1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.611 2010/05/29 19:23:26 nm4 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 JJ/01 installed exipick 20100104.1, adding $max_received_linelength, $data_path, and $header_path variables; fixed documentation bugs and typos
11 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow exipick to access non-standard spools, including the "frozen" queue (Finput)
13 NM/01 Bugzilla 965: Support mysql stored procedures.
14 Patch from Alain Williams
16 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
18 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
19 Patch from Andreas Metzler
21 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
22 Patch from Kirill Miazine
24 NM/05 Bugzilla 671: Added umask to procmail example.
26 JJ/03 installed exipick 20100323.0, fixing doc bug
28 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
29 directory. Notification and patch from Dan Rosenberg
31 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
33 TK/02 Improve log output when DKIM signing operation fails.
35 MH/01 Treat the transport option dkim_domain as a colon separated
36 list, not as a single string, and sign the message with each element,
37 omitting multiple occurences of the same signer.
39 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
40 Bugzilla 985, 986. Patch by Simon Arlott
42 NM/08 Bugzilla 967. DKIM DNS TXT record bug fix
49 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
51 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
53 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
55 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
57 NM/04 Bugzilla 915: Flag broken perl installation during build.
63 TK/01 Added patch by Johannes Berg that expands the main option
64 "spamd_address" if it starts with a dollar sign.
66 TK/02 Write list of recipients to X-Envelope-Sender header when building
67 the mbox-format spool file for content scanning (suggested by Jakob
70 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
71 (http://www.dcc-servers.net/) support via dccifd. Activated by
72 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
74 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
75 by Mark Daniel Reidel <mr@df.eu>.
77 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
78 When building exim an external PCRE library is now needed -
79 PCRE is a system library on the majority of modern systems.
80 See entry on PCRE_LIBS in EDITME file.
82 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
83 conversation. Added nologin parameter to request.
84 Patch contributed by Kirill Miazine
86 TF/01 Do not log submission mode rewrites if they do not change the address.
88 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
90 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
91 log files in place. Contributed by Roberto Lima
93 NM/04 Bugzilla 667: close socket used by dovecot authenticator
95 TF/03 Bugzilla 615: When checking the local_parts router precondition
96 after a local_part_suffix or local_part_prefix option, Exim now
97 does not use the address's named list lookup cache, since this
98 contains cached lookups for the whole local part.
100 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
101 Robert Millan. Documentation is in experimental-spec.txt
103 TF/04 Bugzilla 668: Fix parallel build (make -j).
105 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
107 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
108 Patch provided by Jan Srzednicki
110 TF/05 Leading white space used to be stripped from $spam_report which
111 wrecked the formatting. Now it is preserved.
113 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
114 that they are available at delivery time.
116 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
118 TF/08 TLS error reporting now respects the incoming_interface and
119 incoming_port log selectors.
121 TF/09 Produce a more useful error message if an SMTP transport's hosts
122 setting expands to an empty string.
124 NM/06 Bugzilla 744: EXPN did not work under TLS.
125 Patch provided by Phil Pennock
127 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
128 Patch provided by Richard Godbee
130 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
131 acl_smtp_notquit, added index entry.
133 NM/09 Bugzilla 787: Potential buffer overflow in string_format
134 Patch provided by Eugene Bujak
136 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
137 Patch provided by Maxim Dounin
139 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
140 Patch provided by Phil Pennock
142 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
144 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
145 Patch provided by Brad "anomie" Jorsch
147 NM/14 Bugzilla 622: Added timeout setting to transport filter.
148 Patch provided by Dean Brooks
150 TK/05 Add native DKIM support (does not depend on external libraries).
152 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
153 Patch provided by Graeme Fowler
155 NM/16 Bugzilla 851: Documentation example syntax fix.
157 NM/17 Changed NOTICE file to remove references to embedded PCRE.
159 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
161 NM/19 Bugzilla 745: TLS version reporting
162 Patch provided by Phil Pennock
164 NM/20 Bugzilla 167: bool: condition support
165 Patch provided by Phil Pennock
167 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
168 Patch provided by Phil Pennock
170 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
171 Patch provided by Brad "anomie" Jorsch
173 NM/23 Bugzilla 687: Fix misparses in eximstats
174 Patch provided by Heiko Schlittermann
176 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
177 Patch provided by Heiko Schlittermann
179 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
180 plus update to original patch.
182 NM/26 Bugzilla 799: Documentation correction for ratelimit
184 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
185 Patch provided by David Brownlee
187 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
189 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
191 NM/30 Bugzilla 888: TLS documentation bugfixes
193 NM/31 Bugzilla 896: Dovecot buffer overrun fix
195 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
196 Unlike the original bugzilla I have changed all shell scripts in src tree
198 NM/33 Bugzilla 898: Transport filter timeout fix
199 Patch by Todd Rinaldo
201 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
202 Patch by Serge Demonchaux
204 NM/35 Bugzilla 39: Base64 decode bug fixes
205 Patch by Jakob Hirsch
207 NM/36 Bugzilla 909: Correct connect() call in dcc code
209 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
211 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
213 NM/39 Bugzilla 911: Fixed MakeLinks build script
219 TK/01 Add preliminary DKIM support. Currently requires a forked version of
220 ALT-N's libdkim that I have put here:
221 http://duncanthrax.net/exim-experimental/
223 Note to Michael Haardt: I had to rename some vars in sieve.c. They
224 were called 'true' and it seems that C99 defines that as a reserved
225 keyword to be used with 'bool' variable types. That means you could
226 not include C99-style headers which use bools without triggering
227 build errors in sieve.c.
229 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
230 as mailq or other aliases. Changed the --help handling significantly
231 to do whats expected. exim_usage() emits usage/help information.
233 SC/01 Added the -bylocaldomain option to eximstats.
235 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
237 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
239 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
245 PH/01 Another patch from the Sieve maintainer.
247 PH/02 When an IPv6 address is converted to a string for single-key lookup
248 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
249 dots are used instead of colons so that keys in lsearch files need not
250 contain colons. This was done some time before quoting was made available
251 in lsearch files. However, iplsearch files do require colons in IPv6 keys
252 (notated using the quote facility) so as to distinguish them from IPv4
253 keys. This meant that lookups for IP addresses in host lists did not work
254 for iplsearch lookups.
256 This has been fixed by arranging for IPv6 addresses to be expressed with
257 colons if the lookup type is iplsearch. This is not incompatible, because
258 previously such lookups could never work.
260 The situation is now rather anomolous, since one *can* have colons in
261 ordinary lsearch keys. However, making the change in all cases is
262 incompatible and would probably break a number of configurations.
264 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
267 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
268 conversion specification without a maximum field width, thereby enabling
269 a rogue spamd server to cause a buffer overflow. While nobody in their
270 right mind would setup Exim to query an untrusted spamd server, an
271 attacker that gains access to a server running spamd could potentially
272 exploit this vulnerability to run arbitrary code as the Exim user.
274 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
275 $primary_hostname instead of what libspf2 thinks the hosts name is.
277 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
278 a directory entry by the name of the lookup key. Previously, if a
279 symlink pointed to a non-existing file or a file in a directory that
280 Exim lacked permissions to read, a lookup for a key matching that
281 symlink would fail. Now it is enough that a matching directory entry
282 exists, symlink or not. (Bugzilla 503.)
284 PH/03 The body_linecount and body_zerocount variables are now exported in the
287 PH/04 Added the $dnslist_matched variable.
289 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
290 This means they are set thereafter only if the connection becomes
293 PH/06 Added the client_condition to authenticators so that some can be skipped
294 by clients under certain conditions.
296 PH/07 The error message for a badly-placed control=no_multiline_responses left
297 "_responses" off the end of the name.
299 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
301 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
302 (without spaces) instead of just copying the configuration text.
304 PH/10 Added the /noupdate option to the ratelimit ACL condition.
306 PH/11 Added $max_received_linelength.
308 PH/12 Added +ignore_defer and +include_defer to host lists.
310 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
311 way in which PCRE > 7.0 is built.
313 PH/14 Implemented queue_only_load_latch.
315 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
316 MAIL command. The effect was to mangle the value on 64-bit systems.
318 PH/16 Another patch from the Sieve maintainer.
320 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
322 PH/18 If a system quota error occurred while trying to create the file for
323 a maildir delivery, the message "Mailbox is full" was not appended to the
324 bounce if the delivery eventually timed out. Change 4.67/27 below applied
325 only to a quota excession during the actual writing of the file.
327 PH/19 It seems that peer DN values may contain newlines (and other non-printing
328 characters?) which causes problems in log lines. The DN values are now
329 passed through string_printing() before being added to log lines.
331 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
332 and InterBase are left for another time.)
334 PH/21 Added message_body_newlines option.
336 PH/22 Guard against possible overflow in moan_check_errorcopy().
338 PH/23 POSIX allows open() to be a macro; guard against that.
340 PH/24 If the recipient of an error message contained an @ in the local part
341 (suitably quoted, of course), incorrect values were put in $domain and
342 $local_part during the evaluation of errors_copy.
348 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
349 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
352 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
353 issue a MAIL command.
355 PH/02 In an ACL statement such as
357 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
359 if a client was not listed at all, or was listed with a value other than
360 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
361 the condition was not true (as it should be), so access was not denied.
362 The bug was that the ! inversion was incorrectly passed on to the second
363 item. This has been fixed.
365 PH/03 Added additional dnslists conditions == and =& which are different from
366 = and & when the dns lookup returns more than one IP address.
368 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
369 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
371 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
372 FSYNC, which compiles an option called disable_fsync that allows for
373 bypassing fsync(). The documentation is heavily laced with warnings.
375 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
377 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
378 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
379 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
380 including adding "make clean"; (3) Added -fPIC when compiling the test
381 dynamically loaded module, to get rid of a warning.
383 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
384 message fails, move_frozen_messages = true and ignore_bounce_errors_after
385 = 0s. The bug is otherwise harmless.
387 PH/07 There was a bug in the dovecot authenticator such that the value of
388 $auth1 could be overwritten, and so not correctly preserved, after a
389 successful authentication. This usually meant that the value preserved by
390 the server_setid option was incorrect.
392 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
394 PH/09 Installed PCRE release 7.0.
396 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
397 run for batched SMTP input. It is now run at the start of every message
398 in the batch. While fixing this I discovered that the process information
399 (output by running exiwhat) was not always getting set for -bs and -bS
400 input. This is fixed, and it now also says "batched" for BSMTP.
402 PH/11 Added control=no_pipelining.
404 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
405 patch, slightly modified), and move the expansion of helo_data till after
406 the connection is made in the smtp transport (so it can use these
409 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
411 PH/14 Added log_selector = +pid.
413 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
415 PH/16 Add ${if forany and ${if forall.
417 PH/17 Added dsn_from option to vary the From: line in DSNs.
419 PH/18 Flush SMTP output before performing a callout, unless control =
420 no_callout_flush is set.
422 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
423 was true (the default) a successful delivery failed to delete the retry
424 item, thus causing premature timeout of the address. The bug is now
427 PH/20 Added hosts_avoid_pipelining to the smtp transport.
429 PH/21 Long custom messages for fakedefer and fakereject are now split up
430 into multiline reponses in the same way that messages for "deny" and
431 other ACL rejections are.
433 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
434 with slight modification.
436 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
437 draft, changing the syntax and factoring some duplicate code".
439 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
440 for deliveries of the second and subsequent messages over the same SMTP
443 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
444 ${reduce, with only minor "tidies".
446 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
448 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
449 expansion side effects.
451 PH/27 When a message times out after an over-quota error from an Exim-imposed
452 quota, the bounce message says "mailbox is full". This message was not
453 being given when it was a system quota that was exceeded. It now should
456 MH/03 Made $recipients available in local_scan(). local_scan() already has
457 better access to the recipient list through recipients_list[], but
458 $recipients can be useful in postmaster-provided expansion strings.
460 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
461 in the case of a MAIL command with additional options following the
462 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
463 were accidentally chopped off.
465 PH/29 SMTP synchronization checks are implemented when a command is read -
466 there is a check that no more input is waiting when there shouldn't be
467 any. However, for some commands, a delay in an ACL can mean that it is
468 some time before the response is written. In this time, more input might
469 arrive, invalidly. So now there are extra checks after an ACL has run for
470 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
471 pipelining has not been advertised.
473 PH/30 MH's patch to allow iscntrl() characters to be list separators.
475 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
476 returned in the SMTP response when smtp_return_error_details was false.
479 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
480 instead of the C I/O that was originally supplied, because problems were
483 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
484 Exim which did not show up earlier: it was assuming that a call to
485 SSL_CTX_set_info_callback() might give an error value. In fact, there is
486 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
487 was a macro that became an assignment, so it seemed to work. This has
488 changed to a proper function call with a void return, hence the compile
489 error. Exim's code has been fixed.
491 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
494 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
496 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
498 PH/37 If a message is not accepted after it has had an id assigned (e.g.
499 because it turns out to be too big or there is a timeout) there is no
500 "Completed" line in the log. When some messages of this type were
501 selected by exigrep, they were listed as "not completed". Others were
502 picked up by some special patterns. I have improved the selection
503 criteria to be more general.
505 PH/38 The host_find_failed option in the manualroute router can now be set
506 to "ignore", to completely ignore a host whose IP address cannot be
507 found. If all hosts are ignored, the behaviour is controlled by the new
508 host_all_ignored option.
510 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
511 homing or because of multiple MX records with /mx) generated more than
512 one IP address, and the following item turned out to be the local host,
513 all the secondary addresses of the first item were incorrectly removed
514 from the list, along with the local host and any following hosts (which
515 is what is supposed to happen).
517 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
518 whoever called Exim into the -H file. In the case of the daemon it was
519 behaving confusingly. When first started, it used values for whoever
520 started the daemon, but after a SIGHUP it used the Exim user (because it
521 calls itself on a restart). I have changed the code so that it now always
524 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
525 message are rejected with the same error (e.g. no authentication or bad
526 sender address), and a DATA command is nevertheless sent (as can happen
527 with PIPELINING or a stupid MUA), the error message that was given to the
528 RCPT commands is included in the rejection of the DATA command. This is
529 intended to be helpful for MUAs that show only the final error to their
532 PH/42 Another patch from the Sieve maintainer.
534 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
535 Eximstats - Fixed some broken HTML links and added missing column headers
537 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
538 columns for Rejects, Temp Rejects, Ham, and Spam rows.
540 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
542 PH/43 Yet another patch from the Sieve maintainer.
544 PH/44 I found a way to check for a TCP/IP connection going away before sending
545 the response to the final '.' that terminates a message, but only in the
546 case where the client has not sent further data following the '.'
547 (unfortunately, this is allowed). However, in many cases there won't be
548 any further data because there won't be any more messages to send. A call
549 to select() can be used: if it shows that the input is "ready", there is
550 either input waiting, or the socket has been closed. An attempt to read
551 the next input character can distinguish the two cases. Previously, Exim
552 would have sent an OK response which the client would never have see.
553 This could lead to message repetition. This fix should cure that, at
554 least in a lot of common cases.
556 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
557 advertised in response to EHLO.
563 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
564 fixed by 4.65/MH/01 (is this a record?) are fixed:
566 (i) An empty string was always treated as zero by the numeric comparison
567 operators. This behaviour has been restored.
569 (ii) It is documented that the numeric comparison operators always treat
570 their arguments as decimal numbers. This was broken in that numbers
571 starting with 0 were being interpreted as octal.
573 While fixing these problems I realized that there was another issue that
574 hadn't been noticed. Values of message_size_limit (both the global option
575 and the transport option) were treated as octal if they started with 0.
576 The documentation was vague. These values are now always treated as
577 decimal, and I will make that clear in the documentation.
583 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
584 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
587 MH/01 Don't check that the operands of numeric comparison operators are
588 integers when their expansion is in "skipping" mode (fixes bug
589 introduced by 4.64-PH/07).
591 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
592 child addresses, Exim now panics and dies. Previously, because the count
593 is held in a short int, deliveries were likely to be lost. As such a
594 large number of recipients for a single message is ridiculous
595 (performance will be very, very poor), I have chosen to impose a limit
596 rather than extend the field.
602 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
603 leftover -K file (the existence of which was triggered by #402).
604 While we were at it, introduced process PID as part of the -K
605 filename. This should rule out race conditions when creating
608 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
609 processing considerably. Previous code took too long for large mails,
610 triggering a timeout which in turn triggers #401.
612 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
613 in the DK code in transports.c. sendfile() is not really portable,
614 hence the _LINUX specificness.
616 TF/01 In the add_headers option to the mail command in an Exim filter,
617 there was a bug that Exim would claim a syntax error in any
618 header after the first one which had an odd number of characters
621 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
622 callout verification, Exim cached a "reject" for the entire domain. This
623 is correct for most verifications, but it is not correct for a recipient
624 verification with use_sender or use_postmaster set, because in that case
625 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
626 case of MAIL FROM:<> rejection from other early rejections (e.g.
627 rejection of HELO). When verifying a recipient using a non-null MAIL
628 address, the cache is ignored if it shows MAIL FROM:<> rejection.
629 Whatever the result of the callout, the value of the domain cache is
630 left unchanged (for any other kind of callout, getting as far as trying
631 RCPT means that the domain itself is ok).
633 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
636 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
637 manifest itself as EPIPE rather than ECONNECT. When tidying away a
638 session, the daemon ignores ECONNECT errors and logs others; it now
639 ignores EPIPE as well.
641 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
642 (quoted-printable decoding).
644 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
645 later the small subsequent patch to fix an introduced bug.
647 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
649 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
651 PH/08 An error is now given if message_size_limit is specified negative.
653 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
654 to be given (somewhat) arbitrary names.
656 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
659 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
660 miscellaneous code fixes
662 PH/10 Added the log_reject_target ACL modifier to specify where to log
665 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
666 hostname. This is wrong, because it relates to the incoming message (and
667 probably the interface on which it is arriving) and not to the outgoing
668 callout (which could be using a different interface). This has been
669 changed to use the value of the helo_data option from the smtp transport
670 instead - this is what is used when a message is actually being sent. If
671 there is no remote transport (possible with a router that sets up host
672 addresses), $smtp_active_hostname is used.
674 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
675 tweaks were necessary in order to get it to work (see also 21 below):
676 (a) The code assumed that strncpy() returns a negative number on buffer
677 overflow, which isn't the case. Replaced with Exim's string_format()
679 (b) There were several signed/unsigned issues. I just did the minimum
680 hacking in of casts. There is scope for a larger refactoring.
681 (c) The code used strcasecmp() which is not a standard C function.
682 Replaced with Exim's strcmpic() function.
683 (d) The code set only $1; it now sets $auth1 as well.
684 (e) A simple test gave the error "authentication client didn't specify
685 service in request". It would seem that Dovecot has changed its
686 interface. Fortunately there's a specification; I followed it and
687 changed what the client sends and it appears to be working now.
689 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
692 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
693 address A is aliased to B and C, where B exists and C does not. Without
694 -v the output is "A verified" because verification stops after a
695 successful redirection if more than one address is generated. However,
696 with -v the child addresses are also verified. Exim was outputting "A
697 failed to verify" and then showing the successful verification for C,
698 with its parentage. It now outputs "B failed to verify", showing B's
699 parentage before showing the successful verification of C.
701 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
702 look up a TXT record in a specific list after matching in a combined
705 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
706 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
707 they consult the DNS. I had assumed they would set it the way they
708 wanted; and indeed my experiments on Linux seem to show that in some
709 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
710 To be on the safe side, however, I have now made the interface to
711 host_find_byname() similar to host_find_bydns(), with an argument
712 containing the DNS resolver options. The host_find_byname() function now
713 sets these options at its start, just as host_find_bydns() does. The smtp
714 transport options dns_qualify_single and dns_search_parents are passed to
715 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
716 of host_find_byname() use the default settings of RES_DEFNAMES
717 (qualify_single) but not RES_DNSRCH (search_parents).
719 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
720 spool_read_header() do less string testing, by means of a preliminary
721 switch on the second character of optional "-foo" lines. (This is
722 overdue, caused by the large number of possibilities that now exist.
723 Originally there were few.) While I was there, I also converted the
724 str(n)cmp tests so they don't re-test the leading "-" and the first
725 character, in the hope this might squeeze out yet more improvement.
727 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
728 flag allowing group syntax was set by the header_syntax check but not
729 turned off, possible causing trouble later; (2) The flag was not being
730 set at all for the header_verify test, causing "group"-style headers to
731 be rejected. I have now set it in this case, and also caused header_
732 verify to ignore an empty address taken from a group. While doing this, I
733 came across some other cases where the code for allowing group syntax
734 while scanning a header line wasn't quite right (mostly, not resetting
735 the flag correctly in the right place). These bugs could have caused
736 trouble for malformed header lines. I hope it is now all correct.
738 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
739 with the "reply" argument non-NULL. The code, however (which originally
740 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
741 but it didn't always do it. This confused somebody who was copying the
742 code for some other use. I have removed all the tests.
744 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
745 feature that was used to support insecure browsers during the U.S. crypto
746 embargo. It requires special client support, and Exim is probably the
747 only MTA that supported it -- and would never use it because real RSA is
748 always available. This code has been removed, because it had the bad
749 effect of slowing Exim down by computing (never used) parameters for the
750 RSA_EXPORT functionality.
752 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
753 authenticator to fail if there's a tab character in the incoming data
754 (there should never be unless someone is messing about, as it's supposed
755 to be base64-encoded). Also added, on Timo's advice, the "secured" option
756 if the connection is using TLS or if the remote IP is the same as the
757 local IP, and the "valid-client-cert option" if a client certificate has
760 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
761 authenticators. This can be used for authorization after authentication
762 succeeds. (In the case of plaintext, it servers for both authentication
765 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
766 if any retry times were supplied.
768 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
769 connection, where there is no client IP address to check. In this
770 situation, the verify now always succeeds.
772 PH/25 Applied John Jetmore's -Mset patch.
774 PH/26 Added -bem to be like -Mset, but loading a message from a file.
776 PH/27 In a string expansion for a processed (not raw) header when multiple
777 headers of the same name were present, leading whitespace was being
778 removed from all of them, but trailing whitespace was being removed only
779 from the last one. Now trailing whitespace is removed from each header
780 before concatenation. Completely empty headers in a concatenation (as
783 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
784 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
786 PH/29 [Removed. This was a change that I later backed out, and forgot to
787 correct the ChangeLog entry (that I had efficiently created) before
788 committing the later change.]
790 PH/30 Exim was sometimes attempting to deliver messages that had suffered
791 address errors (4xx response to RCPT) over the same connection as other
792 messages routed to the same hosts. Such deliveries are always "forced",
793 so retry times are not inspected. This resulted in far too many retries
794 for the affected addresses. The effect occurred only when there were more
795 hosts than the hosts_max_try setting in the smtp transport when it had
796 the 4xx errors. Those hosts that it had tried were not added to the list
797 of hosts for which the message was waiting, so if all were tried, there
798 was no problem. Two fixes have been applied:
800 (i) If there are any address or message errors in an SMTP delivery, none
801 of the hosts (tried or untried) are now added to the list of hosts
802 for which the message is waiting, so the message should not be a
803 candidate for sending over the same connection that was used for a
804 successful delivery of some other message. This seems entirely
805 reasonable: after all the message is NOT "waiting for some host".
806 This is so "obvious" that I'm not sure why it wasn't done
807 previously. Hope I haven't missed anything, but it can't do any
808 harm, as the worst effect is to miss an optimization.
810 (ii) If, despite (i), such a delivery is accidentally attempted, the
811 routing retry time is respected, so at least it doesn't keep
812 hammering the server.
814 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
815 in ${readsocket because some servers need this prod.
817 PH/32 Added some extra debug output when updating a wait-xxx database.
819 PH/33 The hint "could be header name not terminated by colon", which has been
820 given for certain expansion errors for a long time, was not being given
821 for the ${if def:h_colon_omitted{... case.
823 PH/34 The spec says: "With one important exception, whenever a domain list is
824 being scanned, $domain contains the subject domain." There was at least
825 one case where this was not true.
827 PH/35 The error "getsockname() failed: connection reset by peer" was being
828 written to the panic log as well as the main log, but it isn't really
829 panic-worthy as it just means the connection died rather early on. I have
830 removed the panic log writing for the ECONNRESET error when getsockname()
833 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
834 runs only) independently of the message's sender address. This meant
835 that, if the 4xx error was in fact related to the sender, a different
836 message to the same recipient with a different sender could confuse
837 things. In particualar, this can happen when sending to a greylisting
838 server, but other circumstances could also provoke similar problems.
839 I have changed the default so that the retry time for these errors is now
840 based a combination of the sender and recipient addresses. This change
841 can be overridden by setting address_retry_include_sender=false in the
844 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
845 remote server are returned as part of bounce messages. This was not
846 happening for LMTP over a pipe (the lmtp transport), but now it is the
847 same for both kinds of LMTP.
849 PH/38 Despite being documented as not happening, Exim was rewriting addresses
850 in header lines that were in fact CNAMEs. This is no longer the case.
852 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
853 and queue runs started by the daemon processed all messages. This has
854 been fixed so that -R and -S can now usefully be given with -q<time>.
856 PH/40 Import PCRE release 6.7 (fixes some bugs).
858 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
860 PH/42 Give an error if -q is specified more than once.
862 PH/43 Renamed the variables $interface_address and $interface_port as
863 $received_ip_address and $received_port, to make it clear that these
864 values apply to message reception, and not to the outgoing interface when
865 a message is delivered. (The old names remain recognized, of course.)
867 PH/44 There was no timeout on the connect() call when using a Unix domain
868 socket in the ${readsocket expansion. There now is.
870 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
871 be meaningful with "accept".
873 SC/01 Eximstats V1.43
874 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
876 SC/02 Eximstats V1.44
877 Use a glob alias rather than an array ref in the generated
878 parser. This improves both readability and performance.
880 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
881 Collect SpamAssassin and rejection statistics.
882 Don't display local sender or destination tables unless
883 there is data to show.
884 Added average volumes into the top table text output.
886 SC/04 Eximstats V1.46
887 Collect data on the number of addresses (recipients)
888 as well as the number of messages.
890 SC/05 Eximstats V1.47
891 Added 'Message too big' to the list of mail rejection
892 reasons (thanks to Marco Gaiarin).
894 SC/06 Eximstats V1.48
895 Mainlog lines which have GMT offsets and are too short to
896 have a flag are now skipped.
898 SC/07 Eximstats V1.49 (Alain Williams)
899 Added the -emptyok flag.
901 SC/08 Eximstats V1.50
902 Fixes for obtaining the IP address from reject messages.
904 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
905 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
906 whitesspace changes from 4.64-PH/27
908 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
911 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
912 are found, allow negative numbers in numeric criteria)
914 JJ/06 exipick.20061117.2, added new $message_body_missing variable
916 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
917 to match changes made in 4.64-PH/43
919 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
921 PH/47 Put in an explicit test for a DNS lookup of an address record where the
922 "domain" is actually an IP address, and force a failure. This locks out
923 those revolvers/nameservers that support "A-for-A" lookups, in
924 contravention of the specifications.
926 PH/48 When a host name was looked up from an IP address, and the subsequent
927 forward lookup of the name timed out, the host name was left in
928 $sender_host_name, contrary to the specification.
930 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
931 restricted to single-key lookups, Exim was not diagnosing an error if
932 * or *@ was used with a query-style lookup.
934 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
936 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
937 long ago, but noone interested enough thought of it. Let's just say that
938 the "1.1" means that there are some new functions that weren't there at
939 some point in the past.
941 PH/51 Error processing for expansion failure of helo_data from an smtp
942 transport during callout processing was broken.
944 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
945 tested/used via the -bh/-bhc/-bs options.
947 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
948 bug, fixed in subsequent PCRE releases).
950 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
951 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
953 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
959 SC/01 Use a glob alias rather than an array ref in eximstats generated
960 parser. This improves both readability and performance.
962 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
963 Don't display local sender or destination tables in eximstats unless
964 there is data to show.
965 Added average volumes into the eximstats top table text output.
967 SC/03 Collect data on the number of addresses (recipients) as well
968 as the number of messages in eximstats.
970 TF/01 Correct an error in the documentation for the redirect router. Exim
971 does (usually) call initgroups() when daemonizing.
973 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
974 with consistent privilege compared to when running as a daemon.
976 TF/03 Note in the spec that $authenticated_id is not set for local
977 submissions from trusted users.
979 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
980 Thanks to Dean Brooks <dean@iglou.com> for the patch.
982 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
983 by adding some example configuration directives to the default
984 configuration file. A little bit of work is required to uncomment the
985 directives and define how usernames and passwords are checked, but
986 there is now a framework to start from.
988 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
989 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
990 without this. I don't know how relevant this is to other LDAP libraries.
992 PH/02 Add the verb name to the "unknown ACL verb" error.
994 PH/03 Magnus Holmgren's patch for filter_prepend_home.
996 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
998 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
999 directory not expanded when it should be if an expanded home directory
1000 was set for the address (which is overridden by the transport).
1002 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1005 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1006 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1007 because it is too late at that time, and has no effect.
1009 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1010 security issue with \' (bugzilla #107). I could not use the
1011 PQescapeStringConn() function, because it needs a PGconn value as one of
1014 PH/08 When testing addresses using -bt, indicate those final addresses that
1015 are duplicates that would not cause an additional delivery. At least one
1016 person was confused, thinking that -bt output corresponded to deliveries.
1017 (Suppressing duplicates isn't a good idea as you lose the information
1018 about possibly different redirections that led to the duplicates.)
1020 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1021 systems where poll() doesn't work, in particular OS X.
1023 PH/10 Added more information to debugging output for retry time not reached.
1025 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1026 operations in malware.c.
1028 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1031 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1032 syslog_duplication set false, log lines that would normally be written
1033 both the the main log and to the reject log were not written to syslog at
1036 PH/14 In the default configuration, change the use of "message" in ACL warn
1037 statements to "add_header".
1039 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1040 not followed by a command (e.g. "seen endif").
1042 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1043 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1046 PH/17 Added extra conditions to the default value of delay_warning_condition
1050 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1051 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1052 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1055 The Auto-Submitted: and various List- headers are standardised, whereas I
1056 don't think Precedence: ever was.
1058 PH/18 Refactored debugging code in route_finduser() to show more information,
1059 in particular, the error code if getpwnam() issues one.
1061 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1062 This is apparently needed in addition to the PH/07 change above to avoid
1063 any possible encoding problems.
1065 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1066 but not after initializing Perl.
1068 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1069 output them only if debugging. By default they are written stderr,
1070 apparently, which is not desirable.
1072 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1075 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1078 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1080 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1081 authenticated or an ident call has been made. Suppress the default
1082 values for $authenticated_id and $authenticated_sender (but permit -oMai
1083 and -oMas) when testing with -bh.
1085 PH/24 Re-jigged the order of the tests in the default configuration so that the
1086 tests for valid domains and recipients precede the DNS black list and CSA
1087 tests, on the grounds that those ones are more expensive.
1089 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1090 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1091 This bug exists in every version of Exim that I still have, right back to
1094 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1095 However, an attempt to turn on case-sensitivity in a regex key by
1096 including (?-i) didn't work because the subject string was already
1097 lowercased, and the effects were non-intuitive. It turns out that a
1098 one-line patch can be used to allow (?-i) to work as expected.
1104 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1105 other effects) broke the use of negated acl sub-conditions.
1107 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1110 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1111 "Deny" causes Exim to reject the incoming connection with a 554 error.
1112 Unfortunately, if there is a major crisis, such as a disk failure,
1113 tcp-wrappers gives "deny", whereas what one would like would be some
1114 kind of temporary error. A kludge has been added to help with this.
1115 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1116 554 error is used if errno is still zero or contains ENOENT (which occurs
1117 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1120 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1122 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1123 errors. Otherwise a message that provokes a temporary error (when other
1124 messages do not) can cause a whole host to time out.
1126 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1127 addresses were routed directly to files or pipes from a redirect router.
1128 File deliveries just didn't batch; pipe deliveries might have suffered
1131 PH/06 A failure to get a lock for a hints database would erroneously always say
1132 "Failed to get write lock", even when it was really a read lock.
1134 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1135 of 0600. This has been changed to use the value of the lockfile_mode
1136 option (which defaults to 0600).
1138 PH/08 Applied small patch from the Sieve maintainer.
1140 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1141 folder from quota calculations, a direct delivery into this folder messed
1142 up the contents of the maildirsize file. This was because the regex was
1143 used only to exclude .Trash (or whatever) when the size of the mailbox
1144 was calculated. There was no check that a delivery was happening into an
1145 excluded directory. This bug has been fixed by ignoring all quota
1146 processing for deliveries into excluded directories.
1148 PH/10 Added the maildirfolder_create_regex option to appendfile.
1154 PH/01 The code for finding all the local interface addresses on a FreeBSD
1155 system running IPv6 was broken. This may well have applied to all BSD
1156 systems, as well as to others that have similar system calls. The broken
1157 code found IPv4 interfaces correctly, but gave incorrect values for the
1158 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1159 that it would not match correctly against @[] and not recognize the IPv6
1162 PH/02 The ipliteral router was not recognizing addresses of the form user@
1163 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1165 PH/03 Added disable_ipv6.
1167 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1168 decoded form, because it is most often used to construct To: headers
1169 lines in autoreplies, and the decoded form may well be syntactically
1170 invalid. However, $reply_address has leading white space removed, and all
1171 newlines turned into spaces so that the autoreply transport does not
1174 PH/05 If group was specified without a user on a router, and no group or user
1175 was specified on a transport, the group from the router was ignored.
1177 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1178 for visible compile-time settings that can be used to change these
1179 numbers, for those that want even more. Backwards compatibility with old
1180 spool files has been maintained. However, going back to a previous Exim
1181 release will lost any variables that are in spool files.
1183 PH/07 Two small changes when running in the test harness: increase delay when
1184 passing a TCP/IP connection to a new process, in case the original
1185 process has to generate a bounce, and remove special handling of
1186 127.0.0.2 (sic), which is no longer necessary.
1188 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1189 be the same on different OS.
1191 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1194 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1195 whether --show-vars was specified or not
1197 JJ/02 exipick: Added support for new ACL variable spool format introduced
1200 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1201 syntactically invalid From: or Reply-to: line, and a filter used this to
1202 generate an autoreply, and therefore failed to obtain an address for the
1203 autoreply, Exim could try to deliver to a non-existent relative file
1204 name, causing unrelated and misleading errors. What now happens is that
1205 it logs this as a hard delivery error, but does not attempt to create a
1208 PH/11 The exinext utility has a -C option for testing purposes, but although
1209 the given file was scanned by exinext itself; it wasn't being passed on
1210 when Exim was called.
1212 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1213 an end-of-file indication when reading a command response.
1215 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1216 compiled. In many other places in Exim, IPv6 addresses are always
1217 recognized, so I have changed this. It also means that IPv4 domain
1218 literals of the form [IPV4:n.n.n.n] are now always recognized.
1220 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1221 used if the router is not running as root, for example, when verifying at
1222 ACL time, or when using -bh. The debugging output from this situation was
1223 non-existent - all you got was a failure to exec. I have made two
1226 (a) Failures to set uid/gid, the current directory, or a process leader
1227 in a subprocess such as that created by queryprogram now generate
1228 suitable debugging ouput when -d is set.
1230 (b) The queryprogram router detects when it is not running as root,
1231 outputs suitable debugging information if -d is set, and then runs
1232 the subprocess without attempting to change uid/gid.
1234 PH/15 Minor change to Makefile for building test_host (undocumented testing
1237 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1238 additional section of a DNS packet that returns MX or SRV records.
1239 Instead, it always explicitly searches for A/AAAA records. This avoids
1240 major problems that occur when a DNS server includes only records of one
1241 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1242 fixed another bug: if SRV records were looked up and the corresponding
1243 address records were *not* found in the additional section, the port
1244 values from the SRV records were lost.
1246 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1247 using the correct key (the original address) when searching the retry
1248 rules in order to find which one to use for generating the retry hint.
1250 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1251 adding the default one. Similarly, if it contains a Reply-To: header, the
1252 errors_reply_to option, if set, is not used.
1254 PH/19 When calculating a retry time, Exim used to measure the "time since
1255 failure" by looking at the "first failed" field in the retry record. Now
1256 it does not use this if it is later than than the arrival time of the
1257 message. Instead it uses the arrival time. This makes for better
1258 behaviour in cases where some deliveries succeed, thus re-setting the
1259 "first failed" field. An example is a quota failure for a huge message
1260 when small messages continue to be delivered. Without this change, the
1261 "time since failure" will always be short, possible causing more frequent
1262 delivery attempts for the huge message than are intended.
1263 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1265 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1266 $1, $2, $3) because the numerical variables can be reset during some
1267 expansion items (e.g. "match"), thereby losing the authentication data.
1269 PH/21 Make -bV show the size of off_t variables so that the test suite can
1270 decide whether to run tests for quotas > 2G.
1272 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1273 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1274 filecount value is greater than 2G or if a quota value is greater than 2G
1275 on a system where the size of off_t is not greater than 4, a panic error
1278 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1279 never match. The debug and -bh output now contains an explicit error
1280 message indicating a malformed IPv4 address or mask.
1282 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1283 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1284 PH/23 above applies.
1286 PH/25 Do not write to syslog when running in the test harness. The only
1287 occasion when this arises is a failure to open the main or panic logs
1288 (for which there is an explicit test).
1290 PH/26 Added the /no_tell option to "control=freeze".
1292 PH/27 If a host name lookup failed very early in a connection, for example, if
1293 the IP address matched host_lookup and the reverse lookup yielded a name
1294 that did not have a forward lookup, an error message of the form "no IP
1295 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1296 could be logged. Now it outputs the IP address instead of "NULL".
1298 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1299 allows the sender and the authenticated sender to be set when
1300 submitting a message from within Exim. Since child_open_exim() is
1301 documented for local_scan(), the new function should be too.
1303 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1304 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1305 results in an empty string is now treated as unset.
1307 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1309 PH/31 Added sender_verify_fail logging option.
1311 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1312 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1313 tidied the source and removed it altogether.
1315 PH/33 When a queue run was abandoned because the load average was too high, a
1316 log line was always written; now it is written only if the queue_run log
1317 selector is set. In addition, the log line for abandonment now contains
1318 information about the queue run such as the pid. This is always present
1319 in "start" and "stop" lines but was omitted from the "abandon" line.
1321 PH/34 Omit spaces between a header name and the colon in the error message that
1322 is given when verify = headers_syntax fails (if there are lots of them,
1323 the message gets confusing).
1325 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1326 names, as there are now some PTR records that contain slashes. This check
1327 is only to protect against broken name servers that fall over on strange
1328 characters, so the fact that it applies to all lookups doesn't matter.
1330 PH/36 Now that the new test suite is complete, we can remove some of the
1331 special code in Exim that was needed for the old test suite. For example,
1332 sorting DNS records because real resolvers return them in an arbitrary
1333 order. The new test suite's fake resolver always returns records in the
1336 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1337 bounces) except when queue_only is set, to avoid logging races between
1338 the different processes.
1340 PH/38 Panic-die if .include specifies a non-absolute path.
1342 PH/39 A tweak to the "H" retry rule from its user.
1344 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1345 a label. They prevented compilation on older perls.
1347 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1348 a warning to be raised on newish perls.
1350 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1351 on queue. Changes to match documented behaviour of showing count of
1352 messages matching specified criteria.
1354 PH/40 Changed the default ident timeout from 30s to 5s.
1356 PH/41 Added support for the use of login_cap features, on those BSD systems
1357 that have them, for controlling the resources used by pipe deliveries.
1359 PH/42 The content-scanning code uses fopen() to create files in which to put
1360 message data. Previously it was not paying any attention to the mode of
1361 the files. Exim runs with umask(0) because the rest of the code creates
1362 files with open(), and sets the required mode explicitly. Thus, these
1363 files were ending up world-writeable. This was not a big issue, because,
1364 being within the spool directory, they were not world-accessible. I have
1365 created a function called modefopen, which takes an additional mode
1366 argument. It sets umask(777), creates the file, chmods it to the required
1367 mode, then resets the umask. All the relevant calls to fopen() in the
1368 content scanning code have been changed to use this function.
1370 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1371 to 24 hours. This avoids potential overflow problems when processing G
1372 and H retry rules. I suspect nobody ever tinkers with this value.
1374 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1376 PH/45 When the plaintext authenticator is running as a client, the server's
1377 challenges are checked to ensure they are valid base64 strings. By
1378 default, the authentication attempt is cancelled if an invalid string is
1379 received. Setting client_ignore_invalid_base64 true ignores these errors.
1380 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1381 they are received. Thus, the responses can be made to depend on the
1382 challenges. If an invalid string is ignored, an empty string is placed in
1385 PH/46 Messages that are created by the autoreply transport now contains a
1386 References: header, in accordance with RFCs 2822 and 3834.
1388 PH/47 Added authenticated_sender_force to the smtp transport.
1390 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1392 PH/49 Installed latest patch from the Sieve maintainer.
1394 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1395 was also set, the appendfile transport was unnecessarily scanning a
1396 directory of message files (e.g. for maildir delivery) to find the count
1397 of files (along with the size), even though it did not need this
1398 information. It now does the scan only if it needs to find either the
1399 size of the count of files.
1401 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1403 PH/52 Two bugs concerned with error handling when the smtp transport is
1406 (i) Exim was not creating retry information for temporary errors given
1407 for individual recipients after the DATA command when the smtp transport
1408 was used in LMTP mode. This meant that they could be retried too
1409 frequently, and not timed out correctly.
1411 (ii) Exim was setting the flag that allows error details to be returned
1412 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1413 recipients that were returned after the DATA command.
1415 PH/53 This is related to PH/52, but is more general: for any failing address,
1416 when detailed error information was permitted to be returned to the
1417 sender, but the error was temporary, then after the final timeout, only
1418 "retry timeout exceeded" was returned. Now it returns the full error as
1419 well as "retry timeout exceeded".
1421 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1422 do this, and (what is worse) MTAs that accept it.
1424 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1425 will now be deprecated.
1427 PH/56 New os.c-cygwin from the Cygwin maintainer.
1429 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1430 formats (previously only available in exim formats via -bpr, -bpru,
1431 and -bpra. Now also available in native and exiqgrep formats)
1433 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1434 with very large, slow to parse queues
1436 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1438 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1440 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1441 responses to authentication challenges, though it was showing the
1442 challenges; (ii) I've removed the CR characters from the debug output for
1445 PH/58 Allow for the insertion of a newline as well as a space when a string
1446 is turned into more than one encoded-word during RFC 2047 encoding. The
1447 Sieve code now uses this.
1449 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1450 data_4xx, lost_connection, tls_required.
1452 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1453 message was being sent as an SMTP response.
1455 PH/61 Add -l and -k options to exicyclog.
1457 PH/62 When verifying, if an address was redirected to one new address, so that
1458 verification continued, and the new address failed or deferred after
1459 having set something in $address_data, the value of $address_data was not
1460 passed back to the ACL. This was different to the case when no
1461 redirection occurred. The value is now passed back in both cases.
1463 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1464 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1465 use login_cap.h, so on its own it isn't the distinguishing feature. The
1466 new name refers directly to the setclassresources() function.
1468 PH/65 Added configuration files for NetBSD3.
1470 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1472 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1473 is preferred over IPv4.
1475 PH/68 The bounce_return_message and bounce_return_body options were not being
1476 honoured for bounces generated during the reception of non-SMTP messages.
1477 In particular, this applied to messages rejected by the ACL. This bug has
1478 been fixed. However, if bounce_return_message is true and bounce_return_
1479 body is false, the headers that are returned for a non-SMTP message
1480 include only those that have been read before the error was detected.
1481 (In the case of an ACL rejection, they have all been read.)
1483 PH/69 The HTML version of the specification is now built in a directory called
1484 spec_html instead of spec.html, because the latter looks like a path with
1485 a MIME-type, and this confuses some software.
1487 PH/70 Catch two compiler warnings in sieve.c.
1489 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1490 function verify_get_ident() calls ip_connect() to connect a socket, but
1491 if the "connect()" function timed out, ip_connect() used to close the
1492 socket. However, verify_get_ident() also closes the socket later, and in
1493 between Exim writes to the log, which may get opened at this point. When
1494 the socket was closed in ip_connect(), the log could get the same file
1495 descriptor number as the socket. This naturally causes chaos. The fix is
1496 not to close the socket in ip_connect(); the socket should be closed by
1497 the function that creates it. There was only one place in the code where
1498 this was missing, in the iplookup router, which I don't think anybody now
1499 uses, but I've fixed it anyway.
1501 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1502 well as to direct DNS lookups. Otherwise the handling of names in host
1503 lists is inconsistent and therefore confusing.
1509 PH/01 Two changes to the default runtime configuration:
1511 (1) Move the checks for relay_from_hosts and authenticated clients from
1512 after to before the (commented out) DNS black list checks.
1514 (2) Add control=submission to the relay_from_hosts and authenticated
1515 clients checks, on the grounds that messages accepted by these
1516 statements are most likely to be submissions.
1518 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1520 (1) Generate an error if the third argument for the ${prvs expansion is
1523 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1526 (3) Reset the variables that are obtained from the first argument of
1527 ${prvscheck and used in the second argument before leaving the code,
1528 because their memory is reclaimed, so using them afterwards may do
1531 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1532 one (it's much easier than Tom thought :-).
1534 (5) Because of (4), we can now allow for the use of $prvscheck_result
1535 inside the third argument.
1537 PH/03 For some reason, the default setting of PATH when running a command from
1538 a pipe transport was just "/usr/bin". I have changed it to
1541 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1542 anything to be listed in the output from -bV.
1544 PH/05 When a filter generated an autoreply, the entire To: header line was
1545 quoted in the delivery log line, like this:
1547 => >A.N.Other <ano@some.domain> <original@ddress> ...
1549 This has been changed so that it extracts the operative address. There
1550 may be more than one such address. If so, they are comma-separated, like
1553 => >ano@some.domain,ona@other.domain <original@ddress> ...
1555 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1556 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1557 not being looked up in the rDNS to get a host name, Exim was showing the
1558 IP address twice in Received: lines, even though the IP addresses were
1559 identical. For example:
1561 Received: from [1.2.3.4] (helo=[1.2.3.4])
1563 However, if the real host name was known, it was omitting the HELO data
1564 if it matched the actual IP address. This has been tidied up so that it
1565 doesn't show the same IP address twice.
1567 PH/07 When both +timestamp and +memory debugging was on, the value given by
1568 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1569 called by the debug printing, thereby overwriting the timestamp buffer.
1570 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1572 PH/08 When the original message was included in an autoreply transport, it
1573 always said "this is a copy of the message, including all the headers",
1574 even if body_only or headers_only was set. It now gives an appropriate
1577 PH/09 Applied a patch from the Sieve maintainer which:
1579 o fixes some comments
1580 o adds the (disabled) notify extension core
1581 o adds some debug output for the result of if/elsif tests
1582 o points to the current vacation draft in the documentation
1583 and documents the missing references header update
1587 o fixes a bug in processing the envelope test (when testing
1588 multiple envelope elements, the last element determinted the
1591 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1592 Electronic Mail") by including:
1594 Auto-submitted: auto-generated
1596 in the messages that it generates (bounce messages and others, such as
1597 warnings). In the case of bounce messages for non-SMTP mesages, there was
1598 also a typo: it was using "Auto_submitted" (underscore instead of
1599 hyphen). Since every message generated by Exim is necessarily in response
1600 to another message, thes have all been changed to:
1602 Auto-Submitted: auto-replied
1604 in accordance with these statements in the RFC:
1606 The auto-replied keyword:
1608 - SHOULD be used on messages sent in direct response to another
1609 message by an automatic process,
1611 - MUST NOT be used on manually-generated messages,
1613 - MAY be used on Delivery Status Notifications (DSNs) and Message
1614 Disposition Notifications (MDNs),
1616 - MUST NOT be used on messages generated by automatic or periodic
1617 processes, except for messages which are automatic responses to
1620 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1621 to the default Received: header definition.
1623 PH/12 Added log selector acl_warn_skipped (default on).
1625 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1626 variables because (a) they are in the wrong storage pool and (b) even if
1627 they were copied, it wouldn't work properly because of the caching.
1629 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1630 checking when decoding. Apparently there are clients that generate
1631 overlong encoded strings. Why am I not surprised?
1633 PH/15 If the first argument of "${if match_address" was not empty, but did not
1634 contain an "@" character, Exim crashed. Now it writes a panic log message
1635 and treats the condition as false.
1637 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1639 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1640 extension "envelope-auth". The code is finished and in agreement with
1641 other implementations, but there is no documentation so far and in fact,
1642 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1643 not changing the active code.
1645 Print executed "if" and "elsif" statements when debugging is used. This
1646 helps a great deal to understand what a filter does.
1648 Document more things not specified clearly in RFC3028. I had all this
1649 sorted out, when out of a sudden new issues came to my mind. Oops."
1651 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1654 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1655 canonical form (as documented). However, after a host name lookup from
1656 the IP address, check_host() was doing a simple string comparison with
1657 addresses acquired from the DNS when checking that the found name did
1658 have the original IP as one of its addresses. Since any found IPv6
1659 addresses are likely to be in abbreviated form, the comparison could
1660 fail. Luckily, there already exists a function for doing the comparison
1661 by converting both addresses to binary, so now that is used instead of
1662 the text comparison.
1664 PH/20 There was another similar case to PH/19, when a complete host name was
1665 given in a host list; looking up its IP address could give an abbreviated
1666 form, whereas the current host's name might or might not be abbreviated.
1667 The same fix has been applied.
1673 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1674 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1677 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1678 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1680 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1682 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1683 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1684 Also, for "Auto-Submitted", treat anything other than "no" as
1685 non-personal, in accordance with RFC 3834. (Previously it treated
1686 anything starting "auto-" as non-personal.)
1688 TF/01 The control=submission/name=... option had a problem with syntax
1689 errors if the name included a slash character. The /name= option
1690 now slurps the rest of the string, so it can include any characters
1691 but it must come last in the list of options (after /sender_retain
1694 PH/05 Some modifications to the interface to the fake nameserver for the new
1702 TK/01 Added the "success_on_redirect" address verification option. See
1703 NewStuff for rationale and an example.
1705 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1707 PH/02 Patch to exigrep to allow it to work on syslog lines.
1709 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1710 fread() to skip over the body file's header line, because in Cygwin the
1711 header line is locked and is inaccessible.
1713 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1714 co-exist for some time) to make it clear that it is the Exim ID that is
1715 referenced, not the Message-ID: header line.
1717 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1718 string_format() function, because snprintf() does not exist on all
1721 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1722 ${stat: expansion item.
1724 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1725 protocol synchronization error", to keep the pedants happy.
1727 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1728 well as for IRIX systems, when gcc is being used. See the host.c source
1731 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1733 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1736 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1737 transport and to the smtp transport in LMTP mode.
1739 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1741 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1742 run a filter in a subprocess. This could lead to confusion in subsequent
1743 lookups in the parent process. There should also be a search_tidyup() at
1744 the end of the subprocess.
1746 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1747 only if the host matched helo_try_verify_hosts, which caused the
1748 verification to occur when the EHLO/HELO command was issued. The ACL just
1749 tested the remembered result. Now, if a previous verification attempt has
1750 not happened, "verify = helo" does it there and then.
1752 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1754 TK/03 Fix log output including CR from clamd.
1756 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1757 exist provoked a memory error which could cause a segfault.
1759 PH/15 Installed PCRE 6.2
1761 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1763 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1764 of the problem. Specifically, suggested +O2 rather than +O1 for the
1767 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1769 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1770 record, the port was not correctly logged when the outgoing_port log
1771 selector was set (it logged the transort's default port).
1773 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1774 fallback_hosts, and "hosts" in the smtp transport.
1776 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1777 host errors such as "Connection refused".
1779 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1780 authentication with radiusclient 0.4.9:
1782 - Error returned from rc_read_config was caught wrongly
1783 - Username/password not passed on to radius server due to wrong length.
1785 The presumption is that some radiusclient API changes for 4.51/PH/17
1786 were not taken care of correctly. The code is still untested by me (my
1787 Linux distribution still has 0.3.2 of radiusclient), but it was
1788 contributed by a Radius user.
1790 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1791 expanding the "port" option of the smtp transport.
1793 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1794 while reading a MIME header. Thanks to Tom Hughes for a patch.
1796 PH/24 Include config.h inside local_scan.h so that configuration settings are
1799 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1800 that in an ACL for RCPT (for example), you can examine exactly what was
1803 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1804 commands, but it was not correctly comparing the address with the actual
1805 client host address. Thus, it would show the EHLO address in Received:
1806 header lines when this was not necessary.
1808 PH/27 Added the % operator to ${eval:}.
1810 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1811 it should be ignoring failures (because with -C, for example, it has lost
1812 privilege). It wasn't ignoring creation failures other than "already
1815 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1818 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1819 because an input line was too long, either on its own, or by virtue of
1820 too many continuations, the temporary file was not being removed, and the
1821 return code was incorrect.
1823 PH/31 Missing "BOOL" in function definition in filtertest.c.
1825 PH/32 Applied Sieve patches from the maintainer.
1827 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1829 PH/33 Added "verify = not_blind".
1831 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1832 Local/Makefile (with some defaults set). These are used in built scripts
1833 such as exicyclog, but they have never been used in the exim_install
1834 script (though there are many overriding facilities there). I have
1835 arranged that the exim_install script now takes note of these two
1838 PH/35 Installed configuration files for Dragonfly.
1840 PH/36 When a locally submitted message by a trusted user did not contain a
1841 From: header, and the sender address was obtained from -f or from an SMTP
1842 MAIL command, and the trusted user did not use -F to supply a sender
1843 name, $originator_name was incorrectly used when constructing a From:
1844 header. Furthermore, $originator_name was used for submission mode
1845 messages from external hosts without From: headers in a similar way,
1846 which is clearly wrong.
1848 PH/37 Added control=suppress_local_fixups.
1850 PH/38 When log_selector = +received_sender was set, and the addition of the
1851 sender made the log line's construction buffer exactly full, or one byte
1852 less than full, an overflow happened when the terminating "\n" was
1855 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1856 when the result of a list match is failure because a DNS lookup failed.
1858 PH/40 RM_COMMAND is now used in the building process.
1860 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1861 the "build-* directories that it finds.
1863 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1864 address in a domain literal was a prefix of an interface address.
1866 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1867 when verifying a sender address, unless rewrite_headers is false.
1869 PH/44 Wrote a long comment about why errors_to addresses are verified as
1870 recipients, not senders.
1872 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1873 the ratelimit ACL was added.
1875 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1877 PH/46 Added extra information about PostgreSQL errors to the error string.
1879 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1880 avoiding the need to install special zones in a real server. This is
1881 backwards compatible; if it can't find the fake resolver, it drops back.
1882 Thus, both old and new test suites can be run.
1884 TF/02 Added util/ratelimit.pl
1886 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1887 clock is set back in time.
1889 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1890 Candler <B.Candler@pobox.com>.
1892 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1893 OK for addresses that are the result of redirections.
1895 PH/48 A number of further additions for the benefit of the new test suite,
1896 including a fake gethostbyname() that interfaces to the fake DNS resolver
1899 TF/06 The fix for widen_domains has also been applied to qualify_single and
1900 search_parents which are the other dnslookup options that can cause
1903 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1906 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1908 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1909 service" if it encountered an MX record with an empty target hostname.
1910 The message is now "an MX or SRV record indicated no SMTP service".
1912 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1913 if the DNS of the sending site is misconfigured. This is quite a
1914 common situation. This change restores the behaviour of treating a
1915 helo verification defer as a failure.
1917 PH/51 If self=fail was set on a router, the bounce message did not include the
1918 actual error message.
1924 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1926 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1927 command itself ended in error, the underlying message about the transport
1928 filter timeout was being overwritten with the pipe command error. Now the
1929 underlying error message should be appended to the second error message.
1931 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1933 PH/02 Reduce the amount of output that "make" produces by default. Full output
1934 can still be requested.
1936 PH/03 The warning log line about a condition test deferring for a "warn" verb
1937 was being output only once per connection, rather than after each
1938 occurrence (because it was using the same function as for successful
1939 "warn" verbs). This seems wrong, so I have changed it.
1941 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1942 it should not have, which might have caused a crash in the right
1943 circumstances, but probably never did.
1945 PH/04 Installed a modified version of Tony Finch's patch to make submission
1946 mode fix the return path as well as the Sender: header line, and to
1947 add a /name= option so that you can make the user's friendly name appear
1950 TF/03 Added the control = fakedefer ACL modifier.
1952 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1953 Mark Lowes for thorough testing.
1955 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1957 TK/03 Merged latest SRS patch from Miles Wilton.
1959 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1960 with the definition in sysexits.h (which is #included earlier).
1961 Fortunately, Exim does not actually use EX_OK. The code used to try to
1962 preserve the sysexits.h value, by assumimg that macro definitions were
1963 scanned for macro replacements. I have been disabused of this notion,
1964 so now the code just undefines EX_OK before #including unistd.h.
1966 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1967 in the smtp transport. When a block could not be written in a single
1968 write() function, the timeout was being re-applied to each part-write.
1969 This seems wrong - if the receiver was accepting one byte at a time it
1970 would take for ever. The timeout is now adjusted when this happens. It
1971 doesn't have to be particularly precise.
1973 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1974 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1976 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1977 without a domain if the check to <postmaster@domain> fails.
1979 SC/01 Eximstats: added -xls and the ability to specify output files
1980 (patch written by Frank Heydlauf).
1982 SC/02 Eximstats: use FileHandles for outputing results.
1984 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1986 SC/04 Eximstats: fixed display of large numbers with -nvr option
1988 SC/05 Eximstats: fixed merging of reports with empty tables.
1990 SC/06 Eximstats: added the -include_original_destination flag
1992 SC/07 Eximstats: removed tabs and trailing whitespace.
1994 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1996 TK/06 MBOX spool code: Add real "From " MBOX separator line
1997 so the .eml file is really in mbox format (even though
1998 most programs do not really care). Patch from Alex Miller.
2000 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2001 The latter is generated from $received_to and is only set if the
2002 message has one envelope recipient. SA can use these headers,
2003 obviously out-of-the-box. Patch from Alex Miller.
2005 PH/08 The ${def test on a variable was returning false if the variable's
2006 value was "0", contrary to what the specification has always said!
2007 The result should be true unless the variable is empty.
2009 PH/09 The syntax error of a character other than { following "${if
2010 def:variable_name" (after optional whitespace) was not being diagnosed.
2011 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2012 accidental colon was present, for example, could give incorrect results.
2014 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2015 result is used (not including appendfile, where other changes are about
2018 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2019 This involved changing a lot of size variables from int to off_t. It
2020 should work with maildirs and everything.
2022 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2023 spamd dying while we are connected to it.
2025 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2028 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2029 function for BDB 4.3.
2031 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2033 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2034 the sources for PCRE have been re-arranged and more files are now
2037 PH/15 The code I had for printing potentially long long variables in PH/11
2038 above was not the best (it lost precision). The length of off_t variables
2039 is now inspected at build time, and an appropriate printing format (%ld
2040 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2041 to be "long long int" or "long int". This is needed for the internal
2042 formatting function string_vformat().
2044 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2045 the configuration file to be ":syslog", then the script "guesses" where
2046 the logs files are, rather than using the compiled in default. In our
2047 case the guess is not the same as the compiled default, so the script
2048 suddenly stopped working when I started to use syslog. The patch checks
2049 to see if log_file_path is "". If so, it attempts to read it from exim
2050 with no configuration file to get the compiled in version, before it
2051 falls back to the previous guessing code."
2053 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2054 implementing BATV in an Exim configuration. See NewStuff for the gory
2057 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2058 Makefile that are specific to HP-UX.
2060 PH/18 If the "use_postmaster" option was set for a recipient callout together
2061 with the "random" option, the postmaster address was used as the MAIL
2062 FROM address for the random test, but not for the subsequent recipient
2063 test. It is now used for both.
2065 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2066 patch removes a few documentation additions to RFC 3028, because the
2067 latest draft now contains them. It adds the new en;ascii-case comparator
2068 and a new error check for 8bit text in MIME parts. Comparator and
2069 require names are now matched exactly. I enabled the subaddress
2070 extension, but it is not well tested yet (read: it works for me)."
2072 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2073 rework some of the code of TK/09 above to avoid the hardwired use of
2074 "%lld" and "long long". Replaced the call to snprintf() with a call to
2077 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2078 records point to non-existent hosts", "retry timeout exceeded", and
2079 "retry time not reached for any host after a long failure period".
2081 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2082 experimental DomainKeys support:
2084 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2085 (2) On an error such as an illegally used "control", the wrong name for
2086 the control was given.
2088 These problems did NOT occur unless DomainKeys support was compiled.
2090 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2092 PH/24 Added ${if match_ip condition.
2094 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2095 databases so that it will be absolutely obvious if a crash occurs in the
2096 DB library. This is a regular occurrence (often caused by mis-matched
2099 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2100 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2101 picked up on a user's system that detects such things. There doesn't seem
2102 to be a gcc warning option for this - only an attribute that has to be
2103 put on the function's prototype. It seems that in Fedora Core 4 they have
2104 set this on a number of new functions. No doubt there will be more in due
2107 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2108 specify a transport. However, if an address that was verified by such a
2109 router was the subject of a callout, Exim crashed because it tried to
2110 read the rcpt_include_affixes from the non-existent transport. Now it
2111 just assumes that the setting of that option is false. This bug was
2112 introduced by 4.51/PH/31.
2114 PH/28 Changed -d+all to exclude +memory, because that information is very
2115 rarely of interest, but it makes the output a lot bigger. People tend to
2116 do -d+all out of habit.
2118 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2119 code in os-type was giving problems when libc.so lives in lib64, like on
2122 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2123 aren't the modern standard, and it seems that some systems' include files
2124 don't always have them. Exim was already checking for some of the newer
2125 ones like T_AAAA, and defining it itself. I've added checks for all the
2126 record types that Exim uses.
2128 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2129 not automatically generating a new one, as it is supposed to. This
2130 prevented TLS from working. If the file did exist, but contained invalid
2131 data, a new version was generated, as expected. It was only the case of a
2132 non-existent file that was broken.
2134 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2135 with a change in libdomainkeys > 0.64.
2137 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2138 from DNS. If the selector record carries the flag, it now has
2139 precedence over the domain-wide flag.
2141 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2143 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2144 the use of a port name that isn't defined in /etc/services) occurred, the
2145 message was deferred as in a normal delivery, and thus remained on the
2146 spool, instead of being failed because of the mua_wrapper setting. This
2147 is now fixed, and I tidied up some of the mua_wrapper messages at the
2150 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2151 the messages in a hash of arrays rather than using individual hashes.
2152 This is a bit cleaner and results in dramatic memory savings, albeit
2153 at a slight CPU cost.
2155 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2156 as requested by Marc Sherman.
2158 SC/10 Eximstats: added histograms for user specified patterns as requested
2161 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2163 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2164 fopen() in the content-scanning modules that did not already have it.
2170 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2171 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2173 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2175 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2177 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2178 to test Sieve filters that use "vacation".
2180 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2181 that changes the way the GnuTLS parameters are stored in the cache file.
2182 The new format can be generated externally. For backward compatibility,
2183 if the data in the cache doesn't make sense, Exim assumes it has read an
2184 old-format file, and it generates new data and writes a new file. This
2185 means that you can't go back to an older release without removing the
2188 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2189 work if there are any delivery delays because "one_time" forces the
2190 parent to be marked "delivered", so its unseen clone is never tried
2191 again. For this reason, Exim now forbids the simultaneous setting of
2194 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2195 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2196 entry if you want to know the details. The fix, however, neglected to
2197 consider the case when local delivery batching is involved. The test for
2198 "previously delivered" was not happening when checking to see if an
2199 address could be batched with a previous (undelivered) one; under
2200 certain circumstances this could lead to multiple deliveries to the same
2203 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2204 in its include files, and this causes problems building Exim.
2206 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2207 header_syntax) but Exim was just ignoring anything given after a slash.
2208 In particular, this caused confusion with an attempt to use "verify =
2209 reverse_host_lookup/defer_ok". An error is now given when options are
2210 supplied for verify items that do not have them. (Maybe reverse_host_
2211 lookup should have a defer_ok option, but that's a different point.)
2213 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2214 defined by RFC 821) to 2048, because there were problems with some AUTH
2215 commands, and RFC 1869 says the size should be increased for extended
2216 SMTP commands that take arguments.
2218 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2221 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2222 "unknown" error; now it says that the functionality isn't in the binary.
2224 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2225 an address' error message when a string expansion fails (syntax or
2226 whatever). Otherwise the password may appear in the log. Following change
2227 PH/42 below, there is no longer a chance of it appearing in a bounce
2230 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2232 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2233 of its addresses was ever tried. (Bugzilla bug #2.)
2235 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2236 the result incorrectly in the debug output. (It correctly added a newline
2237 to what was transported.)
2239 TF/01 Added $received_time.
2241 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2242 commented out examples of how to interface to a virus scanner and to
2243 SpamAssassin. Also added commented examples of av_scanner and
2244 spamd_address settings.
2246 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2247 and controls are allowed in which ACLs. There were a couple of minor
2248 errors. Some of the entries in the conditions table (which is a table of
2249 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2250 negation of where the condition IS allowed.
2252 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2254 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2255 header file does not have a version number, so I've had to invent a new
2256 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2257 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2258 radiusclient), but it was contributed by a Radius user.
2260 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2261 files or directories, for OpenSSL.
2263 PH/19 When an Exim process that is running as root has to create an Exim log
2264 file, it does so in a subprocess that runs as exim:exim so as to get the
2265 ownership right at creation (otherwise, other Exim processes might see
2266 the file with the wrong ownership). There was no test for failure of this
2267 fork() call, which would lead to the process getting stuck as it waited
2268 for a non-existent subprocess. Forks do occasionally fail when resources
2269 run out. I reviewed all the other calls to fork(); they all seem to check
2272 PH/20 When checking for unexpected SMTP input at connect time (before writing
2273 the banner), Exim was not dealing correctly with a non-positive return
2274 from the read() function. If the client had disconnected by this time,
2275 the result was a log entry for a synchronization error with an empty
2276 string after "input=" when read() returned zero. If read() returned -1
2277 (an event I could not check), uninitialized data bytes were printed.
2278 There were reports of junk text (parts of files, etc) appearing after
2281 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2283 PH/22 Added support for macro redefinition, and (re)definition in between
2284 driver and ACL definitions.
2286 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2287 forgetting to use the resulting value; it was using the unexpanded value.
2289 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2290 hadn't been configured. The fix is from Juergen Kreileder, who
2291 understands it better than I do:
2293 "Here's what I see happening with three configured cyrus_sasl
2294 authenticators configured (plain, login, cram-md5):
2296 On startup auth_cyrus_sasl_init() gets called for each of these.
2297 This means three calls to sasl_listmech() without a specified mech_list.
2298 => SASL tests which mechs of all available mechs actually work
2299 => three warnings about OTP not working
2300 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2302 With the patch, sasl_listmech() also gets called three times. But now
2303 SASL's mech_list option is set to the server_mech specified in the the
2304 authenticator. Or in other words, the answer from sasl_listmech()
2305 gets limited to just the mech you're testing for (which is different
2307 => the return list contains just 'plain' or 'login', 'cram-md5' or
2308 nothing depending on the value of ob->server_mech.
2310 I've just tested the patch: Authentication still works fine,
2311 unavailable mechs specified in the exim configuration are still
2312 caught, and the auth.log warnings about OTP are gone."
2314 PH/25 When debugging is enabled, the contents of the command line are added
2315 to the debugging output, even when log_selector=+arguments is not
2318 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2319 answer is "GNU", and only if the return is "GNU/something" is the answer
2322 PH/27 $acl_verify_message is now set immediately after the failure of a
2323 verification in an ACL, and so is available in subsequent modifiers. In
2324 particular, the message can be preserved by coding like this:
2326 warn !verify = sender
2327 set acl_m0 = $acl_verify_message
2329 Previously, $acl_verify_message was set only while expanding "message"
2330 and "log_message" when a very denied access.
2332 PH/28 Modified OS/os.c-Linux with
2334 -#ifndef OS_LOAD_AVERAGE
2335 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2337 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2338 nomenclature these days.)
2340 PH/29 Installed patch from the Sieve maintainer that adds the options
2341 sieve_useraddress and sieve_subaddress to the redirect router.
2343 PH/30 In these circumstances:
2344 . Two addresses routed to the same list of hosts;
2345 . First host does not offer TLS;
2346 . First host accepts first address;
2347 . First host gives temporary error to second address;
2348 . Second host offers TLS and a TLS session is established;
2349 . Second host accepts second address.
2350 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2351 and peerdn, if requested) that were in fact used only for the second
2354 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2355 attention to any local part prefix or suffix that was matched by the
2356 router that accepted the address. It now behaves in the same way as it
2357 does for delivery: the affixes are removed from the local part unless
2358 rcpt_include_affixes is set on the transport.
2360 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2361 timeout during the DATA phase of an incoming message.
2363 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2364 applied a patch sanctioned by the Sieve maintainer.
2366 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2367 the uid or gid is negative. A case of a negative gid caused this to be
2368 noticed. The fix allows for either to be negative.
2370 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2371 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2374 PH/36 The change PH/12 above was broken. Fixed it.
2376 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2377 the grounds that routing the same address twice would always produce the
2378 same answer. This might have been true once, but it is certainly no
2379 longer true now. Routing a child address may depend on the previous
2380 routing that produced that child. Some complicated redirection strategies
2381 went wrong when messages had multiple recipients, and made Exim's
2382 behaviour dependent on the order in which the addresses were given.
2384 I have moved the duplicate checking until after the routing is complete.
2385 Exim scans the addresses that are assigned to local and remote
2386 transports, and removes any duplicates. This means that more work will be
2387 done, as duplicates will always all be routed, but duplicates are
2388 presumably rare, so I don't expect this is of any significance.
2390 For deliveries to pipes, files, and autoreplies, the duplicate checking
2391 still happens during the routing process, since they are not going to be
2394 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2395 It corrects a timeout issue with spamd. This is Ian's comment: "The
2396 background is that sometimes spamd either never reads data from a
2397 connection it has accepted, or it never writes response data. The exiscan
2398 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2399 blindly assumes that writes won't block so it may never time out."
2401 PH/39 Allow G after quota size as well as K and M.
2403 PH/40 The value set for $authenticated_id in an authenticator may not contain
2404 binary zeroes or newlines because the value is written to log lines and
2405 to spool files. There was no check on this. Now the value is run through
2406 the string_printing() function so that such characters are converted to
2407 printable escape sequences.
2409 PH/41 $message_linecount is a new variable that contains the total number of
2410 lines in the message. Compare $body_linecount, which is the count for the
2413 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2414 bounce and delay warning messages, except in certain special cases, which
2417 (a) An SMTP error message from a remote host;
2418 (b) A message specified in a :fail: redirection;
2419 (c) A message specified in a "fail" command in a system filter;
2420 (d) A message specified in a FAIL return from the queryprogram router;
2421 (e) A message specified by the cannot_route_message router option.
2423 In these cases only, Exim does include the error details in bounce and
2424 warning messages. There are also a few cases where bland messages such
2425 as "unrouteable address" or "local delivery error" are given.
2427 PH/43 $value is now also set for the "else" part of a ${run expansion.
2429 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2430 being worked on, but at least Exim now implements the latest version to
2433 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2434 process to complete was treated as a delivery failure, a timeout while
2435 writing the message to the pipe was logged, but erroneously treated as a
2436 successful delivery. Such timeouts include transport filter timeouts. For
2437 consistency with the overall process timeout, these timeouts are now
2438 treated as errors, giving rise to delivery failures by default. However,
2439 there is now a new Boolean option for the pipe transport called
2440 timeout_defer, which, if set TRUE, converts the failures into defers for
2441 both kinds of timeout. A transport filter timeout is now identified in
2444 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2445 systems where "make" and "gmake" are different, calling "gmake" at top
2446 level broke things. I've arranged for the value of $(MAKE) to be passed
2447 from the Makefile to this script so that it can call the same version of
2451 A note about Exim versions 4.44 and 4.50
2452 ----------------------------------------
2454 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2455 changes of various kinds. As a consequence, a big documentation update was
2456 needed. This delayed the release for rather longer than seemed good, especially
2457 in the light of a couple of (minor) security issues. Therefore, the changes
2458 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2459 release. So 4.44 and 4.50 are in effect two different branches that both start
2462 I have left the 4.50 change log unchanged; it contains all the changes since
2463 4.43. The change log for 4.44 is below; many of its items are identical to
2464 those for 4.50. This seems to be the most sensible way to preserve the
2465 historical information.
2471 1. Minor wording change to the doc/README.SIEVE file.
2473 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2474 computation of the current number of files was incorrect.
2476 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2477 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2480 4. Give more explanation in the error message when the command for a transport
2481 filter fails to execute.
2483 5. There are several places where Exim runs a non-Exim command in a
2484 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2485 was being done only for the command run by the queryprogram router. It is
2486 now done for all such subprocesses. The other cases are: ${run, transport
2487 filters, and the commands run by the lmtp and pipe transports.
2489 6. Added CONFIGURE_GROUP build-time option.
2491 7. Some older OS have a limit of 256 on the maximum number of file
2492 descriptors. Exim was using setrlimit() to set 1000 as a large value
2493 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2494 systems. I've change it so that if it can't get 1000, it tries for 256.
2496 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2497 was an oversight, and furthermore, ever since the addition of extra
2498 controls (e.g. 4.43/32), the checks on when to allow different forms of
2499 "control" were broken. There should now be diagnostics for all cases when a
2500 control that does not make sense is encountered.
2502 9. Added the /retain_sender option to "control=submission".
2504 10. $recipients is now available in the predata ACL (oversight).
2506 11. Tidy the search cache before the fork to do a delivery from a message
2507 received from the command line. Otherwise the child will trigger a lookup
2508 failure and thereby defer the delivery if it tries to use (for example) a
2509 cached ldap connection that the parent has called unbind on.
2511 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2512 of $address_data from the recipient verification was clobbered by the
2513 sender verification.
2515 13. The value of address_data from a sender verification is now available in
2516 $sender_address_data in subsequent conditions in the ACL statement.
2518 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2520 15. Added a new option "connect=<time>" to callout options, to set a different
2523 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2524 was its contents. (It was OK if the option was not defined at all.)
2526 17. A "Completed" log line is now written for messages that are removed from
2527 the spool by the -Mrm option.
2529 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2530 information about exactly what failed.
2532 19. Added -dd to debug only the daemon process.
2534 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2535 handles timeouts, both on the server side and network timeouts. Renamed the
2536 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2538 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2539 It is now set to "smtps".
2541 22. $host_address is now set to the target address during the checking of
2542 ignore_target_hosts.
2544 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2545 being passed; this would have caused $sender_host_name to have been used if
2546 matching the list had actually called for a host name (not very likely,
2547 since this list is usually IP addresses). A host name is now passed as
2550 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2551 code that specifies a non-restarting handler (typically sigaction() in
2552 modern systems) in an attempt to fix a rare and obscure crash bug.
2554 25. Narrowed the window for a race in the daemon that could cause it to ignore
2555 SIGCHLD signals. This is not a major problem, because they are used only to
2556 wake it up if nothing else does.
2558 26. A malformed maildirsize file could cause Exim to calculate negative values
2559 for the mailbox size or file count. Odd effects could occur as a result.
2560 The maildirsize information is now recalculated if the size or filecount
2563 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2564 support for a long time. Removed HAVE_SYS_VFS_H.
2566 28. Installed the latest version of exipick from John Jetmore.
2568 29. In an address list, if the pattern was not a regular expression, an empty
2569 subject address (from a bounce message) matched only if the pattern was an
2570 empty string. Non-empty patterns were not even tested. This was the wrong
2571 because it is perfectly reasonable to use an empty address as part of a
2572 database query. An empty address is now tested by patterns that are
2573 lookups. However, all the other forms of pattern expect the subject to
2574 contain a local part and a domain, and therefore, for them, an empty
2575 address still always fails if the pattern is not itself empty.
2577 30. Exim went into a mad DNS loop when attempting to do a callout where the
2578 host was specified on an smtp transport, and looking it up yielded more
2579 than one IP address.
2581 31. Re-factored the code for checking spool and log partition space into a
2582 function that finds that data and another that does the check. The former
2583 is then used to implement four new variables: $spool_space, $log_space,
2584 $spool_inodes, and $log_inodes.
2586 32. The RFC2047 encoding function was originally intended for short strings
2587 such as real names; it was not keeping to the 75-character limit for
2588 encoded words that the RFC imposes. It now respects the limit, and
2589 generates multiple encoded words if necessary. To be on the safe side, I
2590 have increased the buffer size for the ${rfc2047: expansion operator from
2593 33. It is now permitted to omit both strings after an "if" condition; if the
2594 condition is true, the result is "true". As before, when the second string
2595 is omitted, a false condition yields an empty string. This makes it less
2596 cumbersome to write custom ACL and router conditions.
2598 34. Failure to deliver a bounce message always caused it to be frozen, even if
2599 there was an errors_to setting on the router. The errors_to setting is now
2602 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2603 canonical form (fully expanded) before being placed in
2604 $sender_host_address.
2606 36. The table in the code that translates DNS record types into text (T_A to
2607 "A" for instance) was missing entries for NS and CNAME. It is just possible
2608 that this could have caused confusion if both these types were looked up
2609 for the same domain, because the text type is used as part of Exim's
2610 per-process caching. But the chance of anyone hitting this buglet seems
2613 37. The dnsdb lookup has been extended in a number of ways.
2615 (1) There is a new type, "zns", which walks up the domain tree until it
2616 finds some nameserver records. It should be used with care.
2618 (2) There is a new type, "mxh", which is like "mx" except that it returns
2619 just the host names, not the priorities.
2621 (3) It is now possible to give a list of domains (or IP addresses) to be
2622 looked up. The behaviour when one of the lookups defers can be
2623 controlled by a keyword.
2625 (4) It is now possible to specify the separator character for use when
2626 multiple records are returned.
2628 38. The dnslists ACL condition has been extended: it is now possible to supply
2629 a list of IP addresses and/or domains to be looked up in a particular DNS
2632 39. Added log_selector=+queue_time_overall.
2634 40. When running the queue in the test harness, wait just a tad after forking a
2635 delivery process, to get repeatability of debugging output.
2637 41. Include certificate and key file names in error message when GnuTLS fails
2638 to set them up, because the GnuTLS error message doesn't include the name
2639 of the failing file when there is a problem reading it.
2641 42. Allow both -bf and -bF in the same test run.
2643 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2645 44. The "Exiscan patch" is now merged into the mainline Exim source.
2647 45. Sometimes the final signoff response after QUIT could fail to get
2648 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2649 < 0 before doing a fflush(). This bug looks as though it goes back to the
2650 introduction of TLS in release 3.20, but "sometimes" must have been rare
2651 because the tests only now provoked it.
2653 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2654 (this can affect the format of dates).
2656 47. exim_tidydb, when checking for the continued existence of a message for
2657 which it has found a message-specific retry record, was not finding
2658 messages that were in split spool directories. Consequently, it was
2659 deleting retry records that should have stayed in existence.
2661 48. Steve fixed some bugs in eximstats.
2663 49. The SPA authentication driver was not abandoning authentication and moving
2664 on to the next authenticator when an expansion was forced to fail,
2665 contradicting the general specification for all authenticators. Instead it
2666 was generating a temporary error. It now behaves as specified.
2668 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2669 (the order specifies the preference for clients). The order is now AES256,
2670 AES128, 3DES, ARCFOUR128.
2672 51. Small patch to Sieve code - explicitly set From: when generating an
2675 52. Exim crashed if a remote delivery caused a very long error message to be
2676 recorded - for instance if somebody sent an entire SpamAssassin report back
2677 as a large number of 550 error lines. This bug was coincidentally fixed by
2678 increasing the size of one of Exim's internal buffers (big_buffer) that
2679 happened as part of the Exiscan merge. However, to be on the safe side, I
2680 have made the code more robust (and fixed the comments that describe what
2683 53. Now that there can be additional text after "Completed" in log lines (if
2684 the queue_time_overall log selector is set), a one-byte patch to exigrep
2685 was needed to allow it to recognize "Completed" as not the last thing in
2688 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2689 patch that reportedly fixes this has been added. I am not expert enough to
2690 create a test for it. This is what the patch creator wrote:
2692 "I found a little strange behaviour of ldap code when working with
2693 Windows 2003 AD Domain, where users was placed in more than one
2694 Organization Units. When I tried to give exim partial DN, the exit code
2695 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2696 But simultaneously result of request was absolutely normal ldap result,
2697 so I produce this patch..."
2699 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2700 REFERENCE, so I have modified the code to exclude the patch when that macro
2703 55. Some experimental protocols are using DNS PTR records for new purposes. The
2704 keys for these records are domain names, not reversed IP addresses. The
2705 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2706 leaves it alone. Component reversal etc. now happens only for IP addresses.
2709 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2711 57. Double the size of the debug message buffer (to 2048) so that more of very
2712 long debug lines gets shown.
2714 58. The exicyclog utility now does better if the number of log files to keep
2715 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2717 59. Two changes related to the smtp_active_hostname option:
2719 (1) $smtp_active_hostname is now available as a variable.
2720 (2) The default for smtp_banner uses $smtp_active_hostname instead
2721 of $primary_hostname.
2723 60. The host_aton() function is supposed to be passed a string that is known
2724 to be a valid IP address. However, in the case of IPv6 addresses, it was
2725 not checking this. This is a hostage to fortune. Exim now panics and dies
2726 if the condition is not met. A case was found where this could be provoked
2727 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2728 components; fortuitously, this particular loophole had already been fixed
2729 by change 4.50/55 above.
2731 If there are any other similar loopholes, the new check in host_aton()
2732 itself should stop them being exploited. The report I received stated that
2733 data on the command line could provoke the exploit when Exim was running as
2734 exim, but did not say which command line option was involved. All I could
2735 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2736 running as the user.
2739 61. There was a buffer overflow vulnerability in the SPA authentication code
2740 (which came originally from the Samba project). I have added a test to the
2741 spa_base64_to_bits() function which I hope fixes it.
2744 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2745 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2747 63. The daemon start-up calls getloadavg() while still root for those OS that
2748 need the first call to be done as root, but it missed one case: when
2749 deliver_queue_load_max is set with deliver_drop_privilege. This is
2750 necessary for the benefit of the queue runner, because there is no re-exec
2751 when deliver_drop_privilege is set.
2753 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2754 This has been fixed.
2756 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2757 was in use, was not putting the data itself into the right store pool;
2758 consequently, it could be overwritten for a subsequent message in the same
2759 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2762 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2764 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2765 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2766 the function were treating the return as a boolean value, which happened to
2767 work because 0=false and not-0=true, but is not correct code.
2769 68. The host_aton() function was not handling scoped IPv6 addresses (those
2770 with, for example, "%eth0" on the end) correctly.
2772 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2773 negated items (that is, ~something) in unsigned ints. Some compilers
2774 apparently mutter when there is no cast.
2776 70. If an address verification called from an ACL failed, and did not produce a
2777 user-specific message (i.e. there was only a "system" message), nothing was
2778 put in $acl_verify_message. In this situation, it now puts the system
2781 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2782 session; change 4.31/43 added the unwanted input to the log line - except
2783 that it did not do this in the start of session case. It now does.
2785 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2786 This is wrong and can cause the other end to generate a synchronization
2787 error if it is another Exim or anything else that does the synchronization
2788 check. A QUIT command is no longer sent after a timeout.
2790 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2791 during host lookups.
2793 74. The defer_ok option of callout verification was not working if it was used
2794 when verifying addresses in header lines, that is, for this case:
2796 verify = header_sender/callout=defer_ok
2798 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2799 those file descriptors could be used for SMTP connections. If anything
2800 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2801 could be sent to the SMTP client, causing chaos. The daemon now opens
2802 stdin, stdout, and stderr to /dev/null when it puts itself into the
2805 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2806 log by default. The user can override this with suitable Perl magic.
2808 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2809 the log message when discard triggers, was not working for the DATA ACL or
2810 for the non-SMTP ACL.
2812 78. Error message wording change in sieve.c.
2814 79. If smtp_accept_max_per_host was set, the number of connections could be
2815 restricted to fewer than expected, because the daemon was trying to set up
2816 a new connection before checking whether the processes handling previous
2817 connections had finished. The check for completed processes is now done
2818 earlier. On busy systems, this bug wouldn't be noticed because something
2819 else would have woken the daemon, and it would have reaped the completed
2822 80. If a message was submitted locally by a user whose login name contained one
2823 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2824 It caused a spool format error. I have fixed the spool reading code. A
2825 related problem was that the "from" clause in the Received: line became
2826 illegal because of the space(s). It is now covered by ${quote_local_part.
2828 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2831 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2833 83. Patch from Sieve maintainer to fix unterminated string problem in
2834 "vacation" handling.
2836 84. Some minor changes to the Linux configuration files to help with other
2837 OS variants using glibc.
2839 85. One more patch for Sieve to update vacation handling to latest spec.
2842 ----------------------------------------------------
2843 See the note above about the 4.44 and 4.50 releases.
2844 ----------------------------------------------------
2850 1. Change 4.43/35 introduced a bug that caused file counts to be
2851 incorrectly computed when quota_filecount was set in an appendfile
2854 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2855 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2858 3. Give more explanation in the error message when the command for a transport
2859 filter fails to execute.
2861 4. There are several places where Exim runs a non-Exim command in a
2862 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2863 was being done only for the command run by the queryprogram router. It is
2864 now done for all such subprocesses. The other cases are: ${run, transport
2865 filters, and the commands run by the lmtp and pipe transports.
2867 5. Some older OS have a limit of 256 on the maximum number of file
2868 descriptors. Exim was using setrlimit() to set 1000 as a large value
2869 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2870 systems. I've change it so that if it can't get 1000, it tries for 256.
2872 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2873 was an oversight, and furthermore, ever since the addition of extra
2874 controls (e.g. 4.43/32), the checks on when to allow different forms of
2875 "control" were broken. There should now be diagnostics for all cases when a
2876 control that does not make sense is encountered.
2878 7. $recipients is now available in the predata ACL (oversight).
2880 8. Tidy the search cache before the fork to do a delivery from a message
2881 received from the command line. Otherwise the child will trigger a lookup
2882 failure and thereby defer the delivery if it tries to use (for example) a
2883 cached ldap connection that the parent has called unbind on.
2885 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2886 of $address_data from the recipient verification was clobbered by the
2887 sender verification.
2889 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2890 was its contents. (It was OK if the option was not defined at all.)
2892 11. A "Completed" log line is now written for messages that are removed from
2893 the spool by the -Mrm option.
2895 12. $host_address is now set to the target address during the checking of
2896 ignore_target_hosts.
2898 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2899 being passed; this would have caused $sender_host_name to have been used if
2900 matching the list had actually called for a host name (not very likely,
2901 since this list is usually IP addresses). A host name is now passed as
2904 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2905 code that specifies a non-restarting handler (typically sigaction() in
2906 modern systems) in an attempt to fix a rare and obscure crash bug.
2908 15. Narrowed the window for a race in the daemon that could cause it to ignore
2909 SIGCHLD signals. This is not a major problem, because they are used only to
2910 wake it up if nothing else does.
2912 16. A malformed maildirsize file could cause Exim to calculate negative values
2913 for the mailbox size or file count. Odd effects could occur as a result.
2914 The maildirsize information is now recalculated if the size or filecount
2917 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2918 support for a long time. Removed HAVE_SYS_VFS_H.
2920 18. Updated exipick to current release from John Jetmore.
2922 19. Allow an empty sender to be matched against a lookup in an address list.
2923 Previously the only cases considered were a regular expression, or an
2926 20. Exim went into a mad DNS lookup loop when doing a callout where the
2927 host was specified on the transport, if the DNS lookup yielded more than
2930 21. The RFC2047 encoding function was originally intended for short strings
2931 such as real names; it was not keeping to the 75-character limit for
2932 encoded words that the RFC imposes. It now respects the limit, and
2933 generates multiple encoded words if necessary. To be on the safe side, I
2934 have increased the buffer size for the ${rfc2047: expansion operator from
2937 22. Failure to deliver a bounce message always caused it to be frozen, even if
2938 there was an errors_to setting on the router. The errors_to setting is now
2941 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2942 canonical form (fully expanded) before being placed in
2943 $sender_host_address.
2945 24. Updated eximstats to version 1.33
2947 25. Include certificate and key file names in error message when GnuTLS fails
2948 to set them up, because the GnuTLS error message doesn't include the name
2949 of the failing file when there is a problem reading it.
2951 26. Expand error message when OpenSSL has problems setting up cert/key files.
2954 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2955 (this can affect the format of dates).
2957 28. exim_tidydb, when checking for the continued existence of a message for
2958 which it has found a message-specific retry record, was not finding
2959 messages that were in split spool directories. Consequently, it was
2960 deleting retry records that should have stayed in existence.
2962 29. eximstats updated to version 1.35
2963 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2964 1.35 - bugfix such that pie charts by volume are generated correctly
2966 30. The SPA authentication driver was not abandoning authentication and moving
2967 on to the next authenticator when an expansion was forced to fail,
2968 contradicting the general specification for all authenticators. Instead it
2969 was generating a temporary error. It now behaves as specified.
2971 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2972 (the order specifies the preference for clients). The order is now AES256,
2973 AES128, 3DES, ARCFOUR128.
2975 31. Small patch to Sieve code - explicitly set From: when generating an
2978 32. Exim crashed if a remote delivery caused a very long error message to be
2979 recorded - for instance if somebody sent an entire SpamAssassin report back
2980 as a large number of 550 error lines. This bug was coincidentally fixed by
2981 increasing the size of one of Exim's internal buffers (big_buffer) that
2982 happened as part of the Exiscan merge. However, to be on the safe side, I
2983 have made the code more robust (and fixed the comments that describe what
2986 33. Some experimental protocols are using DNS PTR records for new purposes. The
2987 keys for these records are domain names, not reversed IP addresses. The
2988 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2989 leaves it alone. Component reversal etc. now happens only for IP addresses.
2992 34. The host_aton() function is supposed to be passed a string that is known
2993 to be a valid IP address. However, in the case of IPv6 addresses, it was
2994 not checking this. This is a hostage to fortune. Exim now panics and dies
2995 if the condition is not met. A case was found where this could be provoked
2996 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2997 components; fortuitously, this particular loophole had already been fixed
2998 by change 4.50/55 or 4.44/33 above.
3000 If there are any other similar loopholes, the new check in host_aton()
3001 itself should stop them being exploited. The report I received stated that
3002 data on the command line could provoke the exploit when Exim was running as
3003 exim, but did not say which command line option was involved. All I could
3004 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3005 running as the user.
3008 35. There was a buffer overflow vulnerability in the SPA authentication code
3009 (which came originally from the Samba project). I have added a test to the
3010 spa_base64_to_bits() function which I hope fixes it.
3013 36. The daemon start-up calls getloadavg() while still root for those OS that
3014 need the first call to be done as root, but it missed one case: when
3015 deliver_queue_load_max is set with deliver_drop_privilege. This is
3016 necessary for the benefit of the queue runner, because there is no re-exec
3017 when deliver_drop_privilege is set.
3019 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3020 was in use, was not putting the data itself into the right store pool;
3021 consequently, it could be overwritten for a subsequent message in the same
3022 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3025 38. Sometimes the final signoff response after QUIT could fail to get
3026 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3027 < 0 before doing a fflush(). This bug looks as though it goes back to the
3028 introduction of TLS in release 3.20, but "sometimes" must have been rare
3029 because the tests only now provoked it.
3035 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3036 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3037 Now it returns an integer. A number of calls were still expecting a T/F
3038 return. Fortuitously, in all cases, the tests worked in OK situations,
3039 which is the norm. However, things would have gone wrong on any write
3040 failures on the smtp file descriptor. This function is used when sending
3041 messages over SMTP and also when doing verify callouts.
3043 2. When Exim is called to do synchronous delivery of a locally submitted
3044 message (the -odf or -odi options), it no longer closes stderr before doing
3047 3. Implemented the mua_wrapper option.
3049 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3051 5. Implemented the functions header_remove(), header_testname(),
3052 header_add_at_position(), and receive_remove_recipient(), and exported them
3055 6. If an ACL "warn" statement specified the addition of headers, Exim already
3056 inserted X-ACL-Warn: at the start if there was no header name. However, it
3057 was not making this test for the second and subsequent header lines if
3058 there were newlines in the string. This meant that an invalid header could
3059 be inserted if Exim was badly configured.
3061 7. Allow an ACL "warn" statement to add header lines at the start or after all
3062 the Received: headers, as well as at the end.
3064 8. Added the rcpt_4xx retry error code.
3066 9. Added postmaster_mailfrom=xxx to callout verification option.
3068 10. Added mailfrom=xxxx to the callout verification option, for verify=
3071 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3072 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3074 12. Inserted some casts to stop certain compilers warning when using pointer
3075 differences as field lengths or precisions in printf-type calls (mostly
3076 affecting debugging statements).
3078 13. Added optional readline() support for -be (dynamically loaded).
3080 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3081 same clock tick as a message's arrival, so that its received time was the
3082 same as the "first fail" time on the retry record, and that message
3083 remained on the queue past the ultimate address timeout, every queue runner
3084 would try a delivery (because it was past the ultimate address timeout) but
3085 after another failure, the ultimate address timeout, which should have then
3086 bounced the address, did not kick in. This was a "< instead of <=" error;
3087 in most cases the first failure would have been in the next clock tick
3088 after the received time, and all would be well.
3090 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3091 being recognized when the domain list was tested by the match_domain
3092 condition in an expansion string.
3094 16. Added the ${str2b64: operator.
3096 17. Exim was always calling setrlimit() to set a large limit for the number of
3097 processes, without checking whether the existing limit was already
3098 adequate. (It did check for the limit on file descriptors.) Furthermore,
3099 errors from getrlimit() and setrlimit() were being ignored. Now they are
3100 logged to the main and panic logs, but Exim does carry on, to try to do its
3101 job under whatever limits there are.
3103 18. Imported PCRE 5.0.
3105 19. Trivial typo in log message " temporarily refused connection" (the leading
3108 20. If the log selector return_path_on_delivery was set and an address was
3109 redirected to /dev/null, the delivery process crashed because it assumed
3110 that a return path would always be set for a "successful" delivery. In this
3111 case, the whole delivery is bypassed as an optimization, and therefore no
3114 21. Internal re-arrangement: the function for sending a challenge and reading
3115 a response while authentication was assuming a zero-terminated challenge
3116 string. It's now changed to take a pointer and a length, to allow for
3117 binary data in such strings.
3119 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3121 23. Exim was not respecting finduser_retries when seeking the login of the
3122 uid under which it was called; it was always trying 10 times. (The default
3123 setting of finduser_retries is zero.) Also, it was sleeping after the final
3124 failure, which is pointless.
3126 24. Implemented tls_on_connect_ports.
3128 25. Implemented acl_smtp_predata.
3130 26. If the domain in control=submission is set empty, Exim assumes that the
3131 authenticated id is a complete email address when it generates From: or
3132 Sender: header lines.
3134 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3135 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3136 chown and chgrp in /bin and hostname in /usr/bin.
3138 28. Exim was keeping the "process log" file open after each use, just as it
3139 does for the main log. This opens the possibility of it remaining open for
3140 long periods when the USR1 signal hits a daemon. Occasional processlog
3141 errors were reported, that could have been caused by this. Anyway, it seems
3142 much more sensible not to leave this file open at all, so that is what now
3145 29. The long-running daemon process does not normally write to the log once it
3146 has entered its main loop, and it closes the log before doing so. This is
3147 so that log files can straightforwardly be renamed and moved. However,
3148 there are a couple of unusual error situations where the daemon does write
3149 log entries, and I had neglected to close the log afterwards.
3151 30. The text of an SMTP error response that was received during a remote
3152 delivery was being truncated at 512 bytes. This is too short for some of
3153 the long messages that one sometimes sees. I've increased the limit to
3156 31. It is now possible to make retry rules that apply only when a message has a
3157 specific sender, in particular, an empty sender.
3159 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3160 it possible to be selective about when SMTP synchronization is enforced.
3162 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3164 32. Implemented hosts_connection_nolog.
3166 33. Added an ACL for QUIT.
3168 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3171 35. Added mailbox_size and mailbox_filecount to appendfile.
3173 36. Added control = no_multiline_responses to ACLs.
3175 37. There was a bug in the logic of the code that waits for the clock to tick
3176 in the case where the clock went backwards by a substantial amount such
3177 that the microsecond fraction of "now" was more than the microsecond
3178 fraction of "then" (but the whole seconds number was less).
3180 38. Added support for the libradius Radius client library this is found on
3181 FreeBSD (previously only the radiusclient library was supported).
3187 1. When certain lookups returned multiple values in the form name=value, the
3188 quoting of the values was not always being done properly. Specifically:
3189 (a) If the value started with a double quote, but contained no whitespace,
3191 (b) If the value contained whitespace other than a space character (i.e.
3192 tabs or newlines or carriage returns) it was not quoted.
3193 This fix has been applied to the mysql and pgsql lookups by writing a
3194 separate quoting function and calling it from the lookup code. The fix
3195 should probably also be applied to nisplus, ibase and oracle lookups, but
3196 since I cannot test any of those, I have not disturbed their existing code.
3198 2. A hit in the callout cache for a specific address caused a log line with no
3199 reason for rejecting RCPT. Now it says "Previous (cached) callout
3200 verification failure".
3202 3. There was an off-by-one bug in the queryprogram router. An over-long
3203 return line was truncated at 256 instead of 255 characters, thereby
3204 overflowing its buffer with the terminating zero. As well as fixing this, I
3205 have increased the buffer size to 1024 (and made a note to document this).
3207 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3208 when Exim is waiting for an SMTP response from a remote server, Exim
3209 restarts its select() call on the socket, thereby resetting its timeout.
3210 This is not a problem when such interrupts are rare. Somebody set up a cron
3211 job to run exiwhat every 2 minutes, which is less than the normal select()
3212 timeout (5 or 10 minutes). This meant that the select() timeout never
3213 kicked in because it was always reset. I have fixed this by comparing the
3214 time when an interrupt arrives with the time at the start of the first call
3215 to select(). If more time than the timeout has elapsed, the interrupt is
3216 treated as a timeout.
3218 5. Some internal re-factoring in preparation for the addition of Sieve
3219 extensions (by MH). In particular, the "personal" test is moved to a
3220 separate function, and given an option for scanning Cc: and Bcc: (which is
3221 not set for Exim filters).
3223 6. When Exim created an email address using the login of the caller as the
3224 local part (e.g. when creating a From: or Sender: header line), it was not
3225 quoting the local part when it contained special characters such as @.
3227 7. Installed new OpenBSD configuration files.
3229 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3230 try to make them clearer.
3232 9. Callout options, other than the timeout value, were being ignored when
3233 verifying sender addresses in header lines. For example, when using
3235 verify = header_sender/callout=no_cache
3237 the cache was (incorrectly) being used.
3239 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3240 only the Cygwin environment.
3242 11. When return_path_on_delivery was set as a log selector, if different remote
3243 addresses in the same message used different return paths and parallel
3244 remote delivery occurred, the wrong values would sometimes be logged.
3245 (Whenever a remote delivery process finished, the return path value from
3246 the most recently started remote delivery process was logged.)
3248 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3249 lines when AUTH and/or TLS are in use. This is the "received protocol"
3250 field. Exim used to use "asmtp" for authenticated SMTP, without any
3251 indication (in the protocol name) for TLS use. Now it follows the RFC and
3252 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3253 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3254 names appear in log lines as well as in Received: header lines.
3256 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3257 extensions, and comparison tests, and to fix some bugs.
3259 14. Changes to the "personal" filter test:
3261 (1) The test was buggy in that it was just doing the equivalent of
3262 "contains" tests on header lines. For example, if a user's address was
3263 anne@some.where, the "personal" test would incorrectly be true for
3265 To: susanne@some.where
3267 This test is now done by extracting each address from the header in turn,
3268 and checking the entire address. Other tests that are part of "personal"
3269 are now done using regular expressions (for example, to check local parts
3270 of addresses in From: header lines).
3272 (2) The list of non-personal local parts in From: addresses has been
3273 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3274 taken from the Sieve specification recommendations.
3276 (3) If the message contains any header line starting with "List-" it is
3277 treated as non-personal.
3279 (4) The test for "circular" in the Subject: header line has been removed
3280 because it now seems ill-conceived.
3282 15. Minor typos in src/EDITME comments corrected.
3284 16. Installed latest exipick from John Jetmore.
3286 17. If headers_add on a router specified a text string that was too long for
3287 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3288 of string_sprintf() is now avoided.
3290 18. $message_body_size was not set (it was always zero) when running the DATA
3291 ACL and the local_scan() function.
3293 19. For the "mail" command in an Exim filter, no default was being set for
3294 the once_repeat time, causing a random time value to be used if "once" was
3295 specified. (If the value happened to be <= 0, no repeat happened.) The
3296 default is now 0s, meaning "never repeat". The "vacation" command was OK
3297 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3298 (I found it when inspecting the code).
3300 20. There is now an overall timeout for performing a callout verification. It
3301 defaults to 4 times the callout timeout, which applies to individual SMTP
3302 commands during the callout. The overall timeout applies when there is more
3303 than one host that can be tried. The timeout is checked before trying the
3304 next host. This prevents very long delays if there are a large number of
3305 hosts and all are timing out (e.g. when the network connections are timing
3306 out). The value of the overall timeout can be changed by specifying an
3307 additional sub-option for "callout", called "maxwait". For example:
3309 verify = sender/callout=5s,maxwait=20s
3311 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3312 to the end before writing, but this should make it even safer).
3314 22. Exim was forgetting that it had advertised PIPELINING for the second and
3315 subsequent messages on an SMTP connection. It was also not resetting its
3316 memory on STARTTLS and an internal HELO.
3318 23. When Exim logs an SMTP synchronization error within a session, it now
3319 records whether PIPELINING has been advertised or not.
3321 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3322 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3323 rather than long int.
3325 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3327 26. Added the never_mail option to autoreply.
3333 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3334 crash if the getsockname() call failed; this can happen if a connection is
3335 closed very soon after it is established. The problem was simply in the
3336 order in which certain operations were done, causing Exim to try to write
3337 to the SMTP stream before it had set up the file descriptor. The bug has
3338 been fixed by making things happen in the correct order.
3344 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3345 before the connection was closed, thus losing the rejection response.
3347 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3348 some early Solaris releases, but causes trouble in current releases where
3349 socklen_t is defined.
3351 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3354 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3357 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3358 that releasing the top of it at the end releases what was used for sub-
3359 expansions (unless the block got too big). However, discard this block if
3360 the first thing is a variable or header, so that we can use its block when
3361 it is dynamic (useful for very large $message_headers, for example).
3363 6. Lookups now cache *every* query, not just the most recent. A new, separate
3364 store pool is used for this. It can be recovered when all lookup caches are
3365 flushed. Lookups now release memory at the end of their result strings.
3366 This has involved some general refactoring of the lookup sources.
3368 7. Some code has been added to the store_xxx() functions to reduce the amount
3369 of flapping under certain conditions.
3371 8. log_incoming_interface used to affect only the <= reception log lines. Now
3372 it causes the local interface and port to be added to several more SMTP log
3373 lines, for example "SMTP connection from", and rejection lines.
3375 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3377 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3379 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3380 could be overwritten at the end of the current message (or the start of a
3381 new message if it was set in a HELO ACL). The value is now preserved for
3382 the duration of the SMTP connection.
3384 12. If a transport had a headers_rewrite setting, and a matching header line
3385 contained an unqualified address, that address was qualified, even if it
3386 did not match any rewriting rules. The underlying bug was that the values
3387 of the flags that permit the existence of unqualified sender and recipient
3388 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3389 non-local messages, and by -bnq for local messages) were not being
3390 preserved with the message after it was received.
3392 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3393 "next input=" as part of the text comprising the host identity instead of
3394 the correct text. The code was using the same buffer for two different
3395 strings. However, depending on which order the printing function evaluated
3396 its arguments, the bug did not always show up. Under Linux, for example, my
3397 test suite worked just fine.
3399 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3400 doesn't work with some older versions of Perl. It has been changed to "my",
3401 which in any case is probably the better facility to use.
3403 15. A really picky compiler found some instances of statements for creating
3404 error messages that either had too many or two few arguments for the format
3407 16. The size of the buffer for calls to the DNS resolver has been increased
3408 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3409 for addresses that have a lot of PTR records. This alleviates a problem; it
3410 does not fully solve it.
3412 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3413 buffer now truncates the list and logs the incident, which is the same
3414 action as happens when Exim is looking up a host name and its aliases.
3415 Previously in this situation something unpredictable would happen;
3416 sometimes it was "internal error: store_reset failed".
3418 18. If a server dropped the connection unexpectedly when an Exim client was
3419 using GnuTLS and trying to read a response, the client delivery process
3420 crashed while trying to generate an error log message.
3422 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3423 string, for example:
3425 warn message = H1: something\nH2: something
3427 the text was added as a single header line from Exim's point of view
3428 though it ended up OK in the delivered message. However, searching for the
3429 second and subsequent header lines using $h_h2: did not work. This has been
3430 fixed. Similarly, if a system filter added multiple headers in this way,
3431 the routers could not see them.
3433 20. Expanded the error message when iplsearch is called with an invalid key to
3434 suggest using net-iplsearch in a host list.
3436 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3437 ACLs are no longer actually imposed (and a message to that effect is
3440 22. If a "gecos" field in a passwd entry contained escaped characters, in
3441 particular, if it contained a \" sequence, Exim got it wrong when building
3442 a From: or a Sender: header from that name. A second bug also caused
3443 incorrect handling when an unquoted " was present following a character
3444 that needed quoting.
3446 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3447 was not being matched caselessly.
3449 24. Arranged for all hyphens in the exim.8 source to be escaped with
3452 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3453 properly. Recipient callout cache records were still being keyed to include
3454 the sender, even when use_sender was set false. This led to far more
3455 callouts that were necessary. The sender is no longer included in the key
3456 when use_sender is false.
3458 26. Added "control = submission" modifier to ACLs.
3460 27. Added the ${base62d: operator to decode base 62 numbers.
3462 28. dnsdb lookups can now access SRV records.
3464 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3465 the configuration file.
3467 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3468 (-v) mode. This makes the output for a verbose queue run more intelligible.
3470 31. Added a use_postmaster feature to recipient callouts.
3472 32. Added the $body_zerocount variable, containing the number of binary zero
3473 bytes in the message body.
3475 33. The time of last modification of the "new" subdirectory is now used as the
3476 "mailbox time last read" when there is a quota error for a maildir
3479 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3481 35. Added +ignore_unknown as a special item in host lists.
3483 36. Code for decoding IPv6 addresses in host lists is now included, even if
3484 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3485 address was recognized as an IP address, but was then not correctly decoded
3486 into binary, causing unexpected and incorrect effects when compared with
3493 1. Very minor rewording of debugging text in manualroute to say "list of
3494 hosts" instead of "hostlist".
3496 2. If verify=header_syntax was set, and a header line with an unqualified
3497 address (no domain) and a large number of spaces between the end of the
3498 name and the colon was received, the reception process suffered a buffer
3499 overflow, and (when I tested it) crashed. This was caused by some obsolete
3500 code that should have been removed. The fix is to remove it!
3502 3. When running in the test harness, delay a bit after writing a bounce
3503 message to get a bit more predictability in the log output.
3505 4. Added a call to search_tidyup() just before forking a reception process. In
3506 theory, someone could use a lookup in the expansion of smtp_accept_max_
3507 per_host which, without the tidyup, could leave open a database connection.
3509 5. Added the variables $recipient_data and $sender_data which get set from a
3510 lookup success in an ACL "recipients" or "senders" condition, or a router
3511 "senders" option, similar to $domain_data and $local_part_data.
3513 6. Moved the writing of debug_print from before to after the "senders" test
3516 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3517 problems for message scanning, either using a data ACL, or using
3518 local_scan() because the Received: header was not generated till after they
3519 were called (in order to set the time as the time of reception completion).
3520 I have revised the way this works. The header is now generated after the
3521 body is received, but before the ACL or local_scan() are called. After they
3522 are run, the timestamp in the header is updated.
3528 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3529 before starting a queue runner without re-exec. This happened only when
3530 deliver_drop_privilege was set or when the Exim user was set to root. The
3531 effect of the bug was that timeouts during subsequent deliveries caused
3532 crashes instead of being properly handled. The handler is now left at its
3533 default (and expected) setting.
3535 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3536 message, again when deliver_drop_privilege is set or Exim is run as root.
3537 The bug described in (1) was not present in this case, but the tidying up
3538 of the other signals was missing. I have made the two cases consistent.
3540 3. The ignore_target_hosts setting on a manualroute router was being ignored
3541 for hosts that were looked up using the /MX notation.
3543 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3546 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3547 operated on the sender address. After changing the $sender_address to <>
3548 for the sender address verify, Exim was re-instated it as the original
3549 (before rewriting) address, but remembering that it had rewritten it, so it
3550 wasn't rewriting it again. This bug also had the effect of breaking the
3551 sender address verification caching when the sender address was rewritten.
3553 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3554 This has been changed so that if the ip literal address matches
3555 ignore_target_hosts, the router declines.
3557 7. Added expansion conditions match_domain, match_address, and match_local_
3558 part (NOT match_host).
3560 8. The placeholder for the Received: header didn't have a length field set.
3562 9. Added code to Exim itself and to exim_lock to test for a specific race
3563 condition that could lead to file corruption when using MBX delivery. The
3564 issue is with the lockfile that is created in /tmp. If this file is removed
3565 after a process has opened it but before that process has acquired a lock,
3566 there is the potential for a second process to recreate the file and also
3567 acquire a lock. This could lead to two Exim processes writing to the file
3568 at the same time. The added code performs the same test as UW imapd; it
3569 checks after acquiring the lock that its file descriptor still refers to
3570 the same named file.
3572 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3573 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3576 11. Added the smtp_active_hostname option. If used, this will typically be made
3577 to depend on the incoming interface address. Because $interface_address is
3578 not set up until the daemon has forked a reception process, error responses
3579 that can happen earlier (such as "too many connections") no longer contain
3582 12. If an expansion in a condition on a "warn" statement fails because a lookup
3583 defers, the "warn" statement is abandoned, and the next ACL statement is
3584 processed. Previously this caused the whole ACL to be aborted.
3586 13. Added the iplsearch lookup type.
3588 14. Added ident_timeout as a log selector.
3590 15. Added tls_certificate_verified as a log selector.
3592 16. Added a global option tls_require_ciphers (compare the smtp transport
3593 option of the same name). This controls incoming TLS connections.
3595 17. I finally figured out how to make tls_require_ciphers do a similar thing
3596 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3597 before starting the TLS session.
3599 18. Tabs are now shown as \t in -bP output.
3601 19. If the log selector return_path_on_delivery was set, Exim crashed when
3602 bouncing a message because it had too many Received: header lines.
3604 20. If two routers both had headers_remove settings, and the first one included
3605 a superfluous trailing colon, the final name in the first list and the
3606 first name in the second list were incorrectly joined into one item (with a
3607 colon in the middle).
3613 1. Added -C and -D options to the exinext utility, mainly to make it easier
3614 to include in the automated testing, but these could be helpful when
3615 multiple configurations are in use.
3617 2. The exinext utility was not formatting the output nicely when there was
3618 an alternate port involved in the retry record key, nor when there was a
3619 message id as well (for retries that were specific to a specific message
3620 and a specific host). It was also confused by IPv6 addresses, because of
3621 the additional colons they contain. I have fixed the IPv4 problem, and
3622 patched it up to do a reasonable job for IPv6.
3624 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3625 delivery, the log line now contains "pipelined" if PIPELINING was used.
3627 4. An SMTP transport process used to panic and die if the bind() call to set
3628 an explicit outgoing interface failed. This has been changed; it is now
3629 treated in the same way as a connect() failure.
3631 5. A reference to $sender_host_name in the part of a conditional expansion
3632 that was being skipped was still causing a DNS lookup. This no longer
3635 6. The def: expansion condition was not recognizing references to header lines
3636 that used bh_ and bheader_.
3638 7. Added the _cache feature to named lists.
3640 8. The code for checking quota_filecount in the appendfile transport was
3641 allowing one more file than it should have been.
3643 9. For compatibility with Sendmail, the command line option
3651 and sets the incoming protocol and host name (for trusted callers). The
3652 host name and its colon can be omitted when only the protocol is to be set.
3653 Note the Exim already has two private options, -pd and -ps, that refer to
3654 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3655 "s", but I don't think that's a major issue.
3657 10. A number of refactoring changes to the code, none of which should affect
3660 (a) The number of logging options was getting close to filling up the
3661 32-bit word that was used as a bit map. I have split them into two classes:
3662 those that are passed in the argument to log_write(), and those that are
3663 only ever tested independently outside of that function. These are now in
3664 separate 32-bit words, so there is plenty of room for expansion again.
3665 There is no change in the user interface or the logging behaviour.
3667 (b) When building, for example, log lines, the code previously used a
3668 macro that called string_cat() twice, in order to add two strings. This is
3669 not really sufficiently general. Furthermore, there was one instance where
3670 it was actually wrong because one of the argument was used twice, and in
3671 one call a function was used. (As it happened, calling the function twice
3672 did not affect the overall behaviour.) The macro has been replaced by a
3673 function that can join an arbitrary number of extra strings onto a growing
3676 (c) The code for expansion conditions now uses a table and a binary chop
3677 instead of a serial search (which was left over from when there were very
3678 few conditions). Also, it now recognizes conditions like "pam" even when
3679 the relevant support is not compiled in: a suitably worded error message is
3680 given if an attempt is made to use such a condition.
3682 11. Added ${time_interval:xxxxx}.
3684 12. A bug was causing one of the ddress fields not to be passed back correctly
3685 from remote delivery subprocesses. The field in question was not being
3686 subsequently used, so this caused to problems in practice.
3688 13. Added new log selectors queue_time and deliver_time.
3690 14. Might have fixed a bug in maildirsizefile handling that threw up
3691 "unexpected character" debug warnings, and recalculated the data
3692 unnecessarily. In any case, I expanded the warning message to give more
3695 15. Added the message "Restricted characters in address" to the statements in
3696 the default ACL that block characters like @ and % in local parts.
3698 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3699 Three changes have been made:
3701 (a) There was a serious bug; a negative response to MAIL caused the whole
3702 recipient domain to be cached as invalid, thereby blocking all messages
3703 to all local parts at the same domain, from all senders. This bug has
3704 been fixed. The domain is no longer cached after a negative response to
3705 MAIL if the sender used is not empty.
3707 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3710 (c) A new callout option, "use_sender" has been added for people who want
3711 the modified behaviour.
3717 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3720 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3721 indeed breaks things for older releases.
3723 3. Added additional logging to the case where there is a problem reading data
3724 from a filter that is running in a subprocess using a pipe, in order to
3725 try to track down a specific problem.
3727 4. Testing facility fudge: when running in the test harness and attempting
3728 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3729 getting "No route to host". Convert this to a timeout.
3731 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3734 6. Some OS don't have socklen_t but use size_t instead. This affects the
3735 fifth argument of getsockopt() amongst other things. This is now
3736 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3737 can be set for individual OS. I have set it for SunOS5, OSF1, and
3738 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3739 some earlier ones do not.
3741 7. Change 4.30/15 was not doing the test caselessly.
3743 8. The standard form for an IPv6 address literal was being rejected by address
3744 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3745 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3746 this, as well as the form without the "IPv6" on the front (but only when
3747 address literals are enabled, of course).
3749 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3751 10. Exim crashed if a message with an empty sender address specified by -f
3752 encountered a router with an errors_to setting. This could be provoked only
3753 by a command such as
3757 where an empty string was supplied; "<>" did not hit this bug.
3759 11. Installed PCRE release 4.5.
3761 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3762 remained set. It is now erased.
3764 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3765 times from message ids (which are base 36 rather than the normal 62).
3767 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3768 were being counted as actual protocol errors, and logged if the log
3769 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3770 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3771 and DATA following a set of rejected RCPTs do not count as protocol errors.
3772 In other words, Exim assumes they were pipelined, though this may not
3773 actually be the case. Of course, in all cases the client gets an
3774 appropriate error code.
3776 15. If a lookup fails in an ACL condition, a message about the failure may
3777 be available; it is used if testing the ACL cannot continue, because most
3778 such messages specify what the cause of the deferral is. However, some
3779 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3780 that caused an old message to be retained and used if a later statement
3781 caused a defer, replacing the real cause of the deferral.
3783 16. If an IP address had so many PTR records that the DNS lookup buffer
3784 was not large enough to hold them, Exim could crash while trying to process
3785 the truncated data. It now detects and logs this case.
3787 17. Further to 4.21/58, another change has been made: if (and only if) the
3788 first line of a message (the first header line) ends with CRLF, a bare LF
3789 in a subsequent header line has a space inserted after it, so as not to
3790 terminate the header.
3792 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3793 unnecessarily, used atoi() instead of strtol(), and didn't check the
3794 termination when getting file sizes from file names by regex.
3796 19. Completely re-implemented the support for maildirsize files, in the light
3797 of a number of problems with the previous contributed implementation
3798 (4.30/29). In particular:
3800 . If the quota is zero, the maildirsize file is maintained, but no quota is
3803 . If the maildir directory does not exist, it is created before any attempt
3804 to write a maildirsize file.
3806 . The quota value in the file is just a cache; if the quota is changed in
3807 the transport, the new value overrides.
3809 . A regular expression is available for excluding directories from the
3812 20. The autoreply transport checks the characters in options that define the
3813 message's headers; it allows continued headers, but it was checking with
3814 isspace() after an embedded newline instead of explicitly looking for a
3817 21. If all the "regular" hosts to which an address was routed had passed their
3818 expiry times, and had not reached their retry times, the address was
3819 bounced, even if fallback hosts were defined. Now Exim should go on to try
3822 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3823 equivalent code in the SMTP transport. Some hosts send humungous responses
3824 to HELO/EHLO, more than 1024 it seems.
3826 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3827 gives compiler warnings in some environments. I've now done it "properly",
3830 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3831 (because of problems with the built-in one) was declared to return uschar *
3832 instead of char *, causing compiler failure.
3834 25. Fixed a file descriptor leak when processing alias/forward files.
3836 26. Fixed a minor format string issue in dbfn.c.
3838 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3840 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3841 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3844 29. When a sender address is verified, it is cached, to save repeating the test
3845 when there is more than one recipient in a message. However, when the
3846 verification involves a callout, it is possible for different callout
3847 options to be set for different recipients. It is too complicated to keep
3848 track of this in the cache, so now Exim always runs a verification when a
3849 callout is required, relying on the callout cache for the optimization.
3850 The overhead is duplication of the address routing, but this should not be
3853 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3854 to be verified with callout=postmaster, and the main callout worked but the
3855 postmaster check failed, the verification correctly failed. However, if a
3856 subsequent RCPT command asked for sender verification *without* the
3857 postmaster check, incorrect caching caused this verification also to fail,
3860 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3861 it was not caching the DNS options (qualify_single, search_parents) that
3862 were used when the lookup failed. A subsequent lookup with different
3863 options therefore always gave the same answer, though there were cases
3864 where it should not have. (Example: a "domains = !$mx_any" option on a
3865 dnslookup router: the "domains" option is always processed without any
3866 widening, but the router might have qualify_single set.) Now Exim uses the
3867 cached value only when the same options are set.
3869 32. Added John Jetmore's "exipick" utility to the distribution.
3871 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3872 than a timeout (e.g. a certificate is required, and is not provided), an
3873 Exim server now closes the connection immediately. Previously it waited for
3874 the client to close - but if the client is SSL, it seems that they each
3875 wait for each other, leading to a delay before one of them times out.
3877 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3878 maintained 0.8.x compatibility because I don't think many are using it, and
3879 it is clearly obsolete.
3881 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3884 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3885 client certificate was expired. A simple patch fixes this, though I don't
3886 understand the full logic of why the verify callback is called multiple
3889 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3890 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3891 which causes problems with some clients (such as the Certicom SSL Plus
3892 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3893 disables the coutermeasure allowing Eudora to connect."
3895 38. Exim was not checking that a write() to a log file succeeded. This could
3896 lead to Bad Things if a log got too big, in particular if it hit a file
3897 size limit. Exim now panics and dies if it cannot write to a log file, just
3898 as it does if it cannot open a log file.
3900 39. Modified OS/Makefile-Linux so that it now contains
3902 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3904 The two -D definitions ensure that Exim is compiled with large file
3905 support, which makes it possible to handle log files that are bigger than
3908 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3909 instance) a domain was checked against a named list that involved a lookup,
3910 causing $domain_data to be set, then another domain was checked against the
3911 same list, then the first domain was re-checked, the value of $domain_data
3912 after the final check could be wrong. In particular, if the second check
3913 failed, it could be set empty. This bug probably also applied to
3916 41. The strip_trailing_dot option was not being applied to the address given
3917 with the -f command-line option.
3919 42. The code for reading a message's header from the spool was incrementing
3920 $received_count, but never initializing it. This meant that the value was
3921 incorrect (doubled) while delivering a message in the same process in which
3922 it was received. In the most common configuration of Exim, this never
3923 happens - a fresh exec is done - but it can happen when
3924 deliver_drop_privilege is set.
3926 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3927 it now includes up to 150 characters of the unexpected data in the log
3930 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3931 and building data strings. The size of both of these buffers was 10 000
3932 bytes - far larger than anybody would *ever* want, thought I. Needless to
3933 say, somebody hit the limit. I have increased the maximum line length to
3934 20 000 and the maximum data length of concatenated lines to 100 000. I have
3935 also fixed two bugs, because there was no checking on these buffers. Tsk,
3936 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3937 buffer is too small.
3939 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3940 lsearch lookups. Now it does.
3942 46. When parsing a route_list item in a manualroute router, a fixed-length
3943 buffer was used for the list of hosts. I made this 1024 bytes long,
3944 thinking that nobody would ever have a list of hosts that long. Wrong.
3945 Somebody had a whole pile of complicated expansion conditions, and the
3946 string was silently truncated, leading to an expansion error. It turns out
3947 that it is easier to change to an unlimited length (owing to other changes
3948 that have happened since this code was originally written) than to build
3949 structure for giving a limitation error. The length of the item that
3950 expands into the list of hosts is now unlimited.
3952 47. The lsearch lookup could not handle data where the length of text line was
3953 more than 4095 characters. Such lines were truncated, leading to shortened
3954 data being returned. It should now handle lines of any length.
3956 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3957 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3960 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3961 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3962 no longer happens in comment lines. A list of replacements is now placed
3963 at the head of all of the source files, except those whose only change is
3964 to replace PERL_COMMAND in the very first #! line.
3966 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3967 messages on the queue, with a bottom-up merge sort, using code contributed
3968 by Michael Haardt. This should make operations like -bp somewhat faster on
3969 large queues. It won't affect queue runners, except when queue_run_in_order
3972 51. Installed eximstats 1.31 in the distribution.
3974 52. Added support for SRV lookups to the dnslookup router.
3976 53. If an ACL referred to $message_body or $message_body_end, the value was not
3977 reset for any messages that followed in the same SMTP session.
3979 54. The store-handling optimization for building very long strings was not
3980 differentiating between the different store pools. I don't think this
3981 actually made any difference in practice, but I've tidied it.
3983 55. While running the routers to verify a sender address, $sender_address
3984 was still set to the sender address. This is wrong, because when routing to
3985 send a bounce to the sender, it would be empty. Therefore, I have changed
3986 it so that, while verifying a sender address, $sender_address is set to <>.
3987 (There is no change to what happens when verifying a recipient address.)
3989 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3990 target A or AAAA records (if not already returned) without resetting the
3991 qualify_single or search_parents options of the DNS resolver. These are
3992 inappropriate in this case because the targets of MX and SRV records must
3993 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3994 target that, when qualified, matched something in the local domain. These
3995 two options are now turned off when doing these lookups.
3997 57. It seems that at least some releases of Reiserfs (which does not have the
3998 concept of a fixed number of inodes) returns zero and not -1 for the
3999 number of available inodes. This interacted badly with check_spool_inodes,
4000 which assumed that -1 was the "no such thing" setting. What I have done is
4001 to check that the total number of inodes is greater than zero before doing
4002 the test of how many are available.
4004 58. When a "warn" ACL statement has a log_message modifier, the message is
4005 remembered, and not repeated. This is to avoid a lot of repetition when a
4006 message has many recipients that cause the same warning to be written.
4007 Howewer, Exim was preserving the list of already written lines for an
4008 entire SMTP session, which doesn't seem right. The memory is now reset if a
4009 new message is started.
4011 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4012 debugging output unless log_rewrite was also set.
4014 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4015 of (int)(handle) when we know that handle contains (void *)(-1).
4017 61. The Exim daemon panic-logs an error return when it closes the incoming
4018 connection. However "connection reset by peer" seems to be common, and
4019 isn't really an error worthy of noting specially, so that particular error
4022 62. When Exim is trying to find all the local interfaces, it used to panic and
4023 die if the ioctl to get the interface flags failed. However, it seems that
4024 on at least one OS (Solaris 9) it is possible to have an interface that is
4025 included in the list of interfaces, but for which you get a failure error
4026 for this call. This happens when the interface is not "plumbed" into a
4027 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4028 failure of the "get flags" call assumes that the interface is down.
4030 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4031 makes life easier for people who are doing arithmetic on fields extracted
4032 from dates, where you often get leading zeros that should not be
4033 interpreted as octal.
4035 64. Added qualify_domain to the redirect router, to override the global
4038 65. If a pathologically long header line contained very many addresses (the
4039 report of this problem mentioned 10 000) and each of them was rewritten,
4040 Exim could use up a very large amount of memory. (It kept on making new
4041 copies of the header line as it rewrote, and never released the old ones.)
4042 At the expense of a bit more processing, the header rewriting function has
4043 been changed so that it no longer eats memory in this way.
4045 66. The generation of the Received: header has been moved from the time that a
4046 message starts to be received, to the time that it finishes. The timestamp
4047 in the Received: header should now be very close to that of the <= log
4048 line. There are two side-effects of this change:
4050 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4051 logged header lines no longer include the local Received: line, because
4052 it has not yet been created. The same applies to a copy of the message
4053 that is returned to a non-SMTP sender when a message is rejected.
4055 (b) When a filter file is tested using -bf, no additional Received: header
4056 is added to the test message. After some thought, I decided that this
4059 This change does not affect the value of $received_for. It is still set
4060 after address rewriting, but before local_scan() is called.
4062 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4064 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4065 gave an unhelpful panic error message, and a defer error. I have managed to
4066 change this behaviour so that it now rejects any supplied certificate,
4067 which seems right, as the list of acceptable certificates is empty.
4069 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4070 gave an unhelpful defer error. I have not managed to make this reject any
4071 supplied certificates, but the error message it gives is "no certificate
4072 supplied", which is not helpful.
4074 70. exigrep's output now also includes lines that are not associated with any
4075 message, but which match the given pattern. Implemented by a patch from
4076 Martin Sluka, which also tidied up the Perl a bit.
4078 71. Recipient callout verification, like sender verification, was using <> in
4079 the MAIL FROM command. This isn't really the right thing, since the actual
4080 sender may affect whether the remote host accepts the recipient or not. I
4081 have changed it to use the actual sender in the callout; this means that
4082 the cache record is now keyed on a recipient/sender pair, not just the
4083 recipient address. There doesn't seem to be a real danger of callout loops,
4084 since a callout by the remote host to check the sender would use <>.
4085 [SEE ABOVE: changed after hitting problems.]
4087 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4088 temporary errors. However, in the case of such a code being given after
4089 the end of a data transmission (i.e. after ".") Exim was failing to write
4090 a retry record for the message. (Yes, there was some broken host that was
4091 actually sending 8xx at this point.)
4093 73. An unknown lookup type in a host list could cause Exim to panic-die when
4094 the list was checked. (An example that provoked this was putting <; in the
4095 middle of a list instead of at the start.) If this happened during a DATA
4096 ACL check, a -D file could be left lying around. This kind of configuration
4097 error no longer causes Exim to die; instead it causes a defer errror. The
4098 incident is still logged to the main and panic logs.
4100 74. Buglet left over from Exim 3 conversion. The message "too many messages
4101 in one connection" was written to the rejectlog but not the mainlog, except
4102 when address rewriting (yes!) was being logged.
4104 75. Added write_rejectlog option.
4106 76. When a system filter was run not as root (that is, when system_filter_user
4107 was set), the values of the $n variables were not being returned to the
4108 main process; thus, they were not subsequently available in the $sn
4111 77. Added +return_path_on_delivery log selector.
4113 78. A connection timeout was being treated differently from recipients deferred
4114 when testing hosts_max_try with a message that was older than the host's
4115 retry timeout. (The host should not be counted, thus allowing all hosts to
4116 be tried at least once before bouncing.) This may have been the cause of an
4117 occasionally reported bug whereby a message would remain on the queue
4118 longer than the retry timeout, but would be bounced if a delivery was
4119 forced. I say "may" because I never totally pinned down the problem;
4120 setting up timeout/retry tests is difficult. See also the next item.
4122 79. The ultimate address timeout was not being applied to errors that involved
4123 a combination of host plus message (for example, a timeout on a MAIL
4124 command). When an address resolved to a number of possible hosts, and they
4125 were not all tried for each delivery (e.g. because of hosts_max_try), a
4126 message could remain on the queue longer than the retry timeout.
4128 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4131 81. Fixed an obscure SMTP outgoing bug which required at least the following
4132 conditions: (a) there was another message waiting for the same server;
4133 (b) the server returned 5xx to all RCPT commands in the first message so
4134 that the message was not completed; (c) the server dropped the connection
4135 or gave a negative response to the RSET that Exim sends to abort the
4136 transaction. The observed case was a dropped connection after DATA that had
4137 been sent in pipelining mode. That is, the server had advertised PIPELINING
4138 but was not implementing it correctly. The effect of the bug was incorrect
4139 behaviour, such as trying another host, and this could lead to a crash.
4145 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4146 and daemon.c were passed as pointers to ints; they should have been
4147 pointers to socklen_t variables (which are typically unsigned ints).
4149 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4152 3. Fixed a really odd bug that affected only the testing scheme; patching a
4153 certain fixed string in the binary changed the value of another string that
4154 happened to be identical to the end of the original first string.
4156 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4157 name", it returns that address as the IP address. On some operating
4158 systems (e.g. Solaris), it also passes back the IP address string as the
4159 "host name". However, on others (e.g. Linux), it passes back an empty
4160 string. Exim wasn't checking for this, and was changing the host name to an
4161 empty string, assuming it had been canonicized.
4163 5. Although rare, it is permitted to have more than one PTR record for a given
4164 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4165 all the names associated with an address, because they do in Solaris.
4166 However, it seems that they do not in Linux for data that comes from the
4167 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4168 I found this out when I moved to a new Linux workstation and tried to run
4169 the Exim test suite.
4171 To get round this problem I have changed the code so that it now does its
4172 own call to the DNS to look up PTR records when searching for a host name.
4173 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4174 addresses that are only in /etc/hosts are still found.
4176 This behaviour is, however, controlled by an option called host_lookup_
4177 order, which defaults to "bydns:byaddr". If people want to use the other
4178 order, or indeed, just use one or the other means of lookup, they can
4179 specify it in this variable.
4181 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4182 some operating systems, this comes back from gethostbyaddr() as an empty
4183 string, and this is what Exim used to test for. However, it seems that in
4184 other systems, "." is yielded. Exim now tests for this case too.
4186 7. The values of check_spool_space and check_log_space are now held internally
4187 as a number of kilobytes instead of an absolute number of bytes. If a
4188 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4189 kilobyte. This means that much larger values can be stored.
4191 8. Exim monitor: an attempt to get the action menu when not actually pointing
4192 at a message produces an empty menu entitled "No message selected". This
4193 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4194 no entries in it ("Shell widget menu has zero width and/or height"). So I
4195 have added a single, blank menu entry in this case.
4197 9. Added ${quote_local_part.
4199 10. MIME decoding is now applied to the contents of Subject: header lines when
4202 11. Now that a reference to $sender_host_address automatically causes a reverse
4203 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4204 host lookup before query-style lookups in lists that might use this
4205 variable. This has therefore been abolished, and the "net-" prefix is no
4206 longer necessary for query-style lookups.
4208 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4209 have been a typo for LFLAGS, so it has been changed.
4211 13. The install script calls Exim with "-C /dev/null" in order to find the
4212 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4213 to be output. Howeve, since Exim outputs its version number before the
4214 error, it didn't break the script. It just looked ugly. I fixed this by
4215 always allowing "-C /dev/null" if the caller is root.
4217 14. Ignore overlarge ACL variable number when reading spool file - insurance
4218 against a later release with more variables having written the file.
4220 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4221 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4222 this, as well as the form without the "IPv6" on the front.
4224 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4225 OS/Makefile-Darwin file.
4227 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4228 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4230 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4231 variable. However, some OS use ints for inodes. I've added cast to long int
4232 to get rid of the compiler warning.
4234 19. I had forgotten to lock out "/../" in configuration file names when
4235 ALT_CONFIG_PREFIX was set.
4237 20. Routers used for verification do not need to specify transports. However,
4238 if such a router generated a host list, and callout was configured, Exim
4239 crashed, because it could not find a port number from the (non-existent)
4240 transport. It now assumes port 25 in this circumstance.
4242 21. Added the -t option to exigrep.
4244 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4245 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4246 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4247 them is set, LOOKUP_LSEARCH is forced.
4249 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4251 24. Added sender and host information to the "rejected by local_scan()" log
4252 line; previously there was no indication of these.
4254 25. Added .include_if_exists.
4256 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4257 Linux. It turns out that not all file systems support this. Apparently some
4258 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4259 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4260 not supported on the file descriptor, is now ignored when Exim is trying to
4261 sync a directory. This applies only to Linux.
4263 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4265 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4266 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4269 29. Added support for maildirsize files from supplied patch (modified a bit).
4271 30. The use of :fail: followed by an empty string could lead Exim to respond to
4272 sender verification failures with (e.g.):
4274 550 Verification failed for <xxx>
4275 550 Sender verify failed
4277 where the first response line was missing the '-' that indicates it is not
4278 the final line of the response.
4280 31. The loop for finding the name of the user that called Exim had a hardwired
4281 limit of 10; it now uses the value of finduser_retries, which is used for
4282 all other user lookups.
4284 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4287 33. Exim was neglecting to zero errno before one call of strtol() when
4288 expanding a string and expecting an integer value. On some systems this
4289 resulted in spurious "integer overflow" errors. Also, it was casting the
4290 result into an int without checking.
4292 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4293 did not work. The code looks as if it has *never* worked, though it appears
4294 to have been documented since at least releast 1.62. I have made it work.
4296 35. The "timeout_DNS" error in retry rules, also documented since at least
4297 1.62, also never worked. As it isn't clear exactly what this means, and
4298 clearly it isn't a major issue, I have abolished the feature by treating it
4299 as "timeout", and writing a warning to the main and panic logs.
4301 36. The display of retry rules for -brt wasn't always showing the error code
4304 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4305 timeout_connect_A, timeout_connect_MX.
4307 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4308 to the empty sender.
4310 39. The daemon was not analysing the content of -oX till after it had closed
4311 stderr and disconnected from the controlling terminal. This meant that any
4312 syntax errors were only noted on the panic log, and the return code from
4313 the command was 0. By re-arranging the code a little, I've made the
4314 decoding happen first, so such errors now appear on stderr, and the return
4315 code is 1. However, the actual setting up of the sockets still happens in
4316 the disconnected process, so errors there are still only recorded on the
4319 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4320 connections (as happens on some IP stacks) was logged at start up time as
4321 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4322 it from "IPv6 and IPv4", which means that two separate sockets are being
4325 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4326 says whether AF_INET or AF_INET6 was passed as an argument.
4328 42. Exiwhat output was messed up when time zones were included in log
4331 43. Exiwhat now gives more information about the daemon's listening ports,
4332 and whether -tls-on-connect was used.
4334 44. The "port" option of the smtp transport is now expanded.
4336 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4337 silently ignored. Now an error message is written to the main and panic
4340 46. There's a new ACL modifier called "logwrite" which writes to a log file
4341 as soon as it is encountered.
4343 47. Added $local_user_uid and $local_user_gid at routing time.
4345 48. Exim crashed when trying to verify a sender address that was being
4348 49. Exim was recognizing only a space character after ".include". It now also
4349 recognizes a tab character.
4351 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4352 extracting the relevant information from the specification. The man page no
4353 longer contains scrambled data for the -d option, and I've added a section
4354 at the front about calling Exim under different names.
4356 51. Added "extra_headers" argument to the "mail" command in filter files.
4358 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4361 53. Installed eximstats 1.29.
4363 54. Added transport_filter_timeout as a generic transport option.
4365 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4366 Cc: header lines. This was required by RFC 822, but it not required by RFC
4369 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4370 incoming messages that did not have them. Now it does so only if the
4371 message originates locally, that is, if there is no associated remote host
4372 address. When Resent- header lines are present, this applies to the Resent-
4373 lines rather than the non-Resent- lines.
4375 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4376 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4378 58. Messages for configuration errors now include the name of the main
4379 configuration file - useful now that there may be more than one file in a
4380 list (.included file names were always shown).
4382 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4383 for those rare installations that do not start the daemon as root or run it
4384 setuid root. I've cut out the call to initgroups() if the daemon is not
4387 60. The Exim user and group can now be bound into the binary as text strings
4388 that are looked up at the start of Exim's processing.
4390 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4392 62. Added $mailstore_basename variable.
4394 63. Installed patch to sieve.c from Michael Haardt.
4396 64. When Exim failed to open the panic log after failing to open the main log,
4397 the original message it was trying to log was written to stderr and debug
4398 output, but if they were not available (the usual case in production), it
4399 was lost. Now it is written to syslog before the two lines that record the
4400 failures to open the logs.
4402 65. Users' Exim filters run in subprocesses under the user's uid. It is
4403 possible for a "deliver" command or an alias in a "personal" command to
4404 provoke an address rewrite. If logging of address rewriting is configured,
4405 this fails because the process is not running as root or exim. There may be
4406 a better way of dealing with this, but for the moment (because 4.30 needs
4407 to be released), I have disabled address rewrite logging when running a
4408 filter in a non-root, non-exim process.
4414 1. The buildconfig auxiliary program wasn't quoting the value set for
4415 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4416 not defined. This bug was masked in 4.22 by the effect that was fixed in
4419 2. Some messages that were rejected after a message id was allocated were
4420 shown as "incomplete" by exigrep. It no longer does this for messages that
4421 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4423 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4424 have allow_domain_literals set, the ID did not get logged in the <= line.
4425 Domain literals are now always recognized in Message-ID: header lines.
4427 4. The first argument for a ${extract expansion item is the key name or field
4428 number. Leading and trailing spaces in this item were not being ignored,
4429 causing some misleading effects.
4431 5. When deliver_drop_privilege was set, single queue runner processes started
4432 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4433 same command in the process it spins off) were not dropping privilege.
4435 6. When the daemon running as "exim" started a queue runner, it always
4436 re-executed Exim in the spun-off process. This is a waste of effort when
4437 deliver_drop_privilege is set. The new process now just calls the
4438 queue-runner function directly.
4444 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4447 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4448 ignored. Though the use of -oP was forcing the writing of a pid file, it
4449 was always written to the default place.
4451 3. If the message "no IP address found for host xxxx" is generated during
4452 incoming verification, it is now followed by identification of the incoming
4453 connection (so you can more easily find what provoked it).
4455 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4457 5. Added some features to "harden" Exim a bit more against certain attacks:
4459 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4460 be put in Local/Makefile. This is like the never_users runtime option,
4461 but it cannot be overridden. The default setting is "root".
4463 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4464 prefix string with which any file named in a -C command line option
4467 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4468 is retained for -C and -D only if the caller of Exim is root. Without
4469 it, the exim user may also use -C and -D and retain privilege.
4471 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4472 command line option is disabled.
4474 6. Macro names set by the -D option must start with an upper case letter, just
4475 like macro names defined in the configuration file.
4477 7. Added "dereference=" facility to LDAP.
4479 8. Two instances of the typo "uknown" in the source files are fixed.
4481 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4482 the Configure-Makefile script screwed up while processing it.
4484 10. Incorporated PCRE 4.4.
4486 11. The SMTP synchronization check was not operating right at the start of an
4487 SMTP session. For example, it could not catch a HELO sent before the client
4488 waited for the greeting. There is now a check for outstanding input at the
4489 point when the greeting is written. Because of the duplex, asynchronous
4490 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4491 way, but not yet received, when the check is performed.
4493 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4494 on TCP/IP sockets, because this apparently causes some broken clients to
4497 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4498 unchanged) from the Cygwin maintainer.
4500 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4501 was supported instead of testing for mbx. Effectively a typo.
4503 15. The spa authenticator server code was not checking that the input it
4504 received was valid base64.
4506 16. The debug output line for the "set" modifier in ACLs was not showing the
4507 name of the variable that was being set.
4509 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4511 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4512 lookup on its own. Something else was needed to trigger the lookup. For
4513 example, a match in host_lookup or the need for a host name in a host list.
4514 Now, if $sender_host_name is referenced and the host name has not yet been
4515 looked up, a lookup is performed. If the lookup fails, the variable remains
4516 empty, and $host_lookup_failed is set to "1".
4518 19. Added "eqi" as a case-independent comparison operator.
4520 20. The saslauthd authentication condition could segfault if neither service
4521 nor realm was specified.
4523 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4524 error message that was logged was misleading, and incoming SMTP
4525 connections were dropped. The message is now more accurate, and temporary
4526 errors are given to SMTP connections.
4528 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4529 (except RSET) with a 421 error, until QUIT is received. However, it was
4530 failing to send a response to QUIT.
4532 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4533 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4534 ACL is now run after the helo_try_verify_hosts code.
4536 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4537 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4538 case-independent, so other case variants are also recognized). Apparently
4539 some systems use these upper case variants.
4541 25. If more than two messages were waiting for the same host, and a transport
4542 filter was specified for the transport, Exim sent two messages over the
4543 same TCP/IP connection, and then failed with "socket operation on non-
4544 socket" when it tried to send the third.
4546 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4548 27. The extern definition of crypt16() in expand.c was not being excluded when
4549 the OS had its own crypt16() function.
4551 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4552 as a preferred synonym for return_size_limit, both as an option and as an
4555 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4557 30. Changed the default configuration ACL to relax the local part checking rule
4558 for addresses that are not in any local domains. For these addresses,
4559 slashes and pipe symbols are allowed within local parts, but the sequence
4560 /../ is explicitly forbidden.
4562 31. SPA server authentication was not clearing the challenge buffer before
4565 32. log_message in a "warn" ACL statement was writing to the reject log as
4566 well as to the main log, which contradicts the documentation and doesn't
4567 seem right (because no rejection is happening). So I have stopped it.
4569 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4570 However, I am unable to do any testing of this.
4572 34. Fixed an infelicity in the appendfile transport. When checking directories
4573 for a mailbox, to see if any needed to be created, it was accidentally
4574 using path names with one or more superfluous leading slashes; tracing
4575 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4577 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4578 contents are added to the log line that is written for every discarded
4579 recipient. (Previously a log_message setting was ignored.)
4581 36. The ${quote: operator now quotes the string if it is empty.
4583 37. The install script runs exim in order to find its version number. If for
4584 some reason other than non-existence or emptiness, which it checks, it
4585 could not run './exim', it was installing it with an empty version number,
4586 i.e. as "exim-". This error state is now caught, and the installation is
4589 38. An argument was missing from the function that creates an error message
4590 when Exim fails to connect to the socket for saslauthd authentication.
4591 This could cause Exim to crash, or give a corrupted message.
4593 39. Added isip, isip4, and isip6 to ${if conditions.
4595 40. The ACL variables $acl_xx are now saved with the message, and can be
4596 accessed later in routers, transports, and filters.
4598 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4599 strings in the file are not string-expanded.
4601 42. If a MAIL command specified a SIZE value that was too large to fit into an
4602 int variable, the check against message_size_limit failed. Such values are
4603 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4604 day this will have to be increased, but I don't think I want to be around
4605 when emails are that large.
4612 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4613 iconv() is not standard in FreeBSD.
4615 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4616 IPv6 enabled. The observed symptom was a segmentation fault on return from
4617 the function os_common_find_running_interfaces() in src/os.c.
4619 3. In the check_special_case() function in daemon.c I had used "errno" as an
4620 argument name, which causes warnings on some systems. This was basically a
4621 typo, since it was named "eno" in the comments!
4623 4. The code that waits for the clock to tick (at a resolution of some fraction
4624 of a second) so as to ensure message-id uniqueness was always waiting for
4625 at least one whole tick, when it could have waited for less. [This is
4626 almost certainly not relevant at current processor speeds, where it is
4627 unlikely to ever wait at all. But we try to future-proof.]
4629 5. The function that sleeps for a time interval that includes fractions of a
4630 second contained a race. It did not block SIGALRM between setting the
4631 timer, and suspending (a couple of lines later). If the interval was short
4632 and the sigsuspend() was delayed until after it had expired, the suspension
4633 never ended. On busy systems this could lead to processes getting stuck for
4636 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4637 process, before it forks any delivery processes. The open lookup caching
4638 mechanism meant that the open file or database connection was passed into
4639 the delivery process. The problem was that delivery processes always tidy
4640 up cached lookup data. This could cause a problem for the next delivery
4641 process started by the queue runner, because the external queue runner
4642 process does not know about the closure. So the next delivery process
4643 still has data in the lookup cache. In the case of a file lookup, there was
4644 no problem because closing a file descriptor in a subprocess doesn't affect
4645 the parent. However, if the lookup was caching a connection to a database,
4646 the connection was closed, and the second delivery process was likely to
4647 see errors such as "PGSQL: query failed: server closed the connection
4648 unexpectedly". The problem has been fixed by closing all cached lookups
4649 in a queue runner before running a delivery process.
4651 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4652 seem to have the "const" qualifier which it has on other OS. I've
4655 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4656 *of the same type* (client or server) with the same public name that an
4657 error should be diagnosed.
4659 9. When Exim looked up a host name for an IP address, but failed to find the
4660 original IP address when looking up the host name (a safety check), it
4661 output the message "<ip address> does not match any IP for NULL", which was
4662 confusing, to say the least. The bug was that the host name should have
4663 appeared instead of "NULL".
4665 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4666 user that is built into the binary, and the -C or -D options is used, root
4667 privilege is dropped before the configuration file is read. In addition,
4668 logging is switched to stderr instead of the normal log files. If the
4669 configuration then re-defines the Exim user, the unprivileged environment
4670 is probably not what is expected, so Exim logs a panic warning message (but
4673 However, if deliver_drop_privilege is set, the unprivileged state may well
4674 be exactly what is intended, so the warning has been cut out in that case,
4675 and Exim is allowed to try to write to its normal log files.
4681 1. smtp_return_error_details was not giving details for temporary sender
4682 or receiver verification errors.
4684 2. Diagnose a configuration error if two authenticators have the same public
4687 3. Exim used not to create the message log file for a message until the first
4688 delivery attempt. This could be confusing when incoming messages were held
4689 for policy or load reasons. The message log file is now created at the time
4690 the message is received, and an initial "Received" line is written to it.
4692 4. The automatically generated man page for command line options had a minor
4693 bug that caused no ill effects; however, a more serious problem was that
4694 the procedure for building the man page automatically didn't always
4695 operate. Consequently, release 4.20 contains an out-of-date version. This
4696 shouldn't happen again.
4698 5. When building Exim with embedded Perl support, the script that builds the
4699 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4700 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4702 6. The freeze_tell option was not being used for messages that were frozen on
4703 arrival, either by an ACL or by local_scan().
4705 7. Added the smtp_incomplete_transaction log selector.
4707 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4708 was accepting AUTH without a new EHLO.
4710 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4711 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4714 10. Logging of TCP/IP connections (when configured) now happens in the main
4715 daemon process instead of the child process, so that the TCP/IP connection
4716 count is more accurate (but it can never be perfect).
4718 11. The use of "drop" in a nested ACL was not being handled correctly in the
4719 outer ACL. Now, if condition failure induced by the nested "drop" causes
4720 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4721 or "require"), the connection is dropped.
4723 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4724 that yield "discard" can now be used with an "accept" or a "discard" verb,
4725 but an error is generated for any others (because I can't see a useful way
4726 to define what should happen).
4728 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4729 are now processed in the same way as lines in the Exim configuration file.
4730 In particular, continuation lines are supported.
4732 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4734 15. Added -ti meaning -t -i.
4736 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4737 domains, and warn by logging if others are found.
4739 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4740 returned from GIFCONF when Exim is trying to find the list of interfaces on
4741 a host. The code in os.c has been modified to copy each ifreq to an aligned
4742 structure in all cases.
4744 Also, in some cases, the returned ifreq's were being copied to a 'struct
4745 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4746 means the last couple of bytes of an IPv6 address could be chopped if the
4747 ifreq contained only a normal sockaddr (14 bytes storage).
4749 18. Named domain lists were not supported in the hosts_treat_as_local option.
4750 An entry such as +xxxx was not recognized, and was treated as a literal
4753 19. Ensure that header lines added by a DATA ACL are included in the reject log
4754 if the ACL subsequently rejects the message.
4756 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4757 MD5 (which is deprecated).
4759 21. When testing a filter file using -bf, Exim was writing a message when it
4760 took the sender from a "From " line in the message, but it was not doing so
4761 when it took $return_path from a Return-Path: header line. It now does.
4763 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4764 with a valid header line field name (a series of printing characters
4765 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4767 23. Changed "disc" in the source to "disk" to conform to the documentation and
4768 the book and for uniformity.
4770 24. Ignore Sendmail's -Ooption=value command line item.
4772 25. When execve() failed while trying to run a command in a pipe transport,
4773 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4774 could be confused with a return value of 69 from the command itself. This
4775 has been changed to 127, the value the shell returns if it is asked to run
4776 a non-existent command. The wording for the related log line suggests a
4777 non-existent command as the problem.
4779 26. If received_header_text expands to an empty string, do not add a Received:
4780 header line to the message. (Well, it adds a token one on the spool, but
4781 marks it "old" so that it doesn't get used or transmitted.)
4783 27. Installed eximstats 1.28 (addition of -nt option).
4785 28. There was no check for failure on the call to getsockname() in the daemon
4786 code. This can fail if there is a shortage of resources on the system, with
4787 ENOMEM, for example. A temporary error is now given on failure.
4789 29. Contrary to the C standard, it seems that in some environments, the
4790 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4791 program. Exim now does this explicitly; it affects the formatting of
4792 timestamps using strftime().
4794 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4795 complaints. I've now initialized all the variables, to avoid this.
4797 32. Header lines added by a system filter were not being "seen" during
4798 transport-time rewrites.
4800 33. The info_callback() function passed to OpenSSL is set up with type void
4801 (*)(SSL *, int, int), as described somewhere. However, when calling the
4802 function (actually a macro) that sets it up, the type void(*)() is
4803 expected. I've put in a cast to prevent warnings from picky compilers.
4805 34. If a DNS black list lookup found a CNAME record, but there were no A
4806 records associated with the domain it pointed at, Exim crashed.
4808 35. If a DNS black list lookup returned more than one A record, Exim ignored
4809 all but the first. It now scans all returned addresses if a particular IP
4810 value is being sought. In this situation, the contents of the
4811 $dnslist_value variable are a list of all the addresses, separated by a
4814 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4815 to accept a host name and all its aliases if the forward lookup for any of
4816 them yielded the IP address of the incoming connection. Now it accepts only
4817 those names whose forward lookup yields the correct IP address. Any other
4818 names are discarded. This closes a loophole whereby a rogue DNS
4819 administrator could create reverse DNS records to break through a
4820 wildcarded host restriction in an ACL.
4822 37. If a user filter or a system filter that ran in a subprocess used any of
4823 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4824 the wrong values were passed to the pipe command ($thisaddress had the
4825 value of $0, $0 had the value of $1, etc). This bug was introduced by
4826 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4828 38. Improved the line breaking for long SMTP error messages from ACLs.
4829 Previously, if there was no break point between 40 and 75 characters, Exim
4830 left the rest of the message alone. Two changes have been made: (a) I've
4831 reduced the minimum length to 35 characters; (b) if it can't find a break
4832 point between 35 and 75 characters, it looks ahead and uses the first one
4833 that it finds. This may give the occasional overlong line, but at least the
4834 remaining text gets split now.
4836 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4837 file descriptors might be low, and that setting 1000 would always raise it.
4838 It turns out that in some environments, the limit is already over 1000 and
4839 that lowering it causes trouble. So now Exim takes care not to decrease it.
4841 40. When delivering a message, the value of $return_path is set to $sender_
4842 address at the start of routing (routers may change the value). By an
4843 oversight, this default was not being set up when an address was tested by
4844 -bt or -bv, which affected the outcome if any router or filter referred to
4847 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4848 log only when "message" or "log_message" are set. However, if one of the
4849 conditions was an address verification, or a call to a nested ACL, the
4850 messages generated by the underlying test were being passed through. This
4851 no longer happens. The underlying message is available in $acl_verify_
4852 message for both "message" and "log_message" expansions, so it can be
4853 passed through if needed.
4855 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4856 new expansion $bh_ to give the encoded byte string without charset
4857 translation. Translation happens only if iconv() is available; HAVE_ICONV
4858 indicates this at build time. HEADERS_CHARSET gives the charset to
4859 translate to; headers_charset can change it in the configuration, and
4860 "headers charset" can change it in an individual filter file.
4862 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4863 that creates RFC 2047 encoded "words" labels them as that charset instead
4864 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4865 expansion operator; (ii) when Exim creates a From: line for a local
4866 message; (iii) when a header line is rewritten to include a "phrase" part.
4868 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4869 buggy, causing it to skip the first lines of messages whose message ID
4870 ended in 'D'. This would not have bitten before Exim release 4.14, because
4871 message IDs were unlikely to end in 'D' before then. The effect was to have
4872 incorrect size information for certain domains.
4874 45. #include "config.h" was missing at the start of the crypt16.c module. This
4875 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4878 46. If there was a timeout during a "random" callout check, Exim treated it as
4879 a failure of the random address, and carried on sending RSET and the real
4880 address. If the delay was just some slowness somewhere, the response to the
4881 original RCPT would be taken as a response to RSET and so on, causing
4882 mayhem of various kinds.
4884 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4885 when I implemented it. It didn't allow for the fact that some option values
4886 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4887 the right test for positive values.
4889 48. Domain names in DNS records are case-independent. Exim always looks them up
4890 in lower case. Some resolvers return domain names in exactly the case they
4891 appear in the zone file, that is, they may contain uppercase letters. Not
4892 all resolvers do this - some return always lower case. Exim was treating a
4893 change of case by a resolver as a change of domain, similar to a widening
4894 of a domain abbreviation. This triggered its re-routing code and so it was
4895 trying to route what was effectively the same domain again. This normally
4896 caused routing to fail (because the router wouldn't handle the domain
4897 twice). Now Exim checks for this case specially, and just changes the
4898 casing of the domain that it ultimately uses when it transmits the message
4901 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4904 50. If a filter generated a file delivery with a non-absolute name (possible if
4905 no home directory exists for the router), the forbid_file option was not
4908 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4909 the existing equality matching.
4911 52. Exim was using ints instead of ino_t variables in some places where it was
4912 dealing with inode numbers.
4914 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4915 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4916 called TMPDIR, and if it finds it is different, it changes its value.
4918 54. The smtp_printf() function is now made available to local_scan() so
4919 additional output lines can be written before returning. There is also an
4920 smtp_fflush() function to enable the detection of a dropped connection.
4921 The variables smtp_input and smtp_batched_input are exported to
4924 55. Changed the default runtime configuration: the message "Unknown user"
4925 has been removed from the ACL, and instead placed on the localuser router,
4926 using the cannot_route_message feature. This means that any verification
4927 failures that generate their own messages won't get overridden. Similarly,
4928 the "Unrouteable address" message that was in the ACL for unverifiable
4929 relay addresses has also been removed.
4931 56. Added hosts_avoid_esmtp to the smtp transport.
4933 57. The exicyclog script was not checking for the esoteric option
4934 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4935 will work only if exicyclog is run under the appropriate euid.
4937 58. Following a discussion on the list, the rules by which Exim recognises line
4938 endings on incoming messages have been changed. The -dropcr and drop_cr
4939 options are now no-ops, retained only for backwards compatibility. The
4940 following line terminators are recognized: LF CRLF CR. However, special
4941 processing applies to CR:
4943 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4944 nor a local message in the state where . is a terminator.
4946 (ii) If a bare CR is encountered in a header line, an extra space is added
4947 after the line terminator so as not to end the header. The reasoning
4948 behind this is that bare CRs in header lines are most likely either
4949 to be mistakes, or people trying to play silly games.
4951 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4952 was being incorrectly given as 18 bytes larger than it should have been.
4953 This is a VOB (very old bug).
4955 60. This may never have affected anything current, but just in case it has:
4956 When the local host is found other than at the start of a list of hosts,
4957 the local host, those with the same MX, and any that follow, are discarded.
4958 When the list in question was part of a longer list of hosts, the following
4959 hosts (not currently being processed) were also being discarded. This no
4960 longer happens. I'm not sure if this situation could ever has previously
4963 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4966 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4967 header. This is something that is recommended in a new Internet Draft, and
4968 is something that is documented as being done by Sendmail. There are two
4969 possible values. For messages generated by the autoreply transport, Exim
4972 Auto-Submitted: auto-replied
4974 whereas for all other generated messages (e.g. bounces) it adds
4976 Auto-Submitted: auto-generated
4978 63. The "personal" condition in filters now includes a test for the
4979 Auto-Submitted: header. If it contains the string "auto-" the message it
4980 not considered personal.
4982 64. Added rcpt_include_affixes as a generic transport option.
4984 65. Added queue_only_override (default true).
4986 66. Added the syslog_duplication option.
4988 67. If what should have been the first header line of a message consisted of
4989 a space followed by a colon, Exim was mis-interpreting it as a header line.
4990 It isn't of course - it is syntactically invalid and should therefore be
4991 treated as the start of the message body. The misbehaviour could have
4992 caused a number of strange effects, including loss of data in subsequent
4993 header lines, and spool format errors.
4995 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4996 client host had authenticated. This control can now be exercised by an ACL
4997 for more flexibility.
4999 69. By default, callouts do not happen when testing with -bh. There is now a
5000 variant, -bhc, which does actually run the callout code, including
5001 consulting and updating the callout cache.
5003 70. Added support for saslauthd authentication, courtesy of Alexander
5006 71. If statvfs() failed on the spool or log directories while checking their
5007 size for availability, Exim confusingly gave the error "space shortage".
5008 Furthermore, in debugging mode it crashed with a floating point exception.
5009 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5010 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5011 serious problem, Exim now writes to the main and panic logs when this
5012 happens, with details of the failure. It then refuses to accept the
5013 incoming message, giving the message "spool directory problem" or "log
5014 directory problem" with a 421 code for SMTP messages.
5016 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5017 0, 1, and 2 exist, because some OS complain for execs without them (see
5018 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5019 descriptors. However, the code omitted to check that the open succeeded,
5020 causing mysterious errors if for some reason the permissions on /dev/null
5021 got screwed. Now Exim writes a message to the main and panic logs, and
5022 bombs out if it can't open /dev/null.
5024 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5025 interact so that it is all more flexible. It is supposed to remain
5026 backwards compatible. Also added extra_local_interfaces.
5028 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5029 to bomb out with an assertion failure - to the client this appears as a
5030 connection drop. This problem occurs in the part of the code that was taken
5031 from the Samba project. Fortunately, the assertion is in a very simple
5032 function, so I have fixed this by reproducing the function inline in the
5033 one place where it is called, and arranging for authentication to fail
5034 instead of killing the process with assert().
5036 75. The SPA client code was not working when the server requested OEM rather
5037 than Unicode encoding.
5039 76. Added code to make require_files with a specific uid setting more usable in
5040 the case where statting the file as root fails - usually a non-root-mounted
5041 NFS file system. When this happens and the failure is EACCES, Exim now
5042 forks a subprocess and does the per-uid checking as the relevant uid.
5044 77. Added process_log_path.
5046 78. If log_file_path was not explicitly set, a setting of check_log_space or
5047 check_log_inodes was ignored.
5049 79. If a space check for the spool or log partitions fails, the incident is now
5050 logged. Of course, in the latter case the data may get lost...
5052 80. Added the %p formatting code to string_format() so that it can be used to
5053 print addresses in debug_print(). Adjusted all the address printing in the
5054 debugging in store.c to use %p rather than %d.
5056 81. There was a concern that a line of code in smtp_in.c could overflow a
5057 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5058 initially expressed, the concern was not well-founded, because trailing
5059 spaces are removed early. However, if the trailing spaces were followed by
5060 a NULL, they did not get removed, so the overflow was possible. Two fixes
5063 (a) I re-wrote the offending code in a cleaner fashion.
5064 (b) If an incoming SMTP command contains a NULL character, it is rejected
5067 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5068 runs initgroups(), so that if the Exim user is in any additional groups,
5069 they will be used during message reception.
5075 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob