max is set with deliver_drop_privilege. This is
3036 necessary for the benefit of the queue runner, because there is no re-exec
3037 when deliver_drop_privilege is set.
3039 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3040 was in use, was not putting the data itself into the right store pool;
3041 consequently, it could be overwritten for a subsequent message in the same
3042 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3045 38. Sometimes the final signoff response after QUIT could fail to get
3046 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3047 < 0 before doing a fflush(). This bug looks as though it goes back to the
3048 introduction of TLS in release 3.20, but "sometimes" must have been rare
3049 because the tests only now provoked it.
3055 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3056 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3057 Now it returns an integer. A number of calls were still expecting a T/F
3058 return. Fortuitously, in all cases, the tests worked in OK situations,
3059 which is the norm. However, things would have gone wrong on any write
3060 failures on the smtp file descriptor. This function is used when sending
3061 messages over SMTP and also when doing verify callouts.
3063 2. When Exim is called to do synchronous delivery of a locally submitted
3064 message (the -odf or -odi options), it no longer closes stderr before doing
3067 3. Implemented the mua_wrapper option.
3069 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3071 5. Implemented the functions header_remove(), header_testname(),
3072 header_add_at_position(), and receive_remove_recipient(), and exported them
3075 6. If an ACL "warn" statement specified the addition of headers, Exim already
3076 inserted X-ACL-Warn: at the start if there was no header name. However, it
3077 was not making this test for the second and subsequent header lines if
3078 there were newlines in the string. This meant that an invalid header could
3079 be inserted if Exim was badly configured.
3081 7. Allow an ACL "warn" statement to add header lines at the start or after all
3082 the Received: headers, as well as at the end.
3084 8. Added the rcpt_4xx retry error code.
3086 9. Added postmaster_mailfrom=xxx to callout verification option.
3088 10. Added mailfrom=xxxx to the callout verification option, for verify=
3091 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3092 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3094 12. Inserted some casts to stop certain compilers warning when using pointer
3095 differences as field lengths or precisions in printf-type calls (mostly
3096 affecting debugging statements).
3098 13. Added optional readline() support for -be (dynamically loaded).
3100 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3101 same clock tick as a message's arrival, so that its received time was the
3102 same as the "first fail" time on the retry record, and that message
3103 remained on the queue past the ultimate address timeout, every queue runner
3104 would try a delivery (because it was past the ultimate address timeout) but
3105 after another failure, the ultimate address timeout, which should have then
3106 bounced the address, did not kick in. This was a "< instead of <=" error;
3107 in most cases the first failure would have been in the next clock tick
3108 after the received time, and all would be well.
3110 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3111 being recognized when the domain list was tested by the match_domain
3112 condition in an expansion string.
3114 16. Added the ${str2b64: operator.
3116 17. Exim was always calling setrlimit() to set a large limit for the number of
3117 processes, without checking whether the existing limit was already
3118 adequate. (It did check for the limit on file descriptors.) Furthermore,
3119 errors from getrlimit() and setrlimit() were being ignored. Now they are
3120 logged to the main and panic logs, but Exim does carry on, to try to do its
3121 job under whatever limits there are.
3123 18. Imported PCRE 5.0.
3125 19. Trivial typo in log message " temporarily refused connection" (the leading
3128 20. If the log selector return_path_on_delivery was set and an address was
3129 redirected to /dev/null, the delivery process crashed because it assumed
3130 that a return path would always be set for a "successful" delivery. In this
3131 case, the whole delivery is bypassed as an optimization, and therefore no
3134 21. Internal re-arrangement: the function for sending a challenge and reading
3135 a response while authentication was assuming a zero-terminated challenge
3136 string. It's now changed to take a pointer and a length, to allow for
3137 binary data in such strings.
3139 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3141 23. Exim was not respecting finduser_retries when seeking the login of the
3142 uid under which it was called; it was always trying 10 times. (The default
3143 setting of finduser_retries is zero.) Also, it was sleeping after the final
3144 failure, which is pointless.
3146 24. Implemented tls_on_connect_ports.
3148 25. Implemented acl_smtp_predata.
3150 26. If the domain in control=submission is set empty, Exim assumes that the
3151 authenticated id is a complete email address when it generates From: or
3152 Sender: header lines.
3154 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3155 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3156 chown and chgrp in /bin and hostname in /usr/bin.
3158 28. Exim was keeping the "process log" file open after each use, just as it
3159 does for the main log. This opens the possibility of it remaining open for
3160 long periods when the USR1 signal hits a daemon. Occasional processlog
3161 errors were reported, that could have been caused by this. Anyway, it seems
3162 much more sensible not to leave this file open at all, so that is what now
3165 29. The long-running daemon process does not normally write to the log once it
3166 has entered its main loop, and it closes the log before doing so. This is
3167 so that log files can straightforwardly be renamed and moved. However,
3168 there are a couple of unusual error situations where the daemon does write
3169 log entries, and I had neglected to close the log afterwards.
3171 30. The text of an SMTP error response that was received during a remote
3172 delivery was being truncated at 512 bytes. This is too short for some of
3173 the long messages that one sometimes sees. I've increased the limit to
3176 31. It is now possible to make retry rules that apply only when a message has a
3177 specific sender, in particular, an empty sender.
3179 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3180 it possible to be selective about when SMTP synchronization is enforced.
3182 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3184 32. Implemented hosts_connection_nolog.
3186 33. Added an ACL for QUIT.
3188 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3191 35. Added mailbox_size and mailbox_filecount to appendfile.
3193 36. Added control = no_multiline_responses to ACLs.
3195 37. There was a bug in the logic of the code that waits for the clock to tick
3196 in the case where the clock went backwards by a substantial amount such
3197 that the microsecond fraction of "now" was more than the microsecond
3198 fraction of "then" (but the whole seconds number was less).
3200 38. Added support for the libradius Radius client library this is found on
3201 FreeBSD (previously only the radiusclient library was supported).
3207 1. When certain lookups returned multiple values in the form name=value, the
3208 quoting of the values was not always being done properly. Specifically:
3209 (a) If the value started with a double quote, but contained no whitespace,
3211 (b) If the value contained whitespace other than a space character (i.e.
3212 tabs or newlines or carriage returns) it was not quoted.
3213 This fix has been applied to the mysql and pgsql lookups by writing a
3214 separate quoting function and calling it from the lookup code. The fix
3215 should probably also be applied to nisplus, ibase and oracle lookups, but
3216 since I cannot test any of those, I have not disturbed their existing code.
3218 2. A hit in the callout cache for a specific address caused a log line with no
3219 reason for rejecting RCPT. Now it says "Previous (cached) callout
3220 verification failure".
3222 3. There was an off-by-one bug in the queryprogram router. An over-long
3223 return line was truncated at 256 instead of 255 characters, thereby
3224 overflowing its buffer with the terminating zero. As well as fixing this, I
3225 have increased the buffer size to 1024 (and made a note to document this).
3227 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3228 when Exim is waiting for an SMTP response from a remote server, Exim
3229 restarts its select() call on the socket, thereby resetting its timeout.
3230 This is not a problem when such interrupts are rare. Somebody set up a cron
3231 job to run exiwhat every 2 minutes, which is less than the normal select()
3232 timeout (5 or 10 minutes). This meant that the select() timeout never
3233 kicked in because it was always reset. I have fixed this by comparing the
3234 time when an interrupt arrives with the time at the start of the first call
3235 to select(). If more time than the timeout has elapsed, the interrupt is
3236 treated as a timeout.
3238 5. Some internal re-factoring in preparation for the addition of Sieve
3239 extensions (by MH). In particular, the "personal" test is moved to a
3240 separate function, and given an option for scanning Cc: and Bcc: (which is
3241 not set for Exim filters).
3243 6. When Exim created an email address using the login of the caller as the
3244 local part (e.g. when creating a From: or Sender: header line), it was not
3245 quoting the local part when it contained special characters such as @.
3247 7. Installed new OpenBSD configuration files.
3249 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3250 try to make them clearer.
3252 9. Callout options, other than the timeout value, were being ignored when
3253 verifying sender addresses in header lines. For example, when using
3255 verify = header_sender/callout=no_cache
3257 the cache was (incorrectly) being used.
3259 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3260 only the Cygwin environment.
3262 11. When return_path_on_delivery was set as a log selector, if different remote
3263 addresses in the same message used different return paths and parallel
3264 remote delivery occurred, the wrong values would sometimes be logged.
3265 (Whenever a remote delivery process finished, the return path value from
3266 the most recently started remote delivery process was logged.)
3268 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3269 lines when AUTH and/or TLS are in use. This is the "received protocol"
3270 field. Exim used to use "asmtp" for authenticated SMTP, without any
3271 indication (in the protocol name) for TLS use. Now it follows the RFC and
3272 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3273 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3274 names appear in log lines as well as in Received: header lines.
3276 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3277 extensions, and comparison tests, and to fix some bugs.
3279 14. Changes to the "personal" filter test:
3281 (1) The test was buggy in that it was just doing the equivalent of
3282 "contains" tests on header lines. For example, if a user's address was
3283 anne@some.where, the "personal" test would incorrectly be true for
3285 To: susanne@some.where
3287 This test is now done by extracting each address from the header in turn,
3288 and checking the entire address. Other tests that are part of "personal"
3289 are now done using regular expressions (for example, to check local parts
3290 of addresses in From: header lines).
3292 (2) The list of non-personal local parts in From: addresses has been
3293 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3294 taken from the Sieve specification recommendations.
3296 (3) If the message contains any header line starting with "List-" it is
3297 treated as non-personal.
3299 (4) The test for "circular" in the Subject: header line has been removed
3300 because it now seems ill-conceived.
3302 15. Minor typos in src/EDITME comments corrected.
3304 16. Installed latest exipick from John Jetmore.
3306 17. If headers_add on a router specified a text string that was too long for
3307 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3308 of string_sprintf() is now avoided.
3310 18. $message_body_size was not set (it was always zero) when running the DATA
3311 ACL and the local_scan() function.
3313 19. For the "mail" command in an Exim filter, no default was being set for
3314 the once_repeat time, causing a random time value to be used if "once" was
3315 specified. (If the value happened to be <= 0, no repeat happened.) The
3316 default is now 0s, meaning "never repeat". The "vacation" command was OK
3317 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3318 (I found it when inspecting the code).
3320 20. There is now an overall timeout for performing a callout verification. It
3321 defaults to 4 times the callout timeout, which applies to individual SMTP
3322 commands during the callout. The overall timeout applies when there is more
3323 than one host that can be tried. The timeout is checked before trying the
3324 next host. This prevents very long delays if there are a large number of
3325 hosts and all are timing out (e.g. when the network connections are timing
3326 out). The value of the overall timeout can be changed by specifying an
3327 additional sub-option for "callout", called "maxwait". For example:
3329 verify = sender/callout=5s,maxwait=20s
3331 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3332 to the end before writing, but this should make it even safer).
3334 22. Exim was forgetting that it had advertised PIPELINING for the second and
3335 subsequent messages on an SMTP connection. It was also not resetting its
3336 memory on STARTTLS and an internal HELO.
3338 23. When Exim logs an SMTP synchronization error within a session, it now
3339 records whether PIPELINING has been advertised or not.
3341 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3342 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3343 rather than long int.
3345 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3347 26. Added the never_mail option to autoreply.
3353 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3354 crash if the getsockname() call failed; this can happen if a connection is
3355 closed very soon after it is established. The problem was simply in the
3356 order in which certain operations were done, causing Exim to try to write
3357 to the SMTP stream before it had set up the file descriptor. The bug has
3358 been fixed by making things happen in the correct order.
3364 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3365 before the connection was closed, thus losing the rejection response.
3367 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3368 some early Solaris releases, but causes trouble in current releases where
3369 socklen_t is defined.
3371 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3374 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3377 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3378 that releasing the top of it at the end releases what was used for sub-
3379 expansions (unless the block got too big). However, discard this block if
3380 the first thing is a variable or header, so that we can use its block when
3381 it is dynamic (useful for very large $message_headers, for example).
3383 6. Lookups now cache *every* query, not just the most recent. A new, separate
3384 store pool is used for this. It can be recovered when all lookup caches are
3385 flushed. Lookups now release memory at the end of their result strings.
3386 This has involved some general refactoring of the lookup sources.
3388 7. Some code has been added to the store_xxx() functions to reduce the amount
3389 of flapping under certain conditions.
3391 8. log_incoming_interface used to affect only the <= reception log lines. Now
3392 it causes the local interface and port to be added to several more SMTP log
3393 lines, for example "SMTP connection from", and rejection lines.
3395 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3397 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3399 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3400 could be overwritten at the end of the current message (or the start of a
3401 new message if it was set in a HELO ACL). The value is now preserved for
3402 the duration of the SMTP connection.
3404 12. If a transport had a headers_rewrite setting, and a matching header line
3405 contained an unqualified address, that address was qualified, even if it
3406 did not match any rewriting rules. The underlying bug was that the values
3407 of the flags that permit the existence of unqualified sender and recipient
3408 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3409 non-local messages, and by -bnq for local messages) were not being
3410 preserved with the message after it was received.
3412 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3413 "next input=" as part of the text comprising the host identity instead of
3414 the correct text. The code was using the same buffer for two different
3415 strings. However, depending on which order the printing function evaluated
3416 its arguments, the bug did not always show up. Under Linux, for example, my
3417 test suite worked just fine.
3419 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3420 doesn't work with some older versions of Perl. It has been changed to "my",
3421 which in any case is probably the better facility to use.
3423 15. A really picky compiler found some instances of statements for creating
3424 error messages that either had too many or two few arguments for the format
3427 16. The size of the buffer for calls to the DNS resolver has been increased
3428 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3429 for addresses that have a lot of PTR records. This alleviates a problem; it
3430 does not fully solve it.
3432 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3433 buffer now truncates the list and logs the incident, which is the same
3434 action as happens when Exim is looking up a host name and its aliases.
3435 Previously in this situation something unpredictable would happen;
3436 sometimes it was "internal error: store_reset failed".
3438 18. If a server dropped the connection unexpectedly when an Exim client was
3439 using GnuTLS and trying to read a response, the client delivery process
3440 crashed while trying to generate an error log message.
3442 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3443 string, for example:
3445 warn message = H1: something\nH2: something
3447 the text was added as a single header line from Exim's point of view
3448 though it ended up OK in the delivered message. However, searching for the
3449 second and subsequent header lines using $h_h2: did not work. This has been
3450 fixed. Similarly, if a system filter added multiple headers in this way,
3451 the routers could not see them.
3453 20. Expanded the error message when iplsearch is called with an invalid key to
3454 suggest using net-iplsearch in a host list.
3456 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3457 ACLs are no longer actually imposed (and a message to that effect is
3460 22. If a "gecos" field in a passwd entry contained escaped characters, in
3461 particular, if it contained a \" sequence, Exim got it wrong when building
3462 a From: or a Sender: header from that name. A second bug also caused
3463 incorrect handling when an unquoted " was present following a character
3464 that needed quoting.
3466 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3467 was not being matched caselessly.
3469 24. Arranged for all hyphens in the exim.8 source to be escaped with
3472 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3473 properly. Recipient callout cache records were still being keyed to include
3474 the sender, even when use_sender was set false. This led to far more
3475 callouts that were necessary. The sender is no longer included in the key
3476 when use_sender is false.
3478 26. Added "control = submission" modifier to ACLs.
3480 27. Added the ${base62d: operator to decode base 62 numbers.
3482 28. dnsdb lookups can now access SRV records.
3484 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3485 the configuration file.
3487 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3488 (-v) mode. This makes the output for a verbose queue run more intelligible.
3490 31. Added a use_postmaster feature to recipient callouts.
3492 32. Added the $body_zerocount variable, containing the number of binary zero
3493 bytes in the message body.
3495 33. The time of last modification of the "new" subdirectory is now used as the
3496 "mailbox time last read" when there is a quota error for a maildir
3499 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3501 35. Added +ignore_unknown as a special item in host lists.
3503 36. Code for decoding IPv6 addresses in host lists is now included, even if
3504 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3505 address was recognized as an IP address, but was then not correctly decoded
3506 into binary, causing unexpected and incorrect effects when compared with
3513 1. Very minor rewording of debugging text in manualroute to say "list of
3514 hosts" instead of "hostlist".
3516 2. If verify=header_syntax was set, and a header line with an unqualified
3517 address (no domain) and a large number of spaces between the end of the
3518 name and the colon was received, the reception process suffered a buffer
3519 overflow, and (when I tested it) crashed. This was caused by some obsolete
3520 code that should have been removed. The fix is to remove it!
3522 3. When running in the test harness, delay a bit after writing a bounce
3523 message to get a bit more predictability in the log output.
3525 4. Added a call to search_tidyup() just before forking a reception process. In
3526 theory, someone could use a lookup in the expansion of smtp_accept_max_
3527 per_host which, without the tidyup, could leave open a database connection.
3529 5. Added the variables $recipient_data and $sender_data which get set from a
3530 lookup success in an ACL "recipients" or "senders" condition, or a router
3531 "senders" option, similar to $domain_data and $local_part_data.
3533 6. Moved the writing of debug_print from before to after the "senders" test
3536 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3537 problems for message scanning, either using a data ACL, or using
3538 local_scan() because the Received: header was not generated till after they
3539 were called (in order to set the time as the time of reception completion).
3540 I have revised the way this works. The header is now generated after the
3541 body is received, but before the ACL or local_scan() are called. After they
3542 are run, the timestamp in the header is updated.
3548 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3549 before starting a queue runner without re-exec. This happened only when
3550 deliver_drop_privilege was set or when the Exim user was set to root. The
3551 effect of the bug was that timeouts during subsequent deliveries caused
3552 crashes instead of being properly handled. The handler is now left at its
3553 default (and expected) setting.
3555 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3556 message, again when deliver_drop_privilege is set or Exim is run as root.
3557 The bug described in (1) was not present in this case, but the tidying up
3558 of the other signals was missing. I have made the two cases consistent.
3560 3. The ignore_target_hosts setting on a manualroute router was being ignored
3561 for hosts that were looked up using the /MX notation.
3563 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3566 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3567 operated on the sender address. After changing the $sender_address to <>
3568 for the sender address verify, Exim was re-instated it as the original
3569 (before rewriting) address, but remembering that it had rewritten it, so it
3570 wasn't rewriting it again. This bug also had the effect of breaking the
3571 sender address verification caching when the sender address was rewritten.
3573 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3574 This has been changed so that if the ip literal address matches
3575 ignore_target_hosts, the router declines.
3577 7. Added expansion conditions match_domain, match_address, and match_local_
3578 part (NOT match_host).
3580 8. The placeholder for the Received: header didn't have a length field set.
3582 9. Added code to Exim itself and to exim_lock to test for a specific race
3583 condition that could lead to file corruption when using MBX delivery. The
3584 issue is with the lockfile that is created in /tmp. If this file is removed
3585 after a process has opened it but before that process has acquired a lock,
3586 there is the potential for a second process to recreate the file and also
3587 acquire a lock. This could lead to two Exim processes writing to the file
3588 at the same time. The added code performs the same test as UW imapd; it
3589 checks after acquiring the lock that its file descriptor still refers to
3590 the same named file.
3592 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3593 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3596 11. Added the smtp_active_hostname option. If used, this will typically be made
3597 to depend on the incoming interface address. Because $interface_address is
3598 not set up until the daemon has forked a reception process, error responses
3599 that can happen earlier (such as "too many connections") no longer contain
3602 12. If an expansion in a condition on a "warn" statement fails because a lookup
3603 defers, the "warn" statement is abandoned, and the next ACL statement is
3604 processed. Previously this caused the whole ACL to be aborted.
3606 13. Added the iplsearch lookup type.
3608 14. Added ident_timeout as a log selector.
3610 15. Added tls_certificate_verified as a log selector.
3612 16. Added a global option tls_require_ciphers (compare the smtp transport
3613 option of the same name). This controls incoming TLS connections.
3615 17. I finally figured out how to make tls_require_ciphers do a similar thing
3616 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3617 before starting the TLS session.
3619 18. Tabs are now shown as \t in -bP output.
3621 19. If the log selector return_path_on_delivery was set, Exim crashed when
3622 bouncing a message because it had too many Received: header lines.
3624 20. If two routers both had headers_remove settings, and the first one included
3625 a superfluous trailing colon, the final name in the first list and the
3626 first name in the second list were incorrectly joined into one item (with a
3627 colon in the middle).
3633 1. Added -C and -D options to the exinext utility, mainly to make it easier
3634 to include in the automated testing, but these could be helpful when
3635 multiple configurations are in use.
3637 2. The exinext utility was not formatting the output nicely when there was
3638 an alternate port involved in the retry record key, nor when there was a
3639 message id as well (for retries that were specific to a specific message
3640 and a specific host). It was also confused by IPv6 addresses, because of
3641 the additional colons they contain. I have fixed the IPv4 problem, and
3642 patched it up to do a reasonable job for IPv6.
3644 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3645 delivery, the log line now contains "pipelined" if PIPELINING was used.
3647 4. An SMTP transport process used to panic and die if the bind() call to set
3648 an explicit outgoing interface failed. This has been changed; it is now
3649 treated in the same way as a connect() failure.
3651 5. A reference to $sender_host_name in the part of a conditional expansion
3652 that was being skipped was still causing a DNS lookup. This no longer
3655 6. The def: expansion condition was not recognizing references to header lines
3656 that used bh_ and bheader_.
3658 7. Added the _cache feature to named lists.
3660 8. The code for checking quota_filecount in the appendfile transport was
3661 allowing one more file than it should have been.
3663 9. For compatibility with Sendmail, the command line option
3671 and sets the incoming protocol and host name (for trusted callers). The
3672 host name and its colon can be omitted when only the protocol is to be set.
3673 Note the Exim already has two private options, -pd and -ps, that refer to
3674 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3675 "s", but I don't think that's a major issue.
3677 10. A number of refactoring changes to the code, none of which should affect
3680 (a) The number of logging options was getting close to filling up the
3681 32-bit word that was used as a bit map. I have split them into two classes:
3682 those that are passed in the argument to log_write(), and those that are
3683 only ever tested independently outside of that function. These are now in
3684 separate 32-bit words, so there is plenty of room for expansion again.
3685 There is no change in the user interface or the logging behaviour.
3687 (b) When building, for example, log lines, the code previously used a
3688 macro that called string_cat() twice, in order to add two strings. This is
3689 not really sufficiently general. Furthermore, there was one instance where
3690 it was actually wrong because one of the argument was used twice, and in
3691 one call a function was used. (As it happened, calling the function twice
3692 did not affect the overall behaviour.) The macro has been replaced by a
3693 function that can join an arbitrary number of extra strings onto a growing
3696 (c) The code for expansion conditions now uses a table and a binary chop
3697 instead of a serial search (which was left over from when there were very
3698 few conditions). Also, it now recognizes conditions like "pam" even when
3699 the relevant support is not compiled in: a suitably worded error message is
3700 given if an attempt is made to use such a condition.
3702 11. Added ${time_interval:xxxxx}.
3704 12. A bug was causing one of the ddress fields not to be passed back correctly
3705 from remote delivery subprocesses. The field in question was not being
3706 subsequently used, so this caused to problems in practice.
3708 13. Added new log selectors queue_time and deliver_time.
3710 14. Might have fixed a bug in maildirsizefile handling that threw up
3711 "unexpected character" debug warnings, and recalculated the data
3712 unnecessarily. In any case, I expanded the warning message to give more
3715 15. Added the message "Restricted characters in address" to the statements in
3716 the default ACL that block characters like @ and % in local parts.
3718 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3719 Three changes have been made:
3721 (a) There was a serious bug; a negative response to MAIL caused the whole
3722 recipient domain to be cached as invalid, thereby blocking all messages
3723 to all local parts at the same domain, from all senders. This bug has
3724 been fixed. The domain is no longer cached after a negative response to
3725 MAIL if the sender used is not empty.
3727 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3730 (c) A new callout option, "use_sender" has been added for people who want
3731 the modified behaviour.
3737 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3740 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3741 indeed breaks things for older releases.
3743 3. Added additional logging to the case where there is a problem reading data
3744 from a filter that is running in a subprocess using a pipe, in order to
3745 try to track down a specific problem.
3747 4. Testing facility fudge: when running in the test harness and attempting
3748 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3749 getting "No route to host". Convert this to a timeout.
3751 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3754 6. Some OS don't have socklen_t but use size_t instead. This affects the
3755 fifth argument of getsockopt() amongst other things. This is now
3756 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3757 can be set for individual OS. I have set it for SunOS5, OSF1, and
3758 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3759 some earlier ones do not.
3761 7. Change 4.30/15 was not doing the test caselessly.
3763 8. The standard form for an IPv6 address literal was being rejected by address
3764 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3765 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3766 this, as well as the form without the "IPv6" on the front (but only when
3767 address literals are enabled, of course).
3769 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3771 10. Exim crashed if a message with an empty sender address specified by -f
3772 encountered a router with an errors_to setting. This could be provoked only
3773 by a command such as
3777 where an empty string was supplied; "<>" did not hit this bug.
3779 11. Installed PCRE release 4.5.
3781 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3782 remained set. It is now erased.
3784 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3785 times from message ids (which are base 36 rather than the normal 62).
3787 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3788 were being counted as actual protocol errors, and logged if the log
3789 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3790 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3791 and DATA following a set of rejected RCPTs do not count as protocol errors.
3792 In other words, Exim assumes they were pipelined, though this may not
3793 actually be the case. Of course, in all cases the client gets an
3794 appropriate error code.
3796 15. If a lookup fails in an ACL condition, a message about the failure may
3797 be available; it is used if testing the ACL cannot continue, because most
3798 such messages specify what the cause of the deferral is. However, some
3799 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3800 that caused an old message to be retained and used if a later statement
3801 caused a defer, replacing the real cause of the deferral.
3803 16. If an IP address had so many PTR records that the DNS lookup buffer
3804 was not large enough to hold them, Exim could crash while trying to process
3805 the truncated data. It now detects and logs this case.
3807 17. Further to 4.21/58, another change has been made: if (and only if) the
3808 first line of a message (the first header line) ends with CRLF, a bare LF
3809 in a subsequent header line has a space inserted after it, so as not to
3810 terminate the header.
3812 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3813 unnecessarily, used atoi() instead of strtol(), and didn't check the
3814 termination when getting file sizes from file names by regex.
3816 19. Completely re-implemented the support for maildirsize files, in the light
3817 of a number of problems with the previous contributed implementation
3818 (4.30/29). In particular:
3820 . If the quota is zero, the maildirsize file is maintained, but no quota is
3823 . If the maildir directory does not exist, it is created before any attempt
3824 to write a maildirsize file.
3826 . The quota value in the file is just a cache; if the quota is changed in
3827 the transport, the new value overrides.
3829 . A regular expression is available for excluding directories from the
3832 20. The autoreply transport checks the characters in options that define the
3833 message's headers; it allows continued headers, but it was checking with
3834 isspace() after an embedded newline instead of explicitly looking for a
3837 21. If all the "regular" hosts to which an address was routed had passed their
3838 expiry times, and had not reached their retry times, the address was
3839 bounced, even if fallback hosts were defined. Now Exim should go on to try
3842 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3843 equivalent code in the SMTP transport. Some hosts send humungous responses
3844 to HELO/EHLO, more than 1024 it seems.
3846 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3847 gives compiler warnings in some environments. I've now done it "properly",
3850 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3851 (because of problems with the built-in one) was declared to return uschar *
3852 instead of char *, causing compiler failure.
3854 25. Fixed a file descriptor leak when processing alias/forward files.
3856 26. Fixed a minor format string issue in dbfn.c.
3858 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3860 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3861 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3864 29. When a sender address is verified, it is cached, to save repeating the test
3865 when there is more than one recipient in a message. However, when the
3866 verification involves a callout, it is possible for different callout
3867 options to be set for different recipients. It is too complicated to keep
3868 track of this in the cache, so now Exim always runs a verification when a
3869 callout is required, relying on the callout cache for the optimization.
3870 The overhead is duplication of the address routing, but this should not be
3873 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3874 to be verified with callout=postmaster, and the main callout worked but the
3875 postmaster check failed, the verification correctly failed. However, if a
3876 subsequent RCPT command asked for sender verification *without* the
3877 postmaster check, incorrect caching caused this verification also to fail,
3880 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3881 it was not caching the DNS options (qualify_single, search_parents) that
3882 were used when the lookup failed. A subsequent lookup with different
3883 options therefore always gave the same answer, though there were cases
3884 where it should not have. (Example: a "domains = !$mx_any" option on a
3885 dnslookup router: the "domains" option is always processed without any
3886 widening, but the router might have qualify_single set.) Now Exim uses the
3887 cached value only when the same options are set.
3889 32. Added John Jetmore's "exipick" utility to the distribution.
3891 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3892 than a timeout (e.g. a certificate is required, and is not provided), an
3893 Exim server now closes the connection immediately. Previously it waited for
3894 the client to close - but if the client is SSL, it seems that they each
3895 wait for each other, leading to a delay before one of them times out.
3897 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3898 maintained 0.8.x compatibility because I don't think many are using it, and
3899 it is clearly obsolete.
3901 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3904 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3905 client certificate was expired. A simple patch fixes this, though I don't
3906 understand the full logic of why the verify callback is called multiple
3909 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3910 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3911 which causes problems with some clients (such as the Certicom SSL Plus
3912 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3913 disables the coutermeasure allowing Eudora to connect."
3915 38. Exim was not checking that a write() to a log file succeeded. This could
3916 lead to Bad Things if a log got too big, in particular if it hit a file
3917 size limit. Exim now panics and dies if it cannot write to a log file, just
3918 as it does if it cannot open a log file.
3920 39. Modified OS/Makefile-Linux so that it now contains
3922 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3924 The two -D definitions ensure that Exim is compiled with large file
3925 support, which makes it possible to handle log files that are bigger than
3928 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3929 instance) a domain was checked against a named list that involved a lookup,
3930 causing $domain_data to be set, then another domain was checked against the
3931 same list, then the first domain was re-checked, the value of $domain_data
3932 after the final check could be wrong. In particular, if the second check
3933 failed, it could be set empty. This bug probably also applied to
3936 41. The strip_trailing_dot option was not being applied to the address given
3937 with the -f command-line option.
3939 42. The code for reading a message's header from the spool was incrementing
3940 $received_count, but never initializing it. This meant that the value was
3941 incorrect (doubled) while delivering a message in the same process in which
3942 it was received. In the most common configuration of Exim, this never
3943 happens - a fresh exec is done - but it can happen when
3944 deliver_drop_privilege is set.
3946 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3947 it now includes up to 150 characters of the unexpected data in the log
3950 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3951 and building data strings. The size of both of these buffers was 10 000
3952 bytes - far larger than anybody would *ever* want, thought I. Needless to
3953 say, somebody hit the limit. I have increased the maximum line length to
3954 20 000 and the maximum data length of concatenated lines to 100 000. I have
3955 also fixed two bugs, because there was no checking on these buffers. Tsk,
3956 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3957 buffer is too small.
3959 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3960 lsearch lookups. Now it does.
3962 46. When parsing a route_list item in a manualroute router, a fixed-length
3963 buffer was used for the list of hosts. I made this 1024 bytes long,
3964 thinking that nobody would ever have a list of hosts that long. Wrong.
3965 Somebody had a whole pile of complicated expansion conditions, and the
3966 string was silently truncated, leading to an expansion error. It turns out
3967 that it is easier to change to an unlimited length (owing to other changes
3968 that have happened since this code was originally written) than to build
3969 structure for giving a limitation error. The length of the item that
3970 expands into the list of hosts is now unlimited.
3972 47. The lsearch lookup could not handle data where the length of text line was
3973 more than 4095 characters. Such lines were truncated, leading to shortened
3974 data being returned. It should now handle lines of any length.
3976 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3977 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3980 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3981 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3982 no longer happens in comment lines. A list of replacements is now placed
3983 at the head of all of the source files, except those whose only change is
3984 to replace PERL_COMMAND in the very first #! line.
3986 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3987 messages on the queue, with a bottom-up merge sort, using code contributed
3988 by Michael Haardt. This should make operations like -bp somewhat faster on
3989 large queues. It won't affect queue runners, except when queue_run_in_order
3992 51. Installed eximstats 1.31 in the distribution.
3994 52. Added support for SRV lookups to the dnslookup router.
3996 53. If an ACL referred to $message_body or $message_body_end, the value was not
3997 reset for any messages that followed in the same SMTP session.
3999 54. The store-handling optimization for building very long strings was not
4000 differentiating between the different store pools. I don't think this
4001 actually made any difference in practice, but I've tidied it.
4003 55. While running the routers to verify a sender address, $sender_address
4004 was still set to the sender address. This is wrong, because when routing to
4005 send a bounce to the sender, it would be empty. Therefore, I have changed
4006 it so that, while verifying a sender address, $sender_address is set to <>.
4007 (There is no change to what happens when verifying a recipient address.)
4009 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4010 target A or AAAA records (if not already returned) without resetting the
4011 qualify_single or search_parents options of the DNS resolver. These are
4012 inappropriate in this case because the targets of MX and SRV records must
4013 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4014 target that, when qualified, matched something in the local domain. These
4015 two options are now turned off when doing these lookups.
4017 57. It seems that at least some releases of Reiserfs (which does not have the
4018 concept of a fixed number of inodes) returns zero and not -1 for the
4019 number of available inodes. This interacted badly with check_spool_inodes,
4020 which assumed that -1 was the "no such thing" setting. What I have done is
4021 to check that the total number of inodes is greater than zero before doing
4022 the test of how many are available.
4024 58. When a "warn" ACL statement has a log_message modifier, the message is
4025 remembered, and not repeated. This is to avoid a lot of repetition when a
4026 message has many recipients that cause the same warning to be written.
4027 Howewer, Exim was preserving the list of already written lines for an
4028 entire SMTP session, which doesn't seem right. The memory is now reset if a
4029 new message is started.
4031 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4032 debugging output unless log_rewrite was also set.
4034 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4035 of (int)(handle) when we know that handle contains (void *)(-1).
4037 61. The Exim daemon panic-logs an error return when it closes the incoming
4038 connection. However "connection reset by peer" seems to be common, and
4039 isn't really an error worthy of noting specially, so that particular error
4042 62. When Exim is trying to find all the local interfaces, it used to panic and
4043 die if the ioctl to get the interface flags failed. However, it seems that
4044 on at least one OS (Solaris 9) it is possible to have an interface that is
4045 included in the list of interfaces, but for which you get a failure error
4046 for this call. This happens when the interface is not "plumbed" into a
4047 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4048 failure of the "get flags" call assumes that the interface is down.
4050 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4051 makes life easier for people who are doing arithmetic on fields extracted
4052 from dates, where you often get leading zeros that should not be
4053 interpreted as octal.
4055 64. Added qualify_domain to the redirect router, to override the global
4058 65. If a pathologically long header line contained very many addresses (the
4059 report of this problem mentioned 10 000) and each of them was rewritten,
4060 Exim could use up a very large amount of memory. (It kept on making new
4061 copies of the header line as it rewrote, and never released the old ones.)
4062 At the expense of a bit more processing, the header rewriting function has
4063 been changed so that it no longer eats memory in this way.
4065 66. The generation of the Received: header has been moved from the time that a
4066 message starts to be received, to the time that it finishes. The timestamp
4067 in the Received: header should now be very close to that of the <= log
4068 line. There are two side-effects of this change:
4070 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4071 logged header lines no longer include the local Received: line, because
4072 it has not yet been created. The same applies to a copy of the message
4073 that is returned to a non-SMTP sender when a message is rejected.
4075 (b) When a filter file is tested using -bf, no additional Received: header
4076 is added to the test message. After some thought, I decided that this
4079 This change does not affect the value of $received_for. It is still set
4080 after address rewriting, but before local_scan() is called.
4082 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4084 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4085 gave an unhelpful panic error message, and a defer error. I have managed to
4086 change this behaviour so that it now rejects any supplied certificate,
4087 which seems right, as the list of acceptable certificates is empty.
4089 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4090 gave an unhelpful defer error. I have not managed to make this reject any
4091 supplied certificates, but the error message it gives is "no certificate
4092 supplied", which is not helpful.
4094 70. exigrep's output now also includes lines that are not associated with any
4095 message, but which match the given pattern. Implemented by a patch from
4096 Martin Sluka, which also tidied up the Perl a bit.
4098 71. Recipient callout verification, like sender verification, was using <> in
4099 the MAIL FROM command. This isn't really the right thing, since the actual
4100 sender may affect whether the remote host accepts the recipient or not. I
4101 have changed it to use the actual sender in the callout; this means that
4102 the cache record is now keyed on a recipient/sender pair, not just the
4103 recipient address. There doesn't seem to be a real danger of callout loops,
4104 since a callout by the remote host to check the sender would use <>.
4105 [SEE ABOVE: changed after hitting problems.]
4107 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4108 temporary errors. However, in the case of such a code being given after
4109 the end of a data transmission (i.e. after ".") Exim was failing to write
4110 a retry record for the message. (Yes, there was some broken host that was
4111 actually sending 8xx at this point.)
4113 73. An unknown lookup type in a host list could cause Exim to panic-die when
4114 the list was checked. (An example that provoked this was putting <; in the
4115 middle of a list instead of at the start.) If this happened during a DATA
4116 ACL check, a -D file could be left lying around. This kind of configuration
4117 error no longer causes Exim to die; instead it causes a defer errror. The
4118 incident is still logged to the main and panic logs.
4120 74. Buglet left over from Exim 3 conversion. The message "too many messages
4121 in one connection" was written to the rejectlog but not the mainlog, except
4122 when address rewriting (yes!) was being logged.
4124 75. Added write_rejectlog option.
4126 76. When a system filter was run not as root (that is, when system_filter_user
4127 was set), the values of the $n variables were not being returned to the
4128 main process; thus, they were not subsequently available in the $sn
4131 77. Added +return_path_on_delivery log selector.
4133 78. A connection timeout was being treated differently from recipients deferred
4134 when testing hosts_max_try with a message that was older than the host's
4135 retry timeout. (The host should not be counted, thus allowing all hosts to
4136 be tried at least once before bouncing.) This may have been the cause of an
4137 occasionally reported bug whereby a message would remain on the queue
4138 longer than the retry timeout, but would be bounced if a delivery was
4139 forced. I say "may" because I never totally pinned down the problem;
4140 setting up timeout/retry tests is difficult. See also the next item.
4142 79. The ultimate address timeout was not being applied to errors that involved
4143 a combination of host plus message (for example, a timeout on a MAIL
4144 command). When an address resolved to a number of possible hosts, and they
4145 were not all tried for each delivery (e.g. because of hosts_max_try), a
4146 message could remain on the queue longer than the retry timeout.
4148 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4151 81. Fixed an obscure SMTP outgoing bug which required at least the following
4152 conditions: (a) there was another message waiting for the same server;
4153 (b) the server returned 5xx to all RCPT commands in the first message so
4154 that the message was not completed; (c) the server dropped the connection
4155 or gave a negative response to the RSET that Exim sends to abort the
4156 transaction. The observed case was a dropped connection after DATA that had
4157 been sent in pipelining mode. That is, the server had advertised PIPELINING
4158 but was not implementing it correctly. The effect of the bug was incorrect
4159 behaviour, such as trying another host, and this could lead to a crash.
4165 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4166 and daemon.c were passed as pointers to ints; they should have been
4167 pointers to socklen_t variables (which are typically unsigned ints).
4169 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4172 3. Fixed a really odd bug that affected only the testing scheme; patching a
4173 certain fixed string in the binary changed the value of another string that
4174 happened to be identical to the end of the original first string.
4176 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4177 name", it returns that address as the IP address. On some operating
4178 systems (e.g. Solaris), it also passes back the IP address string as the
4179 "host name". However, on others (e.g. Linux), it passes back an empty
4180 string. Exim wasn't checking for this, and was changing the host name to an
4181 empty string, assuming it had been canonicized.
4183 5. Although rare, it is permitted to have more than one PTR record for a given
4184 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4185 all the names associated with an address, because they do in Solaris.
4186 However, it seems that they do not in Linux for data that comes from the
4187 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4188 I found this out when I moved to a new Linux workstation and tried to run
4189 the Exim test suite.
4191 To get round this problem I have changed the code so that it now does its
4192 own call to the DNS to look up PTR records when searching for a host name.
4193 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4194 addresses that are only in /etc/hosts are still found.
4196 This behaviour is, however, controlled by an option called host_lookup_
4197 order, which defaults to "bydns:byaddr". If people want to use the other
4198 order, or indeed, just use one or the other means of lookup, they can
4199 specify it in this variable.
4201 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4202 some operating systems, this comes back from gethostbyaddr() as an empty
4203 string, and this is what Exim used to test for. However, it seems that in
4204 other systems, "." is yielded. Exim now tests for this case too.
4206 7. The values of check_spool_space and check_log_space are now held internally
4207 as a number of kilobytes instead of an absolute number of bytes. If a
4208 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4209 kilobyte. This means that much larger values can be stored.
4211 8. Exim monitor: an attempt to get the action menu when not actually pointing
4212 at a message produces an empty menu entitled "No message selected". This
4213 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4214 no entries in it ("Shell widget menu has zero width and/or height"). So I
4215 have added a single, blank menu entry in this case.
4217 9. Added ${quote_local_part.
4219 10. MIME decoding is now applied to the contents of Subject: header lines when
4222 11. Now that a reference to $sender_host_address automatically causes a reverse
4223 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4224 host lookup before query-style lookups in lists that might use this
4225 variable. This has therefore been abolished, and the "net-" prefix is no
4226 longer necessary for query-style lookups.
4228 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4229 have been a typo for LFLAGS, so it has been changed.
4231 13. The install script calls Exim with "-C /dev/null" in order to find the
4232 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4233 to be output. Howeve, since Exim outputs its version number before the
4234 error, it didn't break the script. It just looked ugly. I fixed this by
4235 always allowing "-C /dev/null" if the caller is root.
4237 14. Ignore overlarge ACL variable number when reading spool file - insurance
4238 against a later release with more variables having written the file.
4240 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4241 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4242 this, as well as the form without the "IPv6" on the front.
4244 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4245 OS/Makefile-Darwin file.
4247 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4248 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4250 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4251 variable. However, some OS use ints for inodes. I've added cast to long int
4252 to get rid of the compiler warning.
4254 19. I had forgotten to lock out "/../" in configuration file names when
4255 ALT_CONFIG_PREFIX was set.
4257 20. Routers used for verification do not need to specify transports. However,
4258 if such a router generated a host list, and callout was configured, Exim
4259 crashed, because it could not find a port number from the (non-existent)
4260 transport. It now assumes port 25 in this circumstance.
4262 21. Added the -t option to exigrep.
4264 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4265 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4266 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4267 them is set, LOOKUP_LSEARCH is forced.
4269 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4271 24. Added sender and host information to the "rejected by local_scan()" log
4272 line; previously there was no indication of these.
4274 25. Added .include_if_exists.
4276 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4277 Linux. It turns out that not all file systems support this. Apparently some
4278 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4279 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4280 not supported on the file descriptor, is now ignored when Exim is trying to
4281 sync a directory. This applies only to Linux.
4283 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4285 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4286 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4289 29. Added support for maildirsize files from supplied patch (modified a bit).
4291 30. The use of :fail: followed by an empty string could lead Exim to respond to
4292 sender verification failures with (e.g.):
4294 550 Verification failed for <xxx>
4295 550 Sender verify failed
4297 where the first response line was missing the '-' that indicates it is not
4298 the final line of the response.
4300 31. The loop for finding the name of the user that called Exim had a hardwired
4301 limit of 10; it now uses the value of finduser_retries, which is used for
4302 all other user lookups.
4304 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4307 33. Exim was neglecting to zero errno before one call of strtol() when
4308 expanding a string and expecting an integer value. On some systems this
4309 resulted in spurious "integer overflow" errors. Also, it was casting the
4310 result into an int without checking.
4312 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4313 did not work. The code looks as if it has *never* worked, though it appears
4314 to have been documented since at least releast 1.62. I have made it work.
4316 35. The "timeout_DNS" error in retry rules, also documented since at least
4317 1.62, also never worked. As it isn't clear exactly what this means, and
4318 clearly it isn't a major issue, I have abolished the feature by treating it
4319 as "timeout", and writing a warning to the main and panic logs.
4321 36. The display of retry rules for -brt wasn't always showing the error code
4324 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4325 timeout_connect_A, timeout_connect_MX.
4327 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4328 to the empty sender.
4330 39. The daemon was not analysing the content of -oX till after it had closed
4331 stderr and disconnected from the controlling terminal. This meant that any
4332 syntax errors were only noted on the panic log, and the return code from
4333 the command was 0. By re-arranging the code a little, I've made the
4334 decoding happen first, so such errors now appear on stderr, and the return
4335 code is 1. However, the actual setting up of the sockets still happens in
4336 the disconnected process, so errors there are still only recorded on the
4339 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4340 connections (as happens on some IP stacks) was logged at start up time as
4341 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4342 it from "IPv6 and IPv4", which means that two separate sockets are being
4345 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4346 says whether AF_INET or AF_INET6 was passed as an argument.
4348 42. Exiwhat output was messed up when time zones were included in log
4351 43. Exiwhat now gives more information about the daemon's listening ports,
4352 and whether -tls-on-connect was used.
4354 44. The "port" option of the smtp transport is now expanded.
4356 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4357 silently ignored. Now an error message is written to the main and panic
4360 46. There's a new ACL modifier called "logwrite" which writes to a log file
4361 as soon as it is encountered.
4363 47. Added $local_user_uid and $local_user_gid at routing time.
4365 48. Exim crashed when trying to verify a sender address that was being
4368 49. Exim was recognizing only a space character after ".include". It now also
4369 recognizes a tab character.
4371 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4372 extracting the relevant information from the specification. The man page no
4373 longer contains scrambled data for the -d option, and I've added a section
4374 at the front about calling Exim under different names.
4376 51. Added "extra_headers" argument to the "mail" command in filter files.
4378 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4381 53. Installed eximstats 1.29.
4383 54. Added transport_filter_timeout as a generic transport option.
4385 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4386 Cc: header lines. This was required by RFC 822, but it not required by RFC
4389 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4390 incoming messages that did not have them. Now it does so only if the
4391 message originates locally, that is, if there is no associated remote host
4392 address. When Resent- header lines are present, this applies to the Resent-
4393 lines rather than the non-Resent- lines.
4395 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4396 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4398 58. Messages for configuration errors now include the name of the main
4399 configuration file - useful now that there may be more than one file in a
4400 list (.included file names were always shown).
4402 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4403 for those rare installations that do not start the daemon as root or run it
4404 setuid root. I've cut out the call to initgroups() if the daemon is not
4407 60. The Exim user and group can now be bound into the binary as text strings
4408 that are looked up at the start of Exim's processing.
4410 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4412 62. Added $mailstore_basename variable.
4414 63. Installed patch to sieve.c from Michael Haardt.
4416 64. When Exim failed to open the panic log after failing to open the main log,
4417 the original message it was trying to log was written to stderr and debug
4418 output, but if they were not available (the usual case in production), it
4419 was lost. Now it is written to syslog before the two lines that record the
4420 failures to open the logs.
4422 65. Users' Exim filters run in subprocesses under the user's uid. It is
4423 possible for a "deliver" command or an alias in a "personal" command to
4424 provoke an address rewrite. If logging of address rewriting is configured,
4425 this fails because the process is not running as root or exim. There may be
4426 a better way of dealing with this, but for the moment (because 4.30 needs
4427 to be released), I have disabled address rewrite logging when running a
4428 filter in a non-root, non-exim process.
4434 1. The buildconfig auxiliary program wasn't quoting the value set for
4435 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4436 not defined. This bug was masked in 4.22 by the effect that was fixed in
4439 2. Some messages that were rejected after a message id was allocated were
4440 shown as "incomplete" by exigrep. It no longer does this for messages that
4441 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4443 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4444 have allow_domain_literals set, the ID did not get logged in the <= line.
4445 Domain literals are now always recognized in Message-ID: header lines.
4447 4. The first argument for a ${extract expansion item is the key name or field
4448 number. Leading and trailing spaces in this item were not being ignored,
4449 causing some misleading effects.
4451 5. When deliver_drop_privilege was set, single queue runner processes started
4452 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4453 same command in the process it spins off) were not dropping privilege.
4455 6. When the daemon running as "exim" started a queue runner, it always
4456 re-executed Exim in the spun-off process. This is a waste of effort when
4457 deliver_drop_privilege is set. The new process now just calls the
4458 queue-runner function directly.
4464 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4467 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4468 ignored. Though the use of -oP was forcing the writing of a pid file, it
4469 was always written to the default place.
4471 3. If the message "no IP address found for host xxxx" is generated during
4472 incoming verification, it is now followed by identification of the incoming
4473 connection (so you can more easily find what provoked it).
4475 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4477 5. Added some features to "harden" Exim a bit more against certain attacks:
4479 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4480 be put in Local/Makefile. This is like the never_users runtime option,
4481 but it cannot be overridden. The default setting is "root".
4483 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4484 prefix string with which any file named in a -C command line option
4487 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4488 is retained for -C and -D only if the caller of Exim is root. Without
4489 it, the exim user may also use -C and -D and retain privilege.
4491 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4492 command line option is disabled.
4494 6. Macro names set by the -D option must start with an upper case letter, just
4495 like macro names defined in the configuration file.
4497 7. Added "dereference=" facility to LDAP.
4499 8. Two instances of the typo "uknown" in the source files are fixed.
4501 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4502 the Configure-Makefile script screwed up while processing it.
4504 10. Incorporated PCRE 4.4.
4506 11. The SMTP synchronization check was not operating right at the start of an
4507 SMTP session. For example, it could not catch a HELO sent before the client
4508 waited for the greeting. There is now a check for outstanding input at the
4509 point when the greeting is written. Because of the duplex, asynchronous
4510 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4511 way, but not yet received, when the check is performed.
4513 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4514 on TCP/IP sockets, because this apparently causes some broken clients to
4517 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4518 unchanged) from the Cygwin maintainer.
4520 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4521 was supported instead of testing for mbx. Effectively a typo.
4523 15. The spa authenticator server code was not checking that the input it
4524 received was valid base64.
4526 16. The debug output line for the "set" modifier in ACLs was not showing the
4527 name of the variable that was being set.
4529 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4531 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4532 lookup on its own. Something else was needed to trigger the lookup. For
4533 example, a match in host_lookup or the need for a host name in a host list.
4534 Now, if $sender_host_name is referenced and the host name has not yet been
4535 looked up, a lookup is performed. If the lookup fails, the variable remains
4536 empty, and $host_lookup_failed is set to "1".
4538 19. Added "eqi" as a case-independent comparison operator.
4540 20. The saslauthd authentication condition could segfault if neither service
4541 nor realm was specified.
4543 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4544 error message that was logged was misleading, and incoming SMTP
4545 connections were dropped. The message is now more accurate, and temporary
4546 errors are given to SMTP connections.
4548 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4549 (except RSET) with a 421 error, until QUIT is received. However, it was
4550 failing to send a response to QUIT.
4552 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4553 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4554 ACL is now run after the helo_try_verify_hosts code.
4556 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4557 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4558 case-independent, so other case variants are also recognized). Apparently
4559 some systems use these upper case variants.
4561 25. If more than two messages were waiting for the same host, and a transport
4562 filter was specified for the transport, Exim sent two messages over the
4563 same TCP/IP connection, and then failed with "socket operation on non-
4564 socket" when it tried to send the third.
4566 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4568 27. The extern definition of crypt16() in expand.c was not being excluded when
4569 the OS had its own crypt16() function.
4571 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4572 as a preferred synonym for return_size_limit, both as an option and as an
4575 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4577 30. Changed the default configuration ACL to relax the local part checking rule
4578 for addresses that are not in any local domains. For these addresses,
4579 slashes and pipe symbols are allowed within local parts, but the sequence
4580 /../ is explicitly forbidden.
4582 31. SPA server authentication was not clearing the challenge buffer before
4585 32. log_message in a "warn" ACL statement was writing to the reject log as
4586 well as to the main log, which contradicts the documentation and doesn't
4587 seem right (because no rejection is happening). So I have stopped it.
4589 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4590 However, I am unable to do any testing of this.
4592 34. Fixed an infelicity in the appendfile transport. When checking directories
4593 for a mailbox, to see if any needed to be created, it was accidentally
4594 using path names with one or more superfluous leading slashes; tracing
4595 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4597 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4598 contents are added to the log line that is written for every discarded
4599 recipient. (Previously a log_message setting was ignored.)
4601 36. The ${quote: operator now quotes the string if it is empty.
4603 37. The install script runs exim in order to find its version number. If for
4604 some reason other than non-existence or emptiness, which it checks, it
4605 could not run './exim', it was installing it with an empty version number,
4606 i.e. as "exim-". This error state is now caught, and the installation is
4609 38. An argument was missing from the function that creates an error message
4610 when Exim fails to connect to the socket for saslauthd authentication.
4611 This could cause Exim to crash, or give a corrupted message.
4613 39. Added isip, isip4, and isip6 to ${if conditions.
4615 40. The ACL variables $acl_xx are now saved with the message, and can be
4616 accessed later in routers, transports, and filters.
4618 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4619 strings in the file are not string-expanded.
4621 42. If a MAIL command specified a SIZE value that was too large to fit into an
4622 int variable, the check against message_size_limit failed. Such values are
4623 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4624 day this will have to be increased, but I don't think I want to be around
4625 when emails are that large.
4632 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4633 iconv() is not standard in FreeBSD.
4635 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4636 IPv6 enabled. The observed symptom was a segmentation fault on return from
4637 the function os_common_find_running_interfaces() in src/os.c.
4639 3. In the check_special_case() function in daemon.c I had used "errno" as an
4640 argument name, which causes warnings on some systems. This was basically a
4641 typo, since it was named "eno" in the comments!
4643 4. The code that waits for the clock to tick (at a resolution of some fraction
4644 of a second) so as to ensure message-id uniqueness was always waiting for
4645 at least one whole tick, when it could have waited for less. [This is
4646 almost certainly not relevant at current processor speeds, where it is
4647 unlikely to ever wait at all. But we try to future-proof.]
4649 5. The function that sleeps for a time interval that includes fractions of a
4650 second contained a race. It did not block SIGALRM between setting the
4651 timer, and suspending (a couple of lines later). If the interval was short
4652 and the sigsuspend() was delayed until after it had expired, the suspension
4653 never ended. On busy systems this could lead to processes getting stuck for
4656 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4657 process, before it forks any delivery processes. The open lookup caching
4658 mechanism meant that the open file or database connection was passed into
4659 the delivery process. The problem was that delivery processes always tidy
4660 up cached lookup data. This could cause a problem for the next delivery
4661 process started by the queue runner, because the external queue runner
4662 process does not know about the closure. So the next delivery process
4663 still has data in the lookup cache. In the case of a file lookup, there was
4664 no problem because closing a file descriptor in a subprocess doesn't affect
4665 the parent. However, if the lookup was caching a connection to a database,
4666 the connection was closed, and the second delivery process was likely to
4667 see errors such as "PGSQL: query failed: server closed the connection
4668 unexpectedly". The problem has been fixed by closing all cached lookups
4669 in a queue runner before running a delivery process.
4671 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4672 seem to have the "const" qualifier which it has on other OS. I've
4675 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4676 *of the same type* (client or server) with the same public name that an
4677 error should be diagnosed.
4679 9. When Exim looked up a host name for an IP address, but failed to find the
4680 original IP address when looking up the host name (a safety check), it
4681 output the message "<ip address> does not match any IP for NULL", which was
4682 confusing, to say the least. The bug was that the host name should have
4683 appeared instead of "NULL".
4685 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4686 user that is built into the binary, and the -C or -D options is used, root
4687 privilege is dropped before the configuration file is read. In addition,
4688 logging is switched to stderr instead of the normal log files. If the
4689 configuration then re-defines the Exim user, the unprivileged environment
4690 is probably not what is expected, so Exim logs a panic warning message (but
4693 However, if deliver_drop_privilege is set, the unprivileged state may well
4694 be exactly what is intended, so the warning has been cut out in that case,
4695 and Exim is allowed to try to write to its normal log files.
4701 1. smtp_return_error_details was not giving details for temporary sender
4702 or receiver verification errors.
4704 2. Diagnose a configuration error if two authenticators have the same public
4707 3. Exim used not to create the message log file for a message until the first
4708 delivery attempt. This could be confusing when incoming messages were held
4709 for policy or load reasons. The message log file is now created at the time
4710 the message is received, and an initial "Received" line is written to it.
4712 4. The automatically generated man page for command line options had a minor
4713 bug that caused no ill effects; however, a more serious problem was that
4714 the procedure for building the man page automatically didn't always
4715 operate. Consequently, release 4.20 contains an out-of-date version. This
4716 shouldn't happen again.
4718 5. When building Exim with embedded Perl support, the script that builds the
4719 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4720 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4722 6. The freeze_tell option was not being used for messages that were frozen on
4723 arrival, either by an ACL or by local_scan().
4725 7. Added the smtp_incomplete_transaction log selector.
4727 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4728 was accepting AUTH without a new EHLO.
4730 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4731 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4734 10. Logging of TCP/IP connections (when configured) now happens in the main
4735 daemon process instead of the child process, so that the TCP/IP connection
4736 count is more accurate (but it can never be perfect).
4738 11. The use of "drop" in a nested ACL was not being handled correctly in the
4739 outer ACL. Now, if condition failure induced by the nested "drop" causes
4740 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4741 or "require"), the connection is dropped.
4743 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4744 that yield "discard" can now be used with an "accept" or a "discard" verb,
4745 but an error is generated for any others (because I can't see a useful way
4746 to define what should happen).
4748 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4749 are now processed in the same way as lines in the Exim configuration file.
4750 In particular, continuation lines are supported.
4752 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4754 15. Added -ti meaning -t -i.
4756 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4757 domains, and warn by logging if others are found.
4759 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4760 returned from GIFCONF when Exim is trying to find the list of interfaces on
4761 a host. The code in os.c has been modified to copy each ifreq to an aligned
4762 structure in all cases.
4764 Also, in some cases, the returned ifreq's were being copied to a 'struct
4765 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4766 means the last couple of bytes of an IPv6 address could be chopped if the
4767 ifreq contained only a normal sockaddr (14 bytes storage).
4769 18. Named domain lists were not supported in the hosts_treat_as_local option.
4770 An entry such as +xxxx was not recognized, and was treated as a literal
4773 19. Ensure that header lines added by a DATA ACL are included in the reject log
4774 if the ACL subsequently rejects the message.
4776 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4777 MD5 (which is deprecated).
4779 21. When testing a filter file using -bf, Exim was writing a message when it
4780 took the sender from a "From " line in the message, but it was not doing so
4781 when it took $return_path from a Return-Path: header line. It now does.
4783 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4784 with a valid header line field name (a series of printing characters
4785 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4787 23. Changed "disc" in the source to "disk" to conform to the documentation and
4788 the book and for uniformity.
4790 24. Ignore Sendmail's -Ooption=value command line item.
4792 25. When execve() failed while trying to run a command in a pipe transport,
4793 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4794 could be confused with a return value of 69 from the command itself. This
4795 has been changed to 127, the value the shell returns if it is asked to run
4796 a non-existent command. The wording for the related log line suggests a
4797 non-existent command as the problem.
4799 26. If received_header_text expands to an empty string, do not add a Received:
4800 header line to the message. (Well, it adds a token one on the spool, but
4801 marks it "old" so that it doesn't get used or transmitted.)
4803 27. Installed eximstats 1.28 (addition of -nt option).
4805 28. There was no check for failure on the call to getsockname() in the daemon
4806 code. This can fail if there is a shortage of resources on the system, with
4807 ENOMEM, for example. A temporary error is now given on failure.
4809 29. Contrary to the C standard, it seems that in some environments, the
4810 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4811 program. Exim now does this explicitly; it affects the formatting of
4812 timestamps using strftime().
4814 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4815 complaints. I've now initialized all the variables, to avoid this.
4817 32. Header lines added by a system filter were not being "seen" during
4818 transport-time rewrites.
4820 33. The info_callback() function passed to OpenSSL is set up with type void
4821 (*)(SSL *, int, int), as described somewhere. However, when calling the
4822 function (actually a macro) that sets it up, the type void(*)() is
4823 expected. I've put in a cast to prevent warnings from picky compilers.
4825 34. If a DNS black list lookup found a CNAME record, but there were no A
4826 records associated with the domain it pointed at, Exim crashed.
4828 35. If a DNS black list lookup returned more than one A record, Exim ignored
4829 all but the first. It now scans all returned addresses if a particular IP
4830 value is being sought. In this situation, the contents of the
4831 $dnslist_value variable are a list of all the addresses, separated by a
4834 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4835 to accept a host name and all its aliases if the forward lookup for any of
4836 them yielded the IP address of the incoming connection. Now it accepts only
4837 those names whose forward lookup yields the correct IP address. Any other
4838 names are discarded. This closes a loophole whereby a rogue DNS
4839 administrator could create reverse DNS records to break through a
4840 wildcarded host restriction in an ACL.
4842 37. If a user filter or a system filter that ran in a subprocess used any of
4843 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4844 the wrong values were passed to the pipe command ($thisaddress had the
4845 value of $0, $0 had the value of $1, etc). This bug was introduced by
4846 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4848 38. Improved the line breaking for long SMTP error messages from ACLs.
4849 Previously, if there was no break point between 40 and 75 characters, Exim
4850 left the rest of the message alone. Two changes have been made: (a) I've
4851 reduced the minimum length to 35 characters; (b) if it can't find a break
4852 point between 35 and 75 characters, it looks ahead and uses the first one
4853 that it finds. This may give the occasional overlong line, but at least the
4854 remaining text gets split now.
4856 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4857 file descriptors might be low, and that setting 1000 would always raise it.
4858 It turns out that in some environments, the limit is already over 1000 and
4859 that lowering it causes trouble. So now Exim takes care not to decrease it.
4861 40. When delivering a message, the value of $return_path is set to $sender_
4862 address at the start of routing (routers may change the value). By an
4863 oversight, this default was not being set up when an address was tested by
4864 -bt or -bv, which affected the outcome if any router or filter referred to
4867 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4868 log only when "message" or "log_message" are set. However, if one of the
4869 conditions was an address verification, or a call to a nested ACL, the
4870 messages generated by the underlying test were being passed through. This
4871 no longer happens. The underlying message is available in $acl_verify_
4872 message for both "message" and "log_message" expansions, so it can be
4873 passed through if needed.
4875 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4876 new expansion $bh_ to give the encoded byte string without charset
4877 translation. Translation happens only if iconv() is available; HAVE_ICONV
4878 indicates this at build time. HEADERS_CHARSET gives the charset to
4879 translate to; headers_charset can change it in the configuration, and
4880 "headers charset" can change it in an individual filter file.
4882 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4883 that creates RFC 2047 encoded "words" labels them as that charset instead
4884 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4885 expansion operator; (ii) when Exim creates a From: line for a local
4886 message; (iii) when a header line is rewritten to include a "phrase" part.
4888 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4889 buggy, causing it to skip the first lines of messages whose message ID
4890 ended in 'D'. This would not have bitten before Exim release 4.14, because
4891 message IDs were unlikely to end in 'D' before then. The effect was to have
4892 incorrect size information for certain domains.
4894 45. #include "config.h" was missing at the start of the crypt16.c module. This
4895 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4898 46. If there was a timeout during a "random" callout check, Exim treated it as
4899 a failure of the random address, and carried on sending RSET and the real
4900 address. If the delay was just some slowness somewhere, the response to the
4901 original RCPT would be taken as a response to RSET and so on, causing
4902 mayhem of various kinds.
4904 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4905 when I implemented it. It didn't allow for the fact that some option values
4906 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4907 the right test for positive values.
4909 48. Domain names in DNS records are case-independent. Exim always looks them up
4910 in lower case. Some resolvers return domain names in exactly the case they
4911 appear in the zone file, that is, they may contain uppercase letters. Not
4912 all resolvers do this - some return always lower case. Exim was treating a
4913 change of case by a resolver as a change of domain, similar to a widening
4914 of a domain abbreviation. This triggered its re-routing code and so it was
4915 trying to route what was effectively the same domain again. This normally
4916 caused routing to fail (because the router wouldn't handle the domain
4917 twice). Now Exim checks for this case specially, and just changes the
4918 casing of the domain that it ultimately uses when it transmits the message
4921 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4924 50. If a filter generated a file delivery with a non-absolute name (possible if
4925 no home directory exists for the router), the forbid_file option was not
4928 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4929 the existing equality matching.
4931 52. Exim was using ints instead of ino_t variables in some places where it was
4932 dealing with inode numbers.
4934 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4935 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4936 called TMPDIR, and if it finds it is different, it changes its value.
4938 54. The smtp_printf() function is now made available to local_scan() so
4939 additional output lines can be written before returning. There is also an
4940 smtp_fflush() function to enable the detection of a dropped connection.
4941 The variables smtp_input and smtp_batched_input are exported to
4944 55. Changed the default runtime configuration: the message "Unknown user"
4945 has been removed from the ACL, and instead placed on the localuser router,
4946 using the cannot_route_message feature. This means that any verification
4947 failures that generate their own messages won't get overridden. Similarly,
4948 the "Unrouteable address" message that was in the ACL for unverifiable
4949 relay addresses has also been removed.
4951 56. Added hosts_avoid_esmtp to the smtp transport.
4953 57. The exicyclog script was not checking for the esoteric option
4954 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4955 will work only if exicyclog is run under the appropriate euid.
4957 58. Following a discussion on the list, the rules by which Exim recognises line
4958 endings on incoming messages have been changed. The -dropcr and drop_cr
4959 options are now no-ops, retained only for backwards compatibility. The
4960 following line terminators are recognized: LF CRLF CR. However, special
4961 processing applies to CR:
4963 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4964 nor a local message in the state where . is a terminator.
4966 (ii) If a bare CR is encountered in a header line, an extra space is added
4967 after the line terminator so as not to end the header. The reasoning
4968 behind this is that bare CRs in header lines are most likely either
4969 to be mistakes, or people trying to play silly games.
4971 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4972 was being incorrectly given as 18 bytes larger than it should have been.
4973 This is a VOB (very old bug).
4975 60. This may never have affected anything current, but just in case it has:
4976 When the local host is found other than at the start of a list of hosts,
4977 the local host, those with the same MX, and any that follow, are discarded.
4978 When the list in question was part of a longer list of hosts, the following
4979 hosts (not currently being processed) were also being discarded. This no
4980 longer happens. I'm not sure if this situation could ever has previously
4983 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4986 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4987 header. This is something that is recommended in a new Internet Draft, and
4988 is something that is documented as being done by Sendmail. There are two
4989 possible values. For messages generated by the autoreply transport, Exim
4992 Auto-Submitted: auto-replied
4994 whereas for all other generated messages (e.g. bounces) it adds
4996 Auto-Submitted: auto-generated
4998 63. The "personal" condition in filters now includes a test for the
4999 Auto-Submitted: header. If it contains the string "auto-" the message it
5000 not considered personal.
5002 64. Added rcpt_include_affixes as a generic transport option.
5004 65. Added queue_only_override (default true).
5006 66. Added the syslog_duplication option.
5008 67. If what should have been the first header line of a message consisted of
5009 a space followed by a colon, Exim was mis-interpreting it as a header line.
5010 It isn't of course - it is syntactically invalid and should therefore be
5011 treated as the start of the message body. The misbehaviour could have
5012 caused a number of strange effects, including loss of data in subsequent
5013 header lines, and spool format errors.
5015 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5016 client host had authenticated. This control can now be exercised by an ACL
5017 for more flexibility.
5019 69. By default, callouts do not happen when testing with -bh. There is now a
5020 variant, -bhc, which does actually run the callout code, including
5021 consulting and updating the callout cache.
5023 70. Added support for saslauthd authentication, courtesy of Alexander
5026 71. If statvfs() failed on the spool or log directories while checking their
5027 size for availability, Exim confusingly gave the error "space shortage".
5028 Furthermore, in debugging mode it crashed with a floating point exception.
5029 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5030 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5031 serious problem, Exim now writes to the main and panic logs when this
5032 happens, with details of the failure. It then refuses to accept the
5033 incoming message, giving the message "spool directory problem" or "log
5034 directory problem" with a 421 code for SMTP messages.
5036 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5037 0, 1, and 2 exist, because some OS complain for execs without them (see
5038 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5039 descriptors. However, the code omitted to check that the open succeeded,
5040 causing mysterious errors if for some reason the permissions on /dev/null
5041 got screwed. Now Exim writes a message to the main and panic logs, and
5042 bombs out if it can't open /dev/null.
5044 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5045 interact so that it is all more flexible. It is supposed to remain
5046 backwards compatible. Also added extra_local_interfaces.
5048 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5049 to bomb out with an assertion failure - to the client this appears as a
5050 connection drop. This problem occurs in the part of the code that was taken
5051 from the Samba project. Fortunately, the assertion is in a very simple
5052 function, so I have fixed this by reproducing the function inline in the
5053 one place where it is called, and arranging for authentication to fail
5054 instead of killing the process with assert().
5056 75. The SPA client code was not working when the server requested OEM rather
5057 than Unicode encoding.
5059 76. Added code to make require_files with a specific uid setting more usable in
5060 the case where statting the file as root fails - usually a non-root-mounted
5061 NFS file system. When this happens and the failure is EACCES, Exim now
5062 forks a subprocess and does the per-uid checking as the relevant uid.
5064 77. Added process_log_path.
5066 78. If log_file_path was not explicitly set, a setting of check_log_space or
5067 check_log_inodes was ignored.
5069 79. If a space check for the spool or log partitions fails, the incident is now
5070 logged. Of course, in the latter case the data may get lost...
5072 80. Added the %p formatting code to string_format() so that it can be used to
5073 print addresses in debug_print(). Adjusted all the address printing in the
5074 debugging in store.c to use %p rather than %d.
5076 81. There was a concern that a line of code in smtp_in.c could overflow a
5077 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5078 initially expressed, the concern was not well-founded, because trailing
5079 spaces are removed early. However, if the trailing spaces were followed by
5080 a NULL, they did not get removed, so the overflow was possible. Two fixes
5083 (a) I re-wrote the offending code in a cleaner fashion.
5084 (b) If an incoming SMTP command contains a NULL character, it is rejected
5087 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5088 runs initgroups(), so that if the Exim user is in any additional groups,
5089 they will be used during message reception.
5095 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob